Good source for amplifier tips
Will, KN6DV
For those who are wondering what the "EIMAC" letter is all about, you will
find a copy of it on the page:
--
http://www.av.qnet.com/~kn6dv
Also you will find the archive of the "Good source for amplifier tips".
REFORM
Unless the reformer can invent something which substisutes attractive
virtues for atractive vices, he will fail.
(Walter Lippmann)
w8j...@aol.com
Hi Rich,
In a message dated 96-11-12 02:48:42 EST, you write:
>
>According to Eimac CURVE 005211, "Eimac 3CX800A7 Constant Current
>Characteristics (Grounded Grid)":
> around 27vrms of cathode-to-grid driving voltage produces the maximum
>anode/plate current of 600mA.
>P=I times E. 27vrms times 150mA = 4.05w(rms).
Rich. This is unbelievable. You have no idea how to use a constant current
characteristic graph, and you are writing articles implying everyone else
in the world doesn't know what they are doing!
The failure to understand simple physics, like photons or the mechanisims
of heat generation are one thing, but misapplying a simple operating graph
is unbelievable!
When the grid current is maximum, anode voltage is below zero in a
mistuned PA. These conditions are not even plotted on the graph, because
they are not obtainable in steady state conditions without destroying the
tube. The graph would be five times larger, or the useful operating area
would be condensed to the point it was unusable with the performance
calculator overlay, if Eimac included abnormal operating ranges in the
data.
You aren't even calculating the values obtained in normal operation
correctly, because you are mixing peak and RMS values.
Let me give an example of one thing (out of several) you have done wrong.
You assumed the average RMS grid current indicated on the meter (150 mA)
is the value applied to the graph. The peak grid current is many times
that value, and varies throughout the RF cycle. For over half the input RF
cycle, the grid is biased out of conduction. The cathode is positive with
respect to the grid, and current flow in the grid is zero.
As the cathode RF driving waveform crosses zero and exceeds the bias
voltage in a negative direction, the grid slowy begins to draw current.
The grid meter displays this value integrated over the entire RF cycle, so
peak grid current is many many times the value you used, it is far off the
curves by the time the meter reads 150 mA!
You selected 27 VRMS, even though that value is low for the cathode
voltage that might be obtained in a typical properly loaded circuit, and
applied it to the average RMS grid current indicated by the meter. Yet
that voltage is clearly not the G-K voltage in a mistuned condition at
all, since the driving impedance and grid voltage soars when anode load
impedance is raised!
Remember, negative feedback is removed when the anode load impedance is
raised, cathode voltage soars. The vast majority of the exciter's power
is diverted to the control grid. Since this condition is NOT indicated on
a typical operation graph, the design engineer needs to measure it. The
grid meter does indicate the correct effective current applied to
dissipation, but your voltage figure is a factor of almost five times off.
You have probably mixed steady state impedance and average driving
impedances as well.
>Tom: Is 4.05w "almost 20 watts" in your opinion?
No. Since you obviously don't understand the dynamics of a PA, the
dissipation is not 4.05 watts either!
>> And yes, it will cause the gold to ball up, evaporate and even flake
off.
>> And yes, Eimac will confirm this.
>>
>When are you going to publish this confirmation from Eimac that supports
>your position, Mr. Rauch?
QST already in effect did that on page 71 of the Sept 1994 issue Eimac's
letter says:
"Oxide cathode tubes using focus cathode design are more easily damaged in
a short time under fault conditions. Newly developed circuits are quite
effective in protecting tubes from excessive grid dissipation, which can
result from loss of load due to failure in the feedline, balun, or
antenna."
If that alone isn't enough to support my position that improper loading
can quickly damage the grid, there is more. They go on to say your grid
circuit suggestions are "ludicrous" (that description also seems to fit
your suggestions that photons destroy tubes in PA's that are on standby),
and that they have yet to see evidence of the parasitic oscillation damage
claims you make.
Let this parasitic fantasia go quietly away Rich. I have no desire to beat
you into the ground technically, and you will remain unbendable no matter
what evidence is presented.
The tuning capacitor measures a very low Xc over the VHF and UHF range,
yet you claim it arcs from VHF energy. The anode impedance actually moves
the wrong way when your suppressors are installed, yet you claim they
improve things. Resistors make horrid fuses (and fuses are bad enough) yet
you recommend removing electronic protection and adding resistors to
protect a grid. Eimac rates a grid at 25 watts, and you say it takes 500
watts. Eimac rates a cathode at many tens of amperes of peak current, and
yet you say a slight amount of overdrive damages it. You claim a parasitic
can drive a tube into super energy oscillation, yet claim the same tube
flat tops when overdriven by a few percent. You claim a photon can drive a
gas-less tube on standby into conduction, yet current in a gas filled
geiger-muller tube is in fractions of a pico-ampere.
Everything you say contradicts science, the manufacturer of tubes, and
even yourself.
Shame on the publisher for letting this stuff ever get started. I hope
they've learned a lesson about technical editing. Let it die Rich, let's
both do something more productive.
73 Tom
Tom C. Brown, Jr.
>>No. Since you obviously don't understand the dynamics of a PA, the
dissipation is not 4.05 watts either!<<
Tom, it was you who said that the grid dissipation was "about 4 watts".
Not Rich.
Read back through your own postings and you will see that I am right
about that.
Were you wrong then, or are you wrong now? Or is the cat both alive
*and* dead? :-)
>>They go on to say your grid
circuit suggestions are "ludicrous" (that description also seems to fit
your suggestions that photons destroy tubes in PA's that are on
standby),
and that they have yet to see evidence of the parasitic oscillation
damage
claims you make.<<
They, they, they..... WHO are "they"? So far, I have seen one, yes
*one* individual at Eimac named here (by you). No more. Your use of
the term "they" indicates more than one. Please supply the names of the
rest of "they". Or stop saying "they".
As I said before, you and Rich are "1 to 1" in named sources so far.
Why won't you tell us your sources' names? (Other than Mr. Brandon)
I'm not saying that anything you have said is wrong, but you do seem to
be keeping a smoke screen between us and the identities of your
sources...... And where there is smoke..... ;-)
73 and thanks again for the great discussion!
Tom KJ5IE
R. L. Measures
In article <19961112163...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> Hi Rich,
>
> In a message dated 96-11-12 02:48:42 EST, you write:
> >
> >According to Eimac CURVE 005211, "Eimac 3CX800A7 Constant Current
> >Characteristics (Grounded Grid)":
> > around 27vrms of cathode-to-grid driving voltage produces the maximum
> >anode/plate current of 600mA.
> >P=I times E. 27vrms times 150mA = 4.05w(rms).
>
> Rich. This is unbelievable. You have no idea how to use a constant current
> characteristic graph, and you are writing articles implying everyone else
> in the world doesn't know what they are doing!
Hmmm. There seems to be another person writing articles who uses my name.
Please direct me to a specific issue of a magazine where one of these
counterfeit articles appears.
> The failure to understand simple physics, like photons .or the mechanisims
> of heat generation are one thing, but misapplying a simple operating graph
> is unbelievable!
The lady doeth protest too much, me thinks.
> When the grid current is maximum, anode voltage is below zero in a
> mistuned PA. These conditions are not even plotted on the graph, because
> they are not obtainable in steady state conditions without destroying the
> tube. The graph would be five times larger, or the useful operating area
> would be condensed to the point it was unusable with the performance
> calculator overlay, if Eimac included abnormal operating ranges in the
> data.
It seems a bit odd that Eimac does not discuss this alleged phenomenon in
*Care and Feeding of Power Grid Tubes*. Certainly, in tests performed on
my own amplifiers, I was unable to reproduce the incredulous peak anode
voltages you said would appear during mistuning of the output network.
> You aren't even calculating the values obtained in normal operation
> correctly, because you are mixing peak and RMS values.
> Let me give an example of one thing (out of several) you have done wrong.
> You assumed the average RMS grid current indicated on the meter (150 mA)
is the value applied to the graph. The peak grid current is many times
that value, and varies throughout the RF cycle
. ...snip...
I already know that the grid, anode, and cathode currents vary during the
cycle. It is abundantly clear that this is what the constant current
curves are all about. However, Eimac's grid dissipation rating is Not
instantaneous watts. The grid dissipation rating is average watts.
In a post dated 3 Nov., Mr. Rauch states:
> Don't overcomplicate grid dissipation Gary. It's given simply as the
> voltage times current at any given instant of time, and the resulting
> average value of dissipation over time. I explained the calculation. It is
> approximated by taking the average voltage grid to cathode voltage over
> the portion of the cycle when the grid is positive compared to the cathode
> (you can use a polarity sensitive and dc biased RMS RF voltmeter to
> measure this), and the average current (which is measured for you on the
> meter).
Tom: you stated that that the meter-indicated DC grid current was valid in
such matters. Certainly, the cathode driving voltage is a sinewave, whose
peak value is around 38v, according to Eimac's graph. The RMS value of a
38v peak sinewave is in the neighborhood of 27v, is it not?
...snip...
> No. Since you obviously don't understand the dynamics of a PA, the
> dissipation is not 4.05 watts either!
According to Tom Rauch, it should be pretty close to 4.05w.
> >> And yes, it will cause the gold to ball up, evaporate and even flake
> off.
> >> And yes, Eimac will confirm this.
> >>
> >When are you going to publish this confirmation from Eimac that supports
> >your position, Mr. Rauch?
>
> QST already in effect did that on page 71 of the Sept 1994 issue Eimac's
> letter says:
There is no official letter from Eimac on page 71 in my copy of the
September 1994 issue of QST. . Am I overlooking a correction or addendum
that appeared in a subsequent issue of QST?
Since we are on the subject of the unique-critique by the "recognized
experts" (your words, Mr. Rauch), Paul Pagel told me that although his name
was on the Technical Correspondence column in the Sept. 1994 issue of QST,
he had NOTHING to do with the column. Indeed, this was unique in itself.
More than a few readers came away with the impression that the
unique-critique was more personal than scientific. I see that Ameritron
was on your "recognized experts" list.
......Congratulations on being recognized as an amplifier expert, Mr. Rauch.
--------------------------
e-mail copy to Mr. Rauch
--
--Rich-- ag6k, 805.386.3734
w8j...@aol.com
Hi Rich,
In a message dated 96-11-14 05:52:20 EST, you write:
>It seems a bit odd that Eimac does not discuss this alleged phenomenon in
>*Care and Feeding of Power Grid Tubes*. Certainly, in tests performed on
>my own amplifiers, I was unable to reproduce the incredulous peak anode
>voltages you said would appear during mistuning of the output network.
If you look at other threads in this homebrew area, you'll find Ray Mack
who observes this effect. Here's what S56A posted:
In a message dated 96-04-14 12:54:03 EDT, Mario wrote:
>Subj: Perpetual motion
>Date: 96-04-14 12:54:03 EDT
>From: s5...@s55tcp.ampr.org (Marijan Miletic)
>To: cq-co...@tgv.com
>CC: W8J...@aol.com
>
>Tom, W8JI refrained from further discussion of exotic QRO topics but in
>a private email convinced me to repeat his tests on my good, old SB-220.
>
>I made simple 10/1000pF = 1:100 divider from tune cap, used 1N4148 diode
>for a peak rectifier and achieved almost the same results as he did!
>I was initially suspecting that some measurement error was involved...
>
>>I have facilities to measure PA stages, and an actual SB-220.
>>With the amplifier in the SSB position on 7 MHz at 1100 watts output I
>>measured 2.4 kV peak on the tuning cap. When I mistuned the amp by fully
>>meshing the loading cap and adjusting the tuning cap for maximum voltage
>>I measured 3.7 kV peak. The supply voltage was ~ 3kV when mistuned.
>Under above conditions, SB-220 tuning capacitor tested for 3.5kV may arc!
Rich writes:
>I already know that the grid, anode, and cathode currents vary during the
>cycle. It is abundantly clear that this is what the constant current
>curves are all about. However, Eimac's grid dissipation rating is Not
>instantaneous watts. The grid dissipation rating is average watts.
Once again you are confused Rich. You claim understanding, and then
misconstrue results. I plainly explained you are using the correct
(indicated on the meter) RMS current, but the voltage is in error. We can
NOT use the voltage at that point of 150 mA with a positive anode voltage
on the graph, because that is not the operating point. Calculating the
true average RMS cathode RF voltage would require a set of curves far off
the graph!
If the grid is only reaching 150 mA peak on the constant current
characteristic curves graph, it may very well be operating at normal or
near normal values as an RF PA.
You seem to ignore the fact the cathode is in series with the PA output,
and when the tank is underloaded the cathode grid voltage soars and the
anode voltage at peaks goes beyond the plotted values.
>Tom: you stated that that the meter-indicated DC grid current was valid
in
>such matters. Certainly, the cathode driving voltage is a sinewave,
whose
>peak value is around 38v, according to Eimac's graph. The RMS value of a
>38v peak sinewave is in the neighborhood of 27v, is it not?
No. You are mis-using the curves. You are taking the peak current at one
point of the operating cycle, using the cathode to grid voltage at that
point with 135 volts or so of anode/grid voltage, and claiming a complete
analysis. You also made the leap to a sine wave, but without a Fourier or
Chaffee analysis you have no idea what the content of the waveform is.
Eimac's constant current curves do not show grossly abnormal operation,
and the plot would have to include peak grid currents at least in the
range of 500-600 mA down to zero under the conditions of an indicated
current of 150 mA. That's because the grid cathode conduction angle is now
much less than 180 degrees (and it has grid bias to overcome), and the G-K
voltage increases when negative feedback is removed.
If it were a 180 degrees conduction angle sine wave in a fixed resistance,
peak grid current would be (150*2.818).
Look carefully at the curves Rich. The tube is not a linear resistance as
operating parameters are changed. When the grid is overdriven, the
waveform squares (more higher frequency sine waves appear in the Fourier
analysis). Energy content increases rapidly, and the 1.414 peak to average
current and voltage ratios do not hold true. We have no idea what it is
when the peak currents are far off the performance curves given.
How you can misapply this simple graph of constant current charateristics
as a linear time constant resistance is fascinating. The curves plainly
indicate non-linear voltage/resistance changes and great interaction
between the elements of the tube.
If that isn't enough, or if it confuses you, consider this. The exciter
supplying 100 watts, and with very low anode current it can't contribute
to the RF output. Where does excess drive power, that used to be
feedthrough power, go?
In article <328A56...@mail.teclink.net>, "Tom C. Brown, Jr."
<mad...@mail.teclink.net> writes:
>
>Tom, it was you who said that the grid dissipation was "about 4 watts".
>Not Rich.
Tom, you have taken something far from context. I have no time for these
little "word games". Rich plainly asked me what the grid dissipation
rating of a 3CX800 was, and I replied around 4 watts. After I repeated 4
watts several tmes, he said he meant an 8877, and it was 25 watts. Let's
not make this a word game debate by pulling things from the context in
which they were said.
I PLAINLY stated the dissipation of the 3CX800 was just under 20 watts
when the PA I was testing was mistuned to indicated 150 mA of tube grid
current. Read back, rather than wasting time going over the same stuff.
>> QST already in effect did that on page 71 of the Sept 1994 issue
Eimac's
>> letter says:
>
>There is no official letter from Eimac on page 71 in my copy of the
>September 1994 issue of QST. . Am I overlooking a correction or addendum
>that appeared in a subsequent issue of QST?
I guess then Rich, we will have to accept your word. As we accept your
word on many other "facts":
1.) Eimac rates the 8877 grid at 25 watts, but you know it takes almost
500 watts to damage it.
2.) A geiger counter has fractions of a pico-amp in the gas filled
detector tube, but a vacuum filled 3-500 can draw amperes from the same
effect.
3.) Gold melts from I^2 *R heating, and not from the E*I of the element,
yet that works ok for anode dissipation.
4.) Reid Brandon is not the applications engineer for power grid tubes,
yet Bill Foote is.
5.) The tank capacitor has a very low Z at VHF/UHF, yet an oscillation can
cause thousands of volts to appear across it
6.) Tank voltage can not overshoot when the tank is mistuned, yet even a
spice analysis or a direct measurement can conform it.
7.) Tubes flat-top (power saturate) when overdriven by a few watts, yet
have enough reserve to provide many kilowatts of parasitic energy
8.) Electronic overload devices need removed, yet resistors make good
fuses.
9.) Eimac's constant current characteristic graph is applicable, by using
one point, to calculate a PA's operating parameters.
Those are your claims.
I think this has run its' course. If you'd like to start at the beginning
and discuss fundamentals, I'd be happy to spend more time. But I'm afraid
without going back to basics, we will never resolve anything.
73 Tom
Tom C. Brown, Jr.
> If you look at other threads in this homebrew area, you'll find Ray Mack
> who observes this effect. Here's what S56A posted:
>
Tom, Rich didn't ask "who else has seen the effect", he said
(paraphrased) "it's funny that ***EIMAC*** didn't write about it in
their material".
Please address *his* comment.
> In article <328A56...@mail.teclink.net>, "Tom C. Brown, Jr."
> <mad...@mail.teclink.net> writes:
> >
> >Tom, it was you who said that the grid dissipation was "about 4 watts".
> >Not Rich.
>
> Tom, you have taken something far from context. I have no time for these
> little "word games".
My apology here, Tom. I don't have time for word games either. I
misinterpreted your statement to mean that the tube's rated grid
dissipation was not 4 watts, when what you were actually saying was that
"at this point" (paraphrased, and referring to some particular point on
the graph) the dissipation was not 4 watts. My fault. I misinterpreted
your statement.
>
> I think this has run its' course. If you'd like to start at the beginning
> and discuss fundamentals, I'd be happy to spend more time. But I'm afraid
> without going back to basics, we will never resolve anything.
>
OK, Tom. Let's start with something VERY fundamental:
For the umpteenth time: WHO IS "XXX" AT EIMAC THAT WROTE THE EMAIL YOU
QUOTED???
That's a pretty simple question, but you keep ignoring and evading it.
If we are going to discuss this rationally, then let's dispense with the
cloak and dagger secrecy, OK?
Rich is not hiding anyone's identity. I can't believe that you have no
idea who sent you that email from Eimac. So, please tell us who wrote
it. Why not?
That oughta be pretty simple from what I can see from here......
73, Tom KJ5IE
R. L. Measures
In article <19961114150...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> Hi Rich,
>
> In a message dated 96-11-14 05:52:20 EST, you write:
>
> >It seems a bit odd that Eimac does not discuss this alleged phenomenon in
> >*Care and Feeding of Power Grid Tubes*. Certainly, in tests performed on
> >my own amplifiers, I was unable to reproduce the incredulous peak anode
> >voltages you said would appear during mistuning of the output network.
>
> If you look at other threads in this homebrew area, you'll find Ray Mack
> who observes this effect. Here's what S56A posted:
>
> In a message dated 96-04-14 12:54:03 EDT, Mario wrote:
> >Subj: Perpetual motion
> >Date: 96-04-14 12:54:03 EDT
> >From: s5...@s55tcp.ampr.org (Marijan Miletic)
> >To: cq-co...@tgv.com
> >CC: W8J...@aol.com
> >
> >Tom, W8JI refrained from further discussion of exotic QRO topics but in
> >a private email convinced me to repeat his tests on my good, old SB-220.
> >
> >I made simple 10/1000pF = 1:100 divider from tune cap, used 1N4148 diode
> >for a peak rectifier and achieved almost the same results as he did!
> >I was initially suspecting that some measurement error was involved...
> >
> >>I have facilities to measure PA stages, and an actual SB-220.
> >>With the amplifier in the SSB position on 7 MHz at 1100 watts output I
> >>measured 2.4 kV peak on the tuning cap. When I mistuned the amp by fully
> >>meshing the loading cap and adjusting the tuning cap for maximum voltage
> >>I measured 3.7 kV peak. The supply voltage was ~ 3kV when mistuned.
.......Tom said:
"The RF voltage in a tank can soar to several times the HV supply when the
tank is underloaded."
My dictionary says:
SEVERAL: Being of a number more than two or three but not many.
Thus, several times 3kV is more than 9kV. 3.7kV is what was measured.
What is the point in citing a test that disproves your statement, Tom?
Furthermore, the cases of destructive arcing I am familiar with did not
occur during tuneup.
> >Under above conditions, SB-220 tuning capacitor tested for 3.5kV may arc!
My SB-220's tuning capacitor does not arc during mistuning with 100w drive.
However, with Heath's factory-stock high vhf Q parasitic suppressors, the
tuning capacitor infrequently arcs, but not during mistuning, during normal
operation. With lower vhf Q suppressors, this is not the case. Why do we
need higher vhf Q if we are trying to build a vhf dampening device, Tom?
Tom has argued that virtually all of the vhf current passes through the
suppressor resistor, and that virtually none of the vhf current passes
through the suppressor inductor. Is this the case? In a typical 8877 vhf
parasitic oscillation suppressor, the resistance is around 50 ohms and the
inductance is roughly 45nH. A typical anode resonant frequency for an 8877
amplifier is about 100MHz. The reactance of a 45nH inductance at 100MHz is
30 ohms. If a 50 ohm resistance is connected in parallel with a 30 ohm
inductance, which component will carry the most current, Tom?
Tom: What is the identity of Eimac's Mr. XXX? Why does Mr. XXX list a
bogus e-mail address?
----------------------------
w8j...@aol.com
Hi Rich,
I can see this is like pulling teeth with a tweezer. Let's cover one area
at a time.
In article <measures-ya023180...@news.vcnet.com>,
meas...@mail.vcnet.com (R. L. Measures ) writes:
>.......Tom said:
> "The RF voltage in a tank can soar to several times the HV supply when
the
>tank is underloaded."
>
>My dictionary says:
>SEVERAL: Being of a number more than two or three but not many.
>Thus, several times 3kV is more than 9kV.
Two times 3 KV is six, not nine. Several times is any amount more than two
or three times.
>3.7kV is what was measured.
RE the SB-220, here is what I said:
In article <5549of$2...@newsbf02.news.aol.com>, w8j...@aol.com (W8JI Tom)
writes:
>Eventually, if equilbrium isn't acheived first in dissipative losses,
>something gives.
>The SB-220, when improperly loaded, can produce over 4 kV peak on the
>input end of the tank. The tube peak anode voltage is over 7 kV. Until
>something "gives" or the energy falls back in equilbrium, the voltage
>builds. Unfortunately it is often the tuning capacitor or bandswitch that
>arcs.
>
>This effect can be measured and documented, it is not a matter guessing
<<End of my statement>>.
Mario measured ~3.7 kV, within 500 volts of what I measured for Heath
during tests of SB-220's and SB-221's. Now let's compare that to your
claims.
In a message dated 96-11-07 12:03:17 EST, Measures wrote:
> However, I do know that in a 3-500Z g-g amplifier that uses a 3000V
anode
>supply, the negative-going change in anode voltage during the conduction
>cycle of the 3-500Z is about 2700vp (volts peak), which coincides with
the
>peak in anode current. At this peak, the anode to (grounded)grid voltage
>is about 300v. {3000v minus 300v = 2700v} When the amplifier is
>underloaded, the tube saturates, and the anode to grid voltage (at the
>anode current peak) is about 150v. In other words, the negative-going
>change in anode voltage is about 2850vp. It's a long way from 2850vp to
>the 5500v or more needed to arc the bandswitch.
So you claim about 2850 volts peak, and I said a SB-220 can produce over
4000 under conditions of improper tuning. Mario verified 3700 volts in a
crude single test. The plate tuning cap had a minimum spec voltage of 3.5
kV at Heath, and the bandswitch was speced at the same voltage between the
ten meter tap and the rotor. In a steady state test of one PA, Mario
verified a voltage exceeding Heat's minimum QC spec breakdown rating of
the tuning cap and switch.
My data comes from a series of tests on six PA's, three 220's and three
221's. I measured as high as 4300 volts peak tank voltage when some PA's
were mistuned.
>Tom has argued that virtually all of the vhf current passes through the
>suppressor resistor, and that virtually none of the vhf current passes
>through the suppressor inductor. Is this the case?
No, I argued that the dominant Q determining component at VHF and higher
was the resistor, not the inductor. I've said repeatedly the primary path
for VHF and UHF signals is through the resistor, and the primary path for
HF signals is through the coil ### IN A PROPERLY DESIGNED SUPPRESSOR ####
.
Before we start on that, let's clear up this other stuff. We now seem to
just go in circles, and that is a waste of time. If you'd like to discuss
parasitic suppressor behavior, continue the SB-220 test discussion, or
discuss PA's that can produce up to 10 times the dc supply voltage we can
do any of that. Take your pick.
After that we can move on to your other questions, but let's agree on
things and get them out of the way so we quit running in a closed loop.
73 Tom
R. L. Measures
In article <19961117011...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> In article <measures-ya023180...@news.vcnet.com>,
> meas...@mail.vcnet.com (R. L. Measures ) writes:
>
> >.......Tom said:
> > "The RF voltage in a tank can soar to several times the HV supply
..snip....
several times 3kV is more than 9kV.
>
> Two times 3 KV is six, not nine.
>
3.7kV is what was measured. Neither 6kV nor 9kV was measured. In my
opinion, most people would take 'a couple of times' as around 6kV, and take
'several times' as around 9kV.
> RE the SB-220, here is what I said:
> In article <5549of$2...@newsbf02.news.aol.com>, w8j...@aol.com (W8JI Tom)
> writes:
>
> >Eventually, if equilbrium isn't acheived first in dissipative losses,
> >something gives.
The measurement of 3700v-p was made in equilibrium.
> >The SB-220, when improperly loaded, can produce over 4 kV peak on the
> >input end of the tank. The tube peak anode voltage is over 7 kV. Until
> >something "gives" or the energy falls back in equilbrium, the voltage
> >builds. Unfortunately it is often the tuning capacitor or bandswitch that
> >arcs.
This voltage does not 'build' in my SB-220, or in my other amplifiers
either Interest on a savings account builds.
Why mention the anode voltage when we are talking about the peak voltage
on the tune capacitor?
None of this discussion about deliberate mistuning really matters, Tom.
In the SB-220, tune capacitor arcing occasionally occurs during normal
operation, not during experiments where the tune and load controls are
mistuned for maximum voltage across the tune capacitor.
My SB-220 tune capacitor used to arc occasionally until I lowered the
VHF-Q of the VHF parasitic oscillation suppressors. After lowering the Q
of the VHF parasitic suppressor, the pecuilar tune-capacitor double-peak in
40m tuneup disappeared. Other folks have experienced similar results. In
the Drake L4-B, the tune-capacitor double-peak in 15m tuneup disappears
when lower Q suppressors are installed.
snip,,
>
> >Tom has argued that virtually all of the vhf current passes through the
> >suppressor resistor, and that virtually none of the vhf current passes
> >through the suppressor inductor. Is this the case?
>
> No, I argued that the dominant Q determining component at VHF and higher
> was the resistor, not the inductor. I've said repeatedly the primary path
> for VHF and UHF signals is through the resistor, and the primary path for
> HF signals is through the coil ### IN A PROPERLY DESIGNED SUPPRESSOR ####
When you telephoned me in 1990, you talked about properly designed
parasitic suppressors. However, you offered no information about such
animals. You brought up the subject of proper suppressor design again
here, and yet you still provide no information.
If an 8877's VHF parasitic oscillation suppressor resistor is 50 ohms and
the (paralleled) suppressor inductor is 30 ohms, the lion's share of the
VHF current would seemingly pass through the 30 ohm inductor and the
lesser, remainder of the total current would seem to pass through the
resistor. However, perhaps I am in error.
Tom: Are you suggesting that the primary path of VHF current is through
the paralleled 50 ohm resistor? If you maintain that most of the VHF
current passes through the 50 ohm resistor, Tom, please show us the
mathematical solution that led you to your conclusion.
---------------------------------------
w8j...@aol.com
W8JI initially said re SB-220:
>> >The SB-220, when improperly loaded, can produce over 4 kV peak on the
>> >input end of the tank. The tube peak anode voltage is over 7 kV. Until
>> >something "gives" or the energy falls back in equilbrium, the voltage
>> >builds. Unfortunately it is often the tuning capacitor or bandswitch
that
>> >arcs.
>
>This voltage does not 'build' in my SB-220, or in my other amplifiers
>either
Everything from photons to grid dissipation is different in you
amplifiers, it only makes sense that PA voltage behaves differently there
also.
> Why mention the anode voltage when we are talking about the peak voltage
>on the tune capacitor?
Why not? It gives users a feel for the stress on the tube.
> None of this discussion about deliberate mistuning really matters, Tom.
It does. If tuning is correct at a given output but...
1.) The antenna relay fails to transfer and/or the loading cap arcs.
2.) Excessive RF drive is applied.
3.) Something in the antenna or tuner arcs
tank voltage and anode voltage can soar. This can cause the tube or tank
components to arc over.
>In the SB-220, tune capacitor arcing occasionally occurs during normal
>operation, not during experiments where the tune and load controls are
>mistuned for maximum voltage across the tune capacitor.
I would suggest someone tune the PA properly then, or correct the ALC
overshoot in the exciter.
> My SB-220 tune capacitor used to arc occasionally until I lowered the
>VHF-Q of the VHF parasitic oscillation suppressors. After lowering the Q
>of the VHF parasitic suppressor, the pecuilar tune-capacitor double-peak
in
>40m tuneup disappeared. Other folks have experienced similar results.
In
>the Drake L4-B, the tune-capacitor double-peak in 15m tuneup disappears
>when lower Q suppressors are installed.
Of course, it lower HF tank Q.
>Tom has argued that virtually all of the vhf current passes through the
>> >suppressor resistor, and that virtually none of the vhf current passes
>> >through the suppressor inductor. Is this the case?
>>
>> No, I argued that the dominant Q determining component at VHF and
higher
>> was the resistor, not the inductor. I've said repeatedly the primary
path
>> for VHF and UHF signals is through the resistor, and the primary path
for
>> HF signals is through the coil ### IN A PROPERLY DESIGNED SUPPRESSOR
####
>
>When you telephoned me in 1990, you talked about properly designed
>parasitic suppressors. However, you offered no information about such
>animals. You brought up the subject of proper suppressor design again
>here, and yet you still provide no information.
Rich, read back. I described methods used to design suppressors. Unlike
you, I believe each system is different, and requires different solutions.
What works best in an 8877 does not work at all with a 3CX1200A7. The
design is dependent on the self-neutralizing frequency of the tube and the
layout.
>If an 8877's VHF parasitic oscillation suppressor resistor is 50 ohms
and
>the (paralleled) suppressor inductor is 30 ohms, the lion's share of the
>VHF current would seemingly pass through the 30 ohm inductor and the
>lesser, remainder of the total current would seem to pass through the
>resistor. However, perhaps I am in error.
Actually, if the grid is properly grounded and the anode leads are short,
the 8877 requires no suppression at all in many cases. When it does
require suppression, it is generally at UHF. So that would be a poor case
study of VHF suppression.
> Tom: Are you suggesting that the primary path of VHF current is through
>the paralleled 50 ohm resistor? If you maintain that most of the VHF
>current passes through the 50 ohm resistor, Tom, please show us the
>mathematical solution that led you to your conclusion.
Let's look at a typical suppressor such as the SB-220, AL-80, or SB-1000
suppressor, since a 3-500Z almost always needs a suppressor in the 100-
130 MHz area. Will that be ok?
73 Tom
w8j...@aol.com
In article <19961118001...@ladder01.news.aol.com>,
w8j...@aol.com writes:
>
>Let's look at a typical suppressor such as the SB-220, AL-80, or SB-1000
>suppressor, since a 3-500Z almost always needs a suppressor in the 100-
>130 MHz area. Will that be ok?
>
>73 Tom
I will be off-net for two or three days.
BTW, Tom Brown keeps asking who my sources at Eimac were. I can not give
that information out (as I have said) because they really don't want to be
involved in this.
They did say they will consider making one more final statement as soon as
they have time, if Mr. Measures agrees to accept their position as the
final word on this subject.
Anyone doubting Eimac's position can investigate things on their own. I
wouldn't wish the kind of technical nonsense being argued here on anyone.
My opinion is anyone believing fantasies about 25 watt grids handling 500
watts, VHF energy making tuning capacitors arc, VHF parasitics causing
glass to melt, Photons making tubes on standby arc over, not understanding
grid dissipation, and so on..... is far beyond any help I can offer.
This thread has outlived any usefulness.
73 Tom
R. L. Measures
> In article <measures-ya023180...@news.vcnet.com>,
> meas...@mail.vcnet.com (R. L. Measures ) writes:
>
......snip.....
> > My SB-220 tune capacitor used to arc occasionally until I lowered the
> >VHF-Q of the VHF parasitic oscillation suppressors. After lowering the Q
>
> >of the VHF parasitic suppressor, the pecuilar tune-capacitor double-peak
> in
> >40m tuneup disappeared. Other folks have experienced similar results.
.....snip.......
> Of course, it lower HF tank Q.
On 40m?........I don't think so, Tom. // A double-peak has nothing to do
with the operating Q of the tank. A double-peak is an indicator that an
anomaly is present.
> >>......... I argued that the dominant Q determining component at VHF and
higher
> >> was the resistor, not the inductor. I've said repeatedly the primary path
> >> for VHF and UHF signals is through the resistor, and the primary path
> for
> >> HF signals is through the coil ### IN A PROPERLY DESIGNED SUPPRESSOR
> ####
> >
> >When you telephoned me in 1990, you talked about properly designed
> >parasitic suppressors. However, you offered no information about such
> >animals. You brought up the subject of proper suppressor design again
> >here, and yet you still provide no information.
>
> Rich, read back. I described methods used to design suppressors. Unlike
> you, I believe each system is different, and requires different solutions.
> What works best in an 8877 does not work at all with a 3CX1200A7. The
> design is dependent on the self-neutralizing frequency of the tube and the
> layout.
Still no specifics.on the 'PROPERLY DESIGNED SUPPRESSOR'. All I have seen
and heard since 1990 has the appearance of ersatz-esoterica......
> >If an 8877's VHF parasitic oscillation suppressor resistor is 50 ohms
> >and the (paralleled) suppressor inductor is 30 ohms, the lion's share of the
> >VHF current would seemingly pass through the 30 ohm inductor and the
> >lesser, remainder of the total current would seem to pass through the
> >resistor. However, perhaps I am in error.
>
> Actually, if the grid is properly grounded and the anode leads are short,
> the 8877 requires no suppression at all in many cases. When it does
> require suppression, it is generally at UHF. So that would be a poor case
> study of VHF suppression.
Who says that no suppression is needed for the 8877? The mysterious Mr.
XXX? It is my opinion that this is porcine shampoo.
The gold-sputtered 8877 grid that is shown on page 16 of the Sept. 1990
issue of QST, was of a failed tube that was removed from a
commercially-made 8877 HF amplifier that used a relatively high VHF Q
parasitic suppressor. The failure occured during normal operation with
normal drive. The failure did not occur during tuneup.
Tom: does the Alpha 77 8877 amplifier utilize a VHF parasitic oscillation
suppressor?
Tom: does the Dentron DTR-2000 8877 amplifier utilize a VHF parasitic
oscillation suppressor?
Tom: does Henry Radio utilize a VHF parasitic oscillation suppressor in
their 8877 amplifiers?
Tom: does the Ameritron AL-1500 8877 amplifier utilize a VHF parasitic
oscillation suppressor?
> > Tom: Are you suggesting that the primary path of VHF current is through
> >the paralleled 50 ohm resistor? If you maintain that most of the VHF
> >current passes through the 50 ohm resistor, Tom, please show us the
> >mathematical solution that led you to your conclusion.
Tom did not solve this problem. Apparently, Tom either 1. didn't have the
time, or 2. he does not know how to solve this problem, or 3. he is fearful
that the answer might jeopardize his title as one of the "amplifier
experts". (See QST, Sept. 1994, page 71, col. 3)
The original problem:
====================
Tom has argued that virtually all of the vhf current passes through the
suppressor resistor, and that virtually none of the vhf current passes
through the suppressor inductor. Is this the case? In a typical 8877 vhf
parasitic oscillation suppressor, the resistance is around 50 ohms and the
inductance is roughly 45nH. A typical anode resonant frequency for an 8877
amplifier is about 100MHz. The reactance of a 45nH inductance at 100MHz is
30 ohms. If a 50 ohm resistance is connected in parallel with a 30 ohm
inductance, which component will carry the most current, Tom?
====================
One solution: Apply 1VRMS to the parallel L-R circuit. The current
through the R is 1V/50 ohms = 20mA. The current through the L is 1V/30
ohms = 33.33mA. The total current is 20 mA plus 33.33 mA = 53.33mA. Thus,
33.33/54.33 = 0.625, or 62.5%, of the total VHF current flows through the
inductor, and 20/53.33 = 0.375, or 37.5% of the total VHF current flows
though the resistor.
Tom stated:
> >>I've said repeatedly the primary path
> >> for VHF and UHF signals is through the resistor, and the primary path for
> >> HF signals is through the coil.
If 62.5% of the VHF current flows through the inductor ("coil"), and 37.5%
of the VHF current flows through the resistor, Tom, please explain how the
primary path for the VHF current is through the resistor?
€ Note to Tom Rauch---after we discuss the above analysis, I plan on
presenting a similar analysis for the 3-500Z parasitic suppressor that you
mentioned in your reply.
Tom C. Brown, Jr.
Partial quote from Tom's message:
>They did say they will consider making one more final statement as soon as
>they have time, if Mr. Measures agrees to accept their position as the
>final word on this subject.
"They" again......
Tom KJ5IE
R. L. Measures
In article <19961119035...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> In article <19961118001...@ladder01.news.aol.com>,
> w8j...@aol.com writes:
>
> >Let's look at a typical suppressor such as the SB-220, AL-80, or SB-1000
> >suppressor, since a 3-500Z almost always needs a suppressor in the 100-
> >130 MHz area. Will that be ok?
> >
> >73 Tom
OK, Tom, let's have a look: A typical anode VHF parasitic suppressor for a
3-500Z, operating at 3kV, uses a 70nH inductor. The suppressor resistor is
typically 50 ohms. At 100MHz, the reactance of 70nH is 45 ohms. To find
out how the current divides between the paralleled inductor and resistor,
we can apply 1VRMS and calculate what happens. The current through the
resistor is 1v/50 ohms = 20mA. The current through the inductor is 1v/45
ohms = 22mA. The total current is 20mA plus 22mA = 42mA. Thus, the
resistor carries 20/42 = 0.476, or 47.6% of the total current. The
inductor carries 22/42 = 52.4% of the total current.
Tom stated:
> >>I've said repeatedly the primary path
> >> for VHF and UHF signals is through the resistor, and the primary path for
> >> HF signals is through the coil.
Tom: If 52.4% of the current flows through the supressor inductor, and
47.6% of the current flows through the suppressor resistor, why have you
'said repeatedly' that the primary path for VHF current is through the
suppressor resistor?
If a 3-500Z were operating at 4kV instead of 3kV, the suppressor inductor
would have to be reduced to around 55nH to prevent the suppressor resistor
from being crispy-crittered during 10m operation.** With a 55nH suppressor
inductance, an even greater percentage of VHF current would pass through
the VHF suppressor's inductor. (** see QST, March, 1989, page 25
"Calculating Power Dissipation in Parasitic-Suppressor resistors")
Since a substantive amount of VHF current flows through the suppressor
inductor, the Q (XL/R) of the inductor should play a sizeable role in the
overall VHF Q of a VHF parasitic suppressor. Is it logical that lower VHF
Q would provide better VHF dampening? If this is true, then the effect of
using a low VHF Q suppressor inductor should be readily observed on a
dipmeter. With a high VHF Q (copper or silver) suppressor inductor, the
observed VHF dip is sharp and deep. With a low VHF Q suppressor inductor,
the dip is broader and less deep. This is nothing new. Mr. F. E. Handy
knew it 70 years ago. You can read what he wrote about VHF
parasitic-suppressors in the 1926 and 1927 Editions of the ARRL's *Radio
Amateurs Handbook*, Mr. Rauch.
> I will be off-net for two or three days.
Looking forward to your return in two or three days, Tom.
> BTW, Tom Brown keeps asking who my sources at Eimac were. I can not give
> that information out (as I have said) because they really don't want to be
> involved in this.
> They did say they will consider making one more final statement as soon as
> they have time, if Mr. Measures agrees to accept their position as the
> final word on this subject.
I do not sign blank cheques, Tom.
Will the real Mr. XXX please stand up and identify himself? Tom: Is Mr.
XXX "they"? When I asked you for the names of "it's the entire staff" at
Eimac, you were evasive. When I asked for the name of Mr. XXX, you wre
evasive. Mr. Brown told me that you appeared to be evading the issue when
he queried you about Mr. XXX via e-mail
The bogus Eimac e-mail address is suggestive of a deception
............Something isn't quite right here, methinks.
> Anyone doubting Eimac's position can investigate things on their own. I
> wouldn't wish the kind of technical nonsense being argued here on anyone.
Nonsense--like assuming that the path for VHF current path is through the
suppressor resistor instead of through both the resistor AND the inductor?
IMO, any "amplifier expert" worth his salt would have calculated the VHF
L/R current distribution before making such a statement.
> My opinion is anyone believing fantasies about 25 watt grids handling 500
> watts,
I did not state that a 25 watt grid handles 500 watts. I stated that 500
to 700 watts would probably be required to raise the 8877's grid structure
to temperatures that boil gold.
VHF energy making tuning capacitors arc,
Certainly not in an AL-80's tuning capacitor, whose VHF resonances are far
removed from the anode's VHF-resonance point, but tune-capacitor arcing
occasionally happens in TL-922's and SB-220s when the factory-stock high
VHF Q suppressors are used. .
VHF parasitics causing
> glass to melt,
During a telephone conversation on 3/28/91 with Jim Riach, Vectronics'
design engineer, Raich told me that Chinese-made 3-500Z glass-envelopes had
melted for no apparent reason. Mr. Raich works directly for Paul Rivnak,
the owner of Vectronics, Ltd. At the time they experienced the
glass-envelope melting problem, Vectronics was building single 3-500Z LA30
amplifiers for AEA. If you would like to contact Vectronics, Tom, the
address is: 31 Estate Drive, Scarborough, Ont., CANADA M1H 2Z2. We sent
them Low VHF Q Suppressor Retrofit Kit s/n 2173. Subsequently, AEA ordered
Low VHF Q Suppressor Retrofit Kits and bulk materials from us so that they
could retrofit lower-Q suppressors into older LA30 amplifiers. We shipped
the first such order on 10/29/91.
>Photons making tubes on standby arc over,
It is hard to believe, but I've talked to physicists and engineers about
this conundrum, and cosmic radiation photons seems to be the collective
best guess.
Ever wonder why lightning follows a zig-zag path instead of a straight
line? Many folks have wondered about this because logic says that any arc
should take place on the shortest distance between the two points--i.e.,
along a straight-line, right? However, some people* have proposed that
high energy photons from cosmic background radiation speeds up electrons in
atmospheric atoms along their paths of propagation. Accelerated electrons
are more prone to escape from their normal orbits and become free
current-carrying electons---provided that the voltage gradient is
sufficient to liberate them. In other words, ionization takes place along
the path that is most easily ionized, which is not necessarily a straight
path.
Tom: If cosmic radiation can change the path of a lightning bolt, does it
seem impossible that cosmic radiation could have some effect on an
amplifier-tube?
---------
* ...Los Alamos National Laboratory physicist Robert Roussel-Dupre, Alex
Gurevich of Moscow's Lebdedev Institute of Physics, Genady Milikh of the
University of Maryland, and Tom Tunnell of EG&G Energy Measurements.
---------
>not understanding
> grid dissipation,
Sure, calculating grid dissipation is not a piece of cake. It is an
integral-calculus problem. However, when I followed your recommendation
for calculating grid dissipation, you argued that the results were not
correct for reasons that did not appear reasonable.
>and so on..... is far beyond any help I can offer.
.......or apparently beyond any help that Eimac's anonymous Mr. XXX will
offer.
> This thread has outlived any usefulness.
>
> 73 Tom
---------------------------------
E-mail copy to Mr. C. Tom Rauch, W8JI
--
--Rich-- ag6k, 805.386.3734
Wes Stewart
In article <measures-ya023180...@news.vcnet.com> meas...@mail.vcnet.com (R. L. Measures ) writes:
>From: meas...@mail.vcnet.com (R. L. Measures )
>Subject: Re: Good source for amplifier tips
>Date: Wed, 20 Nov 1996 10:03:27 -0700
>In article <19961119035...@ladder01.news.aol.com>, w8j...@aol.com
>wrote:
>> In article <19961118001...@ladder01.news.aol.com>,
>> w8j...@aol.com writes:
>>
>> >Let's look at a typical suppressor such as the SB-220, AL-80, or SB-1000
>> >suppressor, since a 3-500Z almost always needs a suppressor in the 100-
>> >130 MHz area. Will that be ok?
>> >
>> >73 Tom
>OK, Tom, let's have a look: A typical anode VHF parasitic suppressor for a
>3-500Z, operating at 3kV, uses a 70nH inductor. The suppressor resistor is
>typically 50 ohms. At 100MHz, the reactance of 70nH is 45 ohms. To find
>out how the current divides between the paralleled inductor and resistor,
>we can apply 1VRMS and calculate what happens. The current through the
>resistor is 1v/50 ohms = 20mA. The current through the inductor is 1v/45
>ohms = 22mA. The total current is 20mA plus 22mA = 42mA. Thus, the
>resistor carries 20/42 = 0.476, or 47.6% of the total current. The
>inductor carries 22/42 = 52.4% of the total current. [snip]
Don't you think it wise to consider the phase angle of these currents. This
ain't DC you know. Furthermore, these aren't isolated components tested in a
measurement system, they're embedded in a system that includes other
reactances and frequency dependent source and load impedances. So, their
functional behavior is quite a bit more complicated that this "Ohm's law"
analysis.
>> This thread has outlived any usefulness.
>>
>> 73 Tom
So it would seem.
>--Rich-- ag6k, 805.386.3734
Wes, N7WS
R. L. Measures
In article <329326...@TECLink.net>, tbr...@TECLink.net wrote:
> Partial quote from Tom's message:
>
> >They did say they will consider making one more final statement as soon as
> >they have time, if Mr. Measures agrees to accept their position as the
> >final word on this subject.
>
>
> "They" again......
>
>
> Tom KJ5IE
..................................................
Who is "they"?
Mr. Rauch can't say .
Who is Mr. XXX with the phony e-mail address?
We can only guess.
On and on.
In an e-mail to me 4/5/96, Mr. Rauch said:
"If you are so sure of your theories, let's debate on Internet. We can go over
your theories point by point in a non-personal debate."
for instance:
I asked Mr. Rauch: what is the frequency of the anode-circuit's resonance
in the AL-80?
I asked Mr. Rauch: does the AL-1500 8877 amplifier (which Mr. Rauch
designed) use a vhf parasitic suppressor?
Mr. Rauch appears not to have answered these seemingly-simple questions.
How can we go over theories point by point if one party is pointedly less
than forthcoming?
---------------------------------------------
e-mail copies to Mr. Brown and Mr. Rauch
--
--Rich-- ag6k, 805.386.3734
w8j...@aol.com
In article <measures-ya023180...@news.vcnet.com>,
>How can we go over theories point by point if one party is pointedly less
>than forthcoming?
You seem to feel it OK for you to not respond to questions, yet after I
plainly posted I had to go off-line for two or three days you fire away
with nonsense about me not replying. So far you've claimed:
1.) The grid dissipation of an 8877 is several hundred watts, never mind
Eimac says 25 watts.
2.) The heating is by grid conductor resistance and not electron
bombardment, never mind the grid's surface resistance would have to be
hundreds of ohms or the grid current hundreds of amperes for this effect
to occur.
3.) Photons make gas-LESS tubes on standby blow up. Never mind a special
gas filled geiger muller tube only has fraction of a pico-ampere.
4.) Glass in Chinese 3-500Z's melts from parasitic current flow in what
you call poor dissipation factor glass. Never mind the thermal lag of the
glass or the electric field required.
5.) A capacitor can arc over from a parasitic but NOT HF energy. Never
mind the reactance of the capacitor is very low at VHF but high at HF.
6.) A band switch can arc over from VHF parasitics. Never mind the low
shunt reactance plate capacitor.
7.) A single point on a constant current characteristic curves can be used
to calculate dissipation. Never mind the operating conditions indicated
are off the charts.
8.) A relative harmonic level graph for a class C 4-65A can be applied to
an 8875 as a direct percentage of harmonic even though it isn't even in
percent.
9.) Adding a slightly resistive nichrome wire (how much Rich? Less than an
few ohms?) in the primary path of HF current affects VHF Q greatly, when
the VHF Q was already about unity at 100 MHz in the out of circuit
"conventional suppressor" you describe.
Yet after all that very obvious technical rubbish you put out, you claim I
am the one putting out the snow job, distorting facts, avoiding questions,
and thinking customers are "drooling idiots".
I am supposed to answer you and spend hours correcting nonsensical claim
after nonsensical claim, while you are free to avoid any question or spit
out any response you choose. You, and in particular the magazine that gave
you a platform to publish such obvious technical garbage, should be
ashamed of what was done to people trying to learn. I think undoing the
damage is impossible.
More time arguing with you is simply wasted time.
73, Tom
R. L. Measures
In article <n7ws.203...@azstarnet.com>, n7...@azstarnet.com (Wes
Stewart) wrote:
> In article <measures-ya023180...@news.vcnet.com>
meas...@mail.vcnet.com (R. L. Measures ) writes:
> >From: meas...@mail.vcnet.com (R. L. Measures )
> >Subject: Re: Good source for amplifier tips
> >Date: Wed, 20 Nov 1996 10:03:27 -0700
>
> >In article <19961119035...@ladder01.news.aol.com>, w8j...@aol.com
> >wrote:
>
> >> In article <19961118001...@ladder01.news.aol.com>,
> >> w8j...@aol.com writes:
> >>
>
> >> >Let's look at a typical suppressor such as the SB-220, AL-80, or SB-1000
> >> >suppressor, since a 3-500Z almost always needs a suppressor in the 100-
> >> >130 MHz area. Will that be ok?
> >> >
> >> >73 Tom
>
> >OK, Tom, let's have a look:
...snip...
> >Thus, the resistor carries 20/42 = 0.476, or 47.6% of the total current. The
> >inductor carries 22/42 = 52.4% of the total current. [snip]
>
> Don't you think it wise to consider the phase angle of these currents. This
> ain't DC you know. Furthermore, these aren't isolated components tested in a
> measurement system, they're embedded in a system that includes other
> reactances and frequency dependent source and load impedances. So, their
> functional behavior is quite a bit more complicated that this "Ohm's law"
> analysis.
True enough, 'it ain't DC', Wes, and the currents are not in phase, the
exercise in question was to calculate the ratio of the current through the
parallel R and L components. In doing this, I obviously ignored the
intrinsic L in the resistor and I ignored the intrinsic R in the inductor.
However, the answer is probably in the ballpark My point is that roughly
half of the VHF current flows through the R and roughly half of the VHF
current flows through the L. I believe it is hubristic to presume that the
VHF current through either the suppressor R or L can be ignored with
impunity, which has been Mr. Rauch's position for at least the past 6
years.
At DC, nichrome alloys are about 57 times more resistive than copper. To
the best of my knowledge, there is no type of wire that has more VHF
resistance, and produces a lower VHF-Q on a Boonton Q-Meter, than
nickle-chromium-iron alloys.
The reason that high-resistance inductors produce a lower-Q VHF parasitic
suppressor is that roughly half of the VHF current passes through a VHF
suppressor's inductor.
All one needs to demonstrate this effect is a dipmeter and an HF
amplifier. Low VHF-Q suppressors produce a shallower, broader VHF dip than
the relatively higher VHF-Q suppressors that most amplifier manufacturers,
including Ameritron, still use. A broader VHF dip means better
VHF-dampening---which is what a VHF-suppressor is supposed to do. Mr. F.
E. Handy realized this in 1926. I realized this in 1986.
. Even if we knew that the phase angle of the currents in L and R were,
say for instance, 50 degrees apart at 100MHz, how would you use this
information in designing a VHF parasitic-suppressor, Wes?
> >> This thread has outlived any usefulness.
> >>
> >> 73 Tom
>
> So it would seem.
>
Perhaps not. Mr Rauch indicated that he "will be off-net for two or three
days."---or this debate may be near its end. However, in either case, an
epilogue might be appropriate---provided that anyone is interested in the
story behind the smoke and mirrors.
€€ EPILOGUE
Paul Pagel, N1FB, QST's Technical Correspondence editor related the
following story to me:
Paul said that he comes to work at QST about a half hour early to unlock
225 Main Street. Shortly after "Parasitics Revisited" was published in the
Sept. and Oct. 1990 issues of QST, as Paul opened up one morning, he could
hear a telephone ringing in the back of the darkened building. He switched
on the lights and started walking towards his desk in the back of the
building. As Paul got closer and closer to his desk, the ringing became
louder and louder. When he reached his desk, it was his telephone ringing.
Paul said that this happened repeatedly over a period of weeks. The caller
was always Mr. Tom Rauch, W8JI, from Ameritron, complaining about
"Parasitics Revisited". . Paul said that Tom seemed to be especially
concerned about the photograph of the burned bandswitch in "Parasitics
Revisited" (page 33 in the October 1990 issue of QST). Tom was upset
because owners of Ameritron amplifiers, whose bandswitches had burned up,
saw the photo of the burned bandswitch on page 33 in QST and were
apparently calling Ameritron to complain. Paul said that Tom blamed the
burned bandswitches in Ameritron amplifiers on bad antennas, bad coax and
bad operators. After many such telephone calls, Paul began to realize
that he could not follow Mr. Rauch's technical logic, so Paul recommended
that he telephone the author of the article, namely me, to resolve
technical issues. .
Tom Rauch telephoned me. After identifying himself, he somewhat brusquely
demanded to know if I was tape recording the telephone call. I answered no.
Mr. Rauch was skeptical. After additional interrogation, he eventually
became convinced that I was NOT recording the conversation, and we moved on
to the matter at hand. Eventually, it became apparent that Mr. Rauch's
concept of AC Circuit Analysis and my concept of AC Circuit Analysis are
different. Mention VHF, and Mr. Rauch sees "coils" (inductors) as virtual
open circuits. In other words, in Mr. Rauch's mind, VHF plus "coil" equals
insignificant current flow, PERIOD. Next question. ..............
OTOH, I believe that inductance and frequency should be entered into the
appropriate AC Circuit Analysis formulae, and the results calculated.
I was amazed that QST Editor Mark Wilson never questioned Mr. Rauch's
simplistic views on AC Circuit Analysis. See page 74 in the September, 1994
issue. Quoting Mr. Rauch:
------------------
"The coil's reactance increases with frequency, and at VHF it is plainly
evident that the dominant component at VHF is the resistor, not the coil.
Changing the coil has very little effect on overall Q."
------------------
Mr. Rauch is so sure that he is always right, he presented me with a 3-500Z
L/R VHF suppressor current-division problem that he was certain would prove
him right, and presumably bring me to my knees, technically-speaking. It
backfired on Mr. Rauch.
I would think that one of our "recognized amplifier experts" would have
taken the minute or so necessary to calculate the current distribution. If
Tom Rauch had spent spent such a minute six years ago, he would have saved
himself, QST, and yours truly much unpleasantness.
================================================
e-mail copies to Messrs. Stewart, Brown, and Rauch
--
--Rich-- ag6k, 805.386.3734
Filip M Gieszczykiewicz
In Article <19961122120...@ladder01.news.aol.com>, through puissant locution, w8j...@aol.com soliloquized:
>In article <measures-ya023180...@news.vcnet.com>,
>>How can we go over theories point by point if one party is pointedly less
>>than forthcoming?
>I am supposed to answer you and spend hours correcting nonsensical claim
>after nonsensical claim, while you are free to avoid any question or spit
>out any response you choose. You, and in particular the magazine that gave
>you a platform to publish such obvious technical garbage, should be
>ashamed of what was done to people trying to learn. I think undoing the
>damage is impossible. More time arguing with you is simply wasted time.
Greetings. I would just like to interject here that as someone who
grew up after tubes "died" but is still MOST interested in these
discussions... I'm now totally confused and lost. On the bright
side, several explanations have been posted here that should have
been in the books I read... clear and simple enough to digest and
understand without a bottle of aspirin... While I REALLY want this
information to come out... I now have to ask myself, "Why pay
attention if it could be all wrong?"... Or, rather, who is right?
Heck, I'm so confused now that I don't even know which part you're
arguing about: the test results or the physics involved... or both.
Take care.
P.S. Feel free to ignore me... today is my "designated bad day" :-)
--
+-->Filip "I'll buy a vowel" Gieszczykiewicz | E-mail: fil...@paranoia.com
| http://www.paranoia.com/~filipg/ |SCI.ELECTRONICS.REPAIR FAQ + LOTS MORE!
| Enjoy your job, work within the law, make lots of money : Choose any two.
| I think for myself. I listen. I make decisions. I speak what I believe.
Tom C. Brown, Jr.
Wes Stewart wrote:
[Partial]
>
> Don't you think it wise to consider the phase angle of these currents. This
>ain't DC you know. Furthermore, these aren't isolated components tested in a
>measurement system, they're embedded in a system that includes other
>reactances and frequency dependent source and load impedances. So, their
>functional behavior is quite a bit more complicated that this "Ohm's law"
>analysis.
Hi Wes-
Are you saying that Tom was making a hasty (and incorrect) assumption
when he said that he majority of any VHF current would go through the
resistor rather than the suppressor coil? If that's what you are
saying, then I agree. If not, then please explain how the difference in
phase angle between the reactive path and the resistive path (the
parallel resistor and inductor in the suppressor) would force the VHF
current to flow through the higher impedance path rather than the lower.
I think I must be missing something....... (of course, it wouldn't be
the first time THAT happened! <g>)
73, Tom KJ5IE
R. L. Measures
In article <19961122120...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> In article <measures-ya023180...@news.vcnet.com>,
> meas...@mail.vcnet.com (R. L. Measures ) writes:
> >Mr. Rauch appears not to have answered these seemingly-simple questions.
> >How can we go over theories point by point if one party is pointedly less
> >than forthcoming?
>
> You seem to feel it OK for you to not respond to questions, yet after I
> plainly posted I had to go off-line for two or three days you fire away
> with nonsense about me not replying. So far you've claimed:
>
> 1.) The grid dissipation of an 8877 is several hundred watts, never mind
> Eimac says 25 watts.
I estimated that it would probably take on the order of 500 to 700 watts to
elevate the 8877 grid to a temperature that would boil gold, which is
clearly what happened when one views a gold-sputtered 8877 grid through a
microscope.
> 2.) The heating is by grid conductor resistance and not electron
> bombardment, never mind the grid's surface resistance would have to be
> hundreds of ohms or the grid current hundreds of amperes for this effect
> to occur.
You seem to be ignoring the fact that Eimac determined than an oscillation
condition was the cause of gold-sputtering from the 8877's gold-plated
grid.
> 3.) Photons make gas-LESS tubes on standby blow up. Never mind a special
> gas filled geiger muller tube only has fraction of a pico-ampere.
Tom: Do you believe that there are no gas atoms inside a vacuum tube?
Your opinions on photons seem to be somewhat at odds with the findings of
physicist Robert Roussel-Dupre, Alex Gurevich of Moscow's Lebdedev
Institute of Physics, Genady Milikh of the University of Maryland, and Tom
Tunnell of EG&G Energy Measurements
Tom: What do you suppose causes high speed x-ray film to slowly fog up
when it is parked in a lead-lined container?
> 4.) Glass in Chinese 3-500Z's melts from parasitic current flow in what
> you call poor dissipation factor glass. Never mind the thermal lag of the
> glass or the electric field required.
According to Jim Riach and Paul Rivnak (Vectronics, Ltd.), the glass
envelopes of some Chinese-made (c.1990) 3-500Zs reportedly melted in small
areas a few cm away from the anode-stem, for no apparent reason. Mr. Raich
reported this initially to QST-staffer Rus Healy, NJ2L, and Rus recommended
that Mr. Raich contact me. {Rus Healy edited "Parasitics Revisited" in the
9/90 and 10/90 issues of QST. Rus also furnished the gold-sputter damaged
8874 that is shown in the QST photograph. Rus told me that the
gold-sputtered 8874 was removed from an Alpha amplifier, whose electronic
grid protection had dead-shorted.at the same time that the 8874 failed.)
> 5.) A capacitor can arc over from a parasitic but NOT HF energy. Never
> mind the reactance of the capacitor is very low at VHF but high at HF.
Tom: Do you all tuning capacitors have the same internal VHF resonances?
I do not believe that this is the case. For instance, the tune capacitor
in the AL80 apparently has no VHF resonances that are near the VHF
anode-resonance in that amplifier. Thus, I am not surprised that I have
never heard a report of tune capacitor arcing in the AL80. However, I do
hear occasional reports of bandswitch contact arcing in the AL80, as well
as bandswitch contact arcing in its cousin, the SB-1000---which Tom Rauch
blames on bad antennas, bad coax and bad operators. OTOH, tune capacitor
arcing is common in the SB-220 and the TL-922----but they use a DIFFERENT
tuning capacitor, Tom.
> 6.) A band switch can arc over from VHF parasitics. Never mind the low
> shunt reactance plate capacitor.
I do not understand what you mean by "plate capacitor".
>
> 7.) A single point on a constant current characteristic curves can be used
> to calculate dissipation. Never mind the operating conditions indicated
> are off the charts.
You said that meter-indicated DC grid current should be used to calculate
grid dissipation. I did this. An average DC grid current of 150mA in a
3CX800A7 is hardly equivalent to being "off the charts". The RMS voltage
at the 50 ohm cathode is essentially the same as the RMS output voltage
from the 50 ohm exciter---roughly 27V. 150mA multiplied by roughly 27V is
roughly 4W. Tom: How did you arrive at 20W of grid dissipation? Do you
believe that 20W is capable of elevating the grid to gold-boiling
temperatures?
> 8.) A relative harmonic level graph for a class C 4-65A can be applied to
> an 8875 as a direct percentage of harmonic even though it isn't even in
> percent.
-------
(Note: This item is from a discussion on the Topband reflector group) ....
However, briefly, the "graph" (Figure 21 on page 35 in *Care and Feeding
of PowerGrid Tubes*) does NOT indicate that it applies to a specific Class
of amplifier operation. The point is this: you indicated that a Q of 5.9
was okay. Immediately under Figure 21 Eimac says:
"A resonant circuit with a loaded Q of 10 to 20 is usually considered
optimum. A lower loaded Q will result in greater harmonic output." ...
... ...
Tom: Are you saying that Eimac is wrong, and that Qs substantively lower
than 10 are good engineering practice?
--------
> 9.) Adding a slightly resistive nichrome wire (how much Rich? Less than an
> few ohms?) in the primary path of HF current affects VHF Q greatly, when
> the VHF Q was already about unity at 100 MHz in the out of circuit
> "conventional suppressor" you describe.
"Slightly"?...........Nichrome has at least 50 times more resistance than
copper or silver. Also, nichrome has a slight magnetic property, so it
exhibits additional loss at VHF, especially when iron is added to the
recipe. Why wouldn't an amplifier engineer utilize such a material for a
circuit designed to dampen VHF oscillations? Mr. F. E. Handy wrote about
it the 1926 Edition of the *Radio Amateur's Handbook*.
> Yet after all that very obvious technical rubbish you put out, you claim I
> am the one putting out the snow job, distorting facts, avoiding questions,
> and thinking customers are "drooling idiots".
I happen to be the person that some of such customers called after they
called Ameritron. When one of Ameritron's customers knows that there was
nothing wrong with his coax, antenna, or the way he tuned up an hour before
the arc occured that burned up the bandswitch, what do you expect? //
Kenwood used to tell owners of TL-922s that their bandswitches arced and
burned up due to 'switching the bandswitch while transmitting at full
power'. Yeah, right.
> I am supposed to answer you and spend hours correcting nonsensical claim
> after nonsensical claim, while you are free to avoid any question or spit
> out any response you choose. You, and in particular the magazine that gave
> you a platform to publish such obvious technical garbage, should be
> ashamed of what was done to people trying to learn. I think undoing the
> damage is impossible.
>
This debate was Tom Rauch's idea in 1990. Via an e-mail to me earlier this
year, you said you still wanted a public debate with me. When I was
recently made aware of some the advice you were proffering on the Internet,
I decided to take you up on your offer.
RE: 'the magazine': A critique of my QST articles was published in the
September 1994 issue of QST. In the October 1994 issue of QST,
MFJ-Ameritron had over 30 full-pages of advertising. I wrote a rebuttal to
the critique. I sent copies of my rebuttal you, as well as to the other
five critics,. You promptly wrote me and QST a letter in which you
threatened to sue for libel if my rebuttal was published in QST. //
Anyone who would like to see the rebuttal can find it at:
http://www.vcnet.com/measures/
Tom:
€ Who is Mr. XXX? Why does he list an invalid Eimac e-mail address?.
€ Who are/is "they"?.
€ What are names of the people at Eimac, to whom you refer to as "it's the
entire staff", who allegedly say that Mr. Foote and I are wrong about
gold-sputtering being caused by an oscillation condition?
€ Does the AL-1500 8877 amplifier, which you designed, use a VHF parasitic
oscillation suppressor?
€ What is the frequency of the anode-resonance in the Ameritron
AL-80...amplifier ?
€ You proposed calculating the L/R VHF current-distribution for a typical
3-500Z suppressor. I calculated the VHF current distribution, and posted
the results here. Have you lost interest in this matter?
-----------------------------------
e-mail copy to Mr. C. Tom Rauch, Jr.
--
--Rich-- ag6k, 805.386.3734
w8j...@aol.com
In article <measures-ya023180...@news.vcnet.com>,
meas...@mail.vcnet.com (R. L. Measures ) writes:
>Perhaps not. Mr Rauch indicated that he "will be off-net for two or
three
>days."---or this debate may be near its end. However, in either case,
an
>epilogue might be appropriate---provided that anyone is interested in the
<SNIP>
> I would think that one of our "recognized amplifier experts" would have
>taken the minute or so necessary to calculate the current distribution.
If
>Tom Rauch had spent spent such a minute six years ago, he would have
saved
>himself, QST, and yours truly much unpleasantness.
Hi Rich,
Doesn't that current distribution indicate the Q is unity?? If I have a Q
of unity, how much lower can I make it? What is the RF resistance of the
thing nichrome wires, and what is the goal IN THE CIRCUIT?
Dick Ehrhorn owes me one. Last year Dick was the technical idiot
conspiring against you. This year it is me.
From Warren Bruene of Collins radio fame to Buzz Miklos of Siemens, from
Chazen and Brandon of Eimac to Orr, everyone disagrees with Rich Measures
just because they WANT to sell people bad ideas.
Harris, Rockwell, Siemens, GE, ETO, Ten Tec, Heathkit, Drake, Sanders,
RCA, Ameritron, and all the other manufacturers aren't as knowlegeable as
Rich. Who says this??? Rich.
Companies like those, and dozens of people who contacted QST to point out
the errors in your articles, must just avoid your suggestions because they
just don't understand theory or physics as well as Rich.
Discussing circuit theory or physics is a waste of time, because even
textbooks are part of the anti-Rich conspiracy.
73, Tom
R. L. Measures
> In article <measures-ya023180...@news.vcnet.com>,
> meas...@mail.vcnet.com (R. L. Measures ) writes:
Tom: Do all tuning capacitors have the same internal VHF resonances? I
don't think so. The tune capacitor in the AL80 apparently has no VHF
(Note: This item is from a discussion on the Topband-reflector group) ....
R. L. Measures
In article <19961123144...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> In article <measures-ya023180...@news.vcnet.com>,
> meas...@mail.vcnet.com (R. L. Measures ) writes:
>
> >Perhaps not. Mr Rauch indicated that he "will be off-net for two or
> three
> >days."---or this debate may be near its end. However, in either case,
> an
> >epilogue might be appropriate---provided that anyone is interested in the
>
> <SNIP>
>
> > I would think that one of our "recognized amplifier experts" would have
> >taken the minute or so necessary to calculate the current distribution.
> If
> >Tom Rauch had spent spent such a minute six years ago, he would have
> saved
> >himself, QST, and yours truly much unpleasantness.
>
> Hi Rich,
>
> Doesn't that current distribution indicate the Q is unity??
I don't think so. Please explain, Mr. Rauch.
snip...
>
> From Warren Bruene of Collins radio fame to Buzz Miklos of Siemens, from
> Chazen and Brandon of Eimac to Orr, everyone disagrees with Rich Measures
> just because they WANT to sell people bad ideas.
QST sent me a photocopy of Bill Orr's letter. Mr. Orr claimed that there
were 'numerous technical errors' in "Parasitics Revisited". Mr. Orr's
allegation was discussed at a QST staff meeting. However, since Mr. Orr
failed to specify even one error, the staffers at the meeting collectively
agreed that Orr's letter was a "zero". (David Newkirk, WJ1Z-now W9VES)
> Harris, Rockwell, Siemens, GE, ETO, Ten Tec, Heathkit, Drake, Sanders,
> RCA, Ameritron, and all the other manufacturers aren't as knowlegeable as
> Rich. Who says this??? Rich.
No
Hmm........."and all the other manufacturers" too? Wow! I was not aware
of this, Tom. Please provide us with a list.
Vern Baumgartner, W6HW, who worked as an engineer for Art Collins before
Art sold out to Rockwell, told me that Collins used a nickle-chromium
alloy inductor in the Collins Radio 30S-1 amplifier's VHF
parasitic-suppressor.
A Hughes engineer told me that they use low VHF-Q parasitic-oscillation
suppressors in a 100kW HF amplifier design.
I guess it's sorta like air-bags, Tom. Some people say they save lives.
Other people say air-bags kill.
> Companies like those, and dozens of people who contacted QST to point out
> the errors in your articles, must just avoid your suggestions because they
> just don't understand theory or physics as well as Rich.
Big organizations don't always do the right thing, Tom. NASA launched
Shuttle Mission 25 even though NASA had videotapes of 10 previous missions
where fire could be seen coming through the o-rings on the booster rockets,
adjacent to the main hydrogen tank, no less. In fact, NASA had previously
fired the two safety engineers who were trying to get the leaky o-rings
fixed before Mission 25 sent Christa McAuliffe and friends into the
Atlantic Ocean at 200 knots plus.
Ford chose to save a few dollars per Pinto automobile on a gas-tank crash
protector, even though Ford KNEW that Pintos turned into Molotov Cocktails
in rear-end crash tests. More than two hundred people were cremated alive
in Ford Pintos.
One of the alleged errors that was criticized in my article "The Nearly
Perfect Amplifier" (QST, 9/94) was my assertion that R-C "equalizer"
networks across silicon rectifiers could cause premature rectifier failure.
1200 words of space in the 9/94 QST was devoted to criticism of this.
However, in the 1995 Edition of the Radio Amateur's Handbook, the author of
the power supplies chapter says that R-C equalizer networks can cause
premature rectifier failure---and he explains why.
------------------------------------
e-mail copy to Mr. Rauch
--
--Rich-- ag6k, 805.386.3734
w8j...@aol.com
In article <measures-ya023180...@news.vcnet.com>,
meas...@mail.vcnet.com (R. L. Measures ) writes:
>True enough, 'it ain't DC', Wes, and the currents are not in phase, the
>exercise in question was to calculate the ratio of the current through
the
>parallel R and L components. In doing this, I obviously ignored the
>intrinsic L in the resistor and I ignored the intrinsic R in the
inductor.
Rich, you seem to consider circuit design a collection of generic values.
You arbitrarily assigned a value of 50 ohms to the resistor, and .070 uH
to the inductor. The AL-80 series uses a 100 ohm resistor, not 50. The
inductance of later units was indeed reduced to ~85 nH, mainly due to
reduced grid lead and anode lengths, but the earlier models (the "A"
series) used about 120 nH. Early AL-80 series units had longer grid and
anode paths, requiring larger values of inductive reactance. The SB-220,
with its' long grid leads, required more inductance also in combination
with less resistance. The typical inductance required in these PA's varied
somewhat, but was typically in
the 100-200 nH range.
Your calculations are exact for the values "pulled from your hat", but the
actual shift in current varies with the actual supperssor.
A second factor enters the equation, and this is the critical area. Adding
nichrome to the primary HF path does almost NOTHING to system Q. I have no
idea what guage wire you use, but let's look at some typical suppressors
when 5 ohms of ESR is added to the inductor Here are the numbers:
With your suggested values of 70 nH and 50 ohms at 100 MHz.
Q = 1.14 (par equiv values (p eq) are 50r and 70 nH
with 5 ohm esr from nichrome.
Q = 1.0 (p eq) = 44.34r and 70.9 nH
With 70 nH and 100 ohms:
Q = 2.3
with 5 ohm esr nichrome
Q = 1.8 (p eq) = 79.7 r and 70.9 nH
With 150 nH and 50 ohms:
Q = 0.53
with 5 ohms esr nichrome
Q = .5 (p eq) 48.6r and 150.4 nH
(This is the typical 220 suppressor, it has almost a 2:1 current ration in
the suppressor, with resistor dominant)
With 150 nH and 100 ohms:
Q = 1.06
with 5 ohms esr nichrome
Q = 1.02 (p eq) 94.7r and 150.4 nH
In all cases, the slightest reduction in load resistance across the coil
woul do much more than winding the coil out of a 5 ohms loss material. Of
course this considers the non-nichrome coil to be lossless, so the actual
reduction in Q would be less.
From this very small reduction in 100 MHz Q, Rich claims all ills of the
PA are cured. No melted glass, no photons, no oscillations, no arcs.
Rich, please explain why you go through the bother of adding a nichrome
coil when the Q reduction is so very slight? In almost ALL the cases
above, removal of 5-10 ohms of parallel resistance would have changed Q
the same amount or more.
>current flows through the L. I believe it is hubristic to presume that
the
>VHF current through either the suppressor R or L can be ignored with
>impunity, which has been Mr. Rauch's position for at least the past 6
>years.
I believe it is foolish to change a whole suppressor, when a parallel
resistance of several hundred ohms across the coil would accomplish the
same Q reduction.
> At DC, nichrome alloys are about 57 times more resistive than copper. To
>the best of my knowledge, there is no type of wire that has more VHF
>resistance, and produces a lower VHF-Q on a Boonton Q-Meter, than
>nickle-chromium-iron alloys.
If you measured the coil WITHOUT the loading resistor, as I recall you
saying, your measurement method was flawed. Add the loading resistor, and
measure the CIRCUIT. You will find almost no measureable change in Q until
you lower the Boonton's frequency to HF. If you send me a suppressor, I'll
measure it on my HP or Wiltron, they go up over 1 GHz.
I think though we can safely conclude making the coil out of nichrome
mainly lowers HF Q, NOT VHF Q.
> I would think that one of our "recognized amplifier experts" would have
>taken the minute or so necessary to calculate the current distribution.
If
>Tom Rauch had spent spent such a minute six years ago, he would have
saved
>himself, QST, and yours truly much unpleasantness.
I would have thought Mr. Measures would have known how to use a Q meter
properly, and how to calculate Q in a circuit. That would have saved us
all many hours of time listening to wild pie-in-the-sky theories about
photons, melted glass, boiling gold in 500-700 watt 8877 grids (never mind
Eimac says 25 watt) and parasitics.
You've paid too much attention to discrete values Rich, and not to the
system. Nice try, but once again you argument falls apart.
BTW if any of my numbers are wrong, let me know. None of us are perfect,
are we?
73 Tom
R. L. Measures
In article <19961124033...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> In article <measures-ya023180...@news.vcnet.com>,
> meas...@mail.vcnet.com (R. L. Measures ) writes:
>
> >True enough, 'it ain't DC', Wes, and the currents are not in phase, the
> >exercise in question was to calculate the ratio of the current through
> the
> >parallel R and L components. In doing this, I obviously ignored the
> >intrinsic L in the resistor and I ignored the intrinsic R in the
> inductor.
>
> Rich, you seem to consider circuit design a collection of generic values.
> You arbitrarily assigned a value of 50 ohms to the resistor, and .070 uH
> to the inductor. The AL-80 series uses a 100 ohm resistor, not 50. The
> inductance of later units was indeed reduced to ~85 nH, mainly due to
> reduced grid lead and anode lengths,
...snip...
I could do the calculations for the AL-80's 100 ohm suppressor R and 85nH
suppressor L---if you provide the anode resonant frequency for the AL-80,
Tom..
-------------------------------
e-mail copy to Mr. Rauch
--
--Rich-- ag6k, 805.386.3734
comm...@aol.com
>
> I contend that since roughly half of the VHF current flows through the
L,
>a resistive conductor offers a VHF-dampening advantage. Dipmeter-tests
>with such a VHF-suppressor installed in various amplifiers demonstrates
>that with a nichrome L, the VHF dip is broad and shallow.
>
>
Whatever happened to the technique where the
parasitic suppressor is placed in series with
the cathode of a GG amp? I use one in my SB-220,
and it never spits, arcs, or misbehaves.
Pete
R. L. Measures
In article <19961124180...@ladder01.news.aol.com>,
comm...@aol.com wrote:
--------------------------
I include a cathode-resistor (Rc) in my low VHF-Q parasitic suppressor
retrofit-kit for the SB-220. The reason that Rc works in the SB-220 is
that that there are two VHF-resonant circuits---one in the cathode (input)
circuit and one in the anode (output) circuit. The frequency of both
resonances is roughly 110MHz---obviously not a good situation unless one is
trying to build a 110MHz oscillator.
With average-gain 3-500Zs, putting a VHF-dampening device in either the
input or output circuit may produce stability. However with above-average
gain 3-500Zs, using a VHF-dampening device in both the input and in the
output is indicated.
BTW, the SB-220's input-resonance at 110MHz is created by the length of
the RG-58u coaxial-cable that Heath specified between the tuned-input
bandswitch and the cathodes of the 3-500Zs.
-----------------------------------------------
e-mail copy to 'Pete' and to Mr. C. Tom Rauch
--
--Rich-- ag6k, 805.386.3734
R. L. Measures
In article <575aig$8...@villa.fc.net>, fil...@paranoia.com (Filip M
Gieszczykiewicz) wrote:
> In Article <19961122120...@ladder01.news.aol.com>, through
puissant locution, w8j...@aol.com soliloquized:
> >In article <measures-ya023180...@news.vcnet.com>,
> >meas...@mail.vcnet.com (R. L. Measures ) writes:
> >>Mr. Rauch appears not to have answered these seemingly-simple questions.
> >>How can we go over theories point by point if one party is pointedly less
> >>than forthcoming?
> [zap]
> >I am supposed to answer you and spend hours correcting nonsensical claim
> >after nonsensical claim, while you are free to avoid any question or spit
> >out any response you choose. You, and in particular the magazine that gave
> >you a platform to publish such obvious technical garbage, should be
> >ashamed of what was done to people trying to learn. I think undoing the
> >damage is impossible. More time arguing with you is simply wasted time.
>
> Greetings. I would just like to interject here that as someone who
> grew up after tubes "died" but is still MOST interested in these
> discussions... I'm now totally confused and lost.
------------
I understand the confusion, Filip. There has been plenty of smoke and
mirrors in this debate, and I suspect there will be more. Briefly---The
roots of the debate date from 1926, when Mr. F. E. Handy said on page 72 in
the *Radio Amateur's Handbook*:
"The combination of both resistance and inductance is very effective in
limiting parasitic oscillations to a negligible value of current."
In other words, Mr. Handy found that VHF parasitic-oscillation
suppressors performed measurably better when the suppressor's inductor (L)
was made from resistance wire instead of copper wire. Nichrome has the
highest volume-resistivity among resistance wire alloys, but it is not
easily-soldered and is not commonly available..
After a few years, Handy's observation was somehow left out of the
Handbook. After WW2, people apparently forgot what Mr. Handy wrote in 1926
and 1927, and people started making suppressors out of whatever was handy
and easily soldered---i.e., copper. Some folks even started silver-plating
suppressor inductors, evidently for cosmetic reasons. In the 1940's and
1950's, it didn't matter much what the suppressor L was made from because
large amplifier tubes did not exhibit very much VHF gain. With modern,
high-gain amplifier tubes, VHF suppressors with a high VHF-dampening
ability seem to be indicated. High VHF-dampening means low VHF Q.
The difference between VHF suppressor inductors made from resistance wire,
and those made from copper or silver-plated material, can easily be
compared with a dipmeter. The first time I made this comparison was in
1986. With a copper suppressor inductor, the VHF dip was sharp and deep.
With a nichrome suppressor inductor, the VHF dip was broad and shallow.
Subsequent tests with a Boonton Q meter showed a substantive difference in
VHF Q between otherwise-similar inductors made from copper and those made
from nichrome. Additional tests with other alloys demonstrated that if a
small percentage of iron was added to the nichrome recipe, an even lower
VHF Q, and better VHF dampening resulted.
My findings were published in the October 1988 issue of QST magazine. The
reader response was startling. During the next ten months, I received
around 1200 telephone calls and letters from readers. I had to hire a
helper to keep up with the mail.
Other articles I wrote about VHF parasitic oscillation were published in
the March 1989, September 1990, October 1990, and the January, 1994 issues
of QST magazine. The January 1994 article was written at the request of
QST magazine. However, the January, 1994 article reportedly irritated some
amplifier manufacturers, who still use high VHF-Q parasitic suppressors,
who advertise in QST. (see QST, Sept., 1994, page 71)
For at least the past 6 years, Mr. Rauch has stentorianly asserted that
essentially no VHF current flows through the L, virtually all VHF current
flows through the R, and virtually all HF current flows through the L. Mr.
Rauch believes that the suppressor L should be made from copper.
I contend that since roughly half of the VHF current flows through the L,
a resistive conductor offers a VHF-dampening advantage. Dipmeter-tests
with such a VHF-suppressor installed in various amplifiers demonstrates
that with a nichrome L, the VHF dip is broad and shallow.
Now that Mr. Rauch has discovered that his oft-repeated assertion of the
past 6 years is somewhat lacking in technical accuracy, he seems to want to
talk about ANYTHING BUT the VHF current distribution between a suppressor's
L and R. On and on, round and round. Apparently, "recognized amplifier
experts" have a public image to defend at any cost---even though it's
beginning to look like "The combination of both resistance and inductance
is very effective in limiting parasitic oscillations to a negligible value
of current."
-------------------------------------------------------
e-mail copies to Messrs. Gieszczykiewicz and Rauch
--
--Rich-- ag6k, 805.386.3734
w8j...@aol.com
> With average-gain 3-500Zs, putting a VHF-dampening device in either the
>input or output circuit may produce stability. However with
above-average
>gain 3-500Zs, using a VHF-dampening device in both the input and in the
>output is indicated.
This comment puzzles me Rich.
You attribute a wide gain variation to tube differences, indicting it is
an input to output feedback loop.
The worse engineering mistake in the SB-220 was the inclusion of chokes
in paralle with capacitors from the grids to ground in the SB-220. It
makes that PA subject to VLF parasitics, and destroys gain flatness. It
makes the tubes less stable at VHF also. The single most important mod to
an SB-220 is to ground the grids right at the tube pins with very very
short leads.
You attribute problems to the gain of tubes. The gain of a grounded grid
PA, due to fact the input circuit is in series with the output RF path, is
determined by the ratio of cathode driving to anode output impedance.
In thousands of tubes of all types, I see very little gain difference.
There are even reject limits, for example 3CX1200A7's with as little as
.5 dB gain difference are rejected. That amounts to one tube in every few
hundred.
If there was a 3 dB gain difference, it would also show up as a radical
input (going from 1:1 to 2:1 VSWR) impedance difference, since the output
impedance is set by the amount of High Voltage and plate current.
I find your personal attacks an act of desperation. You seem to be unable
to support your claims technically, since most of your statements
contradict yourself. While you openly contradict yourself, you accuse me
of smoke and mirrors.'
Let me give an example of an obvious contradiction. You claim the Chinese
tube's glass is melted by parasitics, yet you also claim any parasitic
will arc the bandswitch, tuning cap or damage the grid of the tube. In
order to melt the glass, the parasitic would have to have many hundreds of
kilovolts sustained from the anode through the glass to an external
conductor.
You allude to Vectronics using your suppressor, yet you failed to mention
after they installed your suppression system field service history did not
improve. Your suppression kit was removed from the Vectronics PA's within
several months of the initial test.
You cling to the idea that photons cause parasitics, even though a
physicist that works in that area disputed your claims. You very clearly
said you tested tubes involved in "standby flashovers", and they had no
gas. Now you have claimed they DO have gas, like the geiger-muller tube.
The gas filled geiger-muller tube has less than a pico-ampere of current
induced by radiation, yet you claim the almost perfect vacuum in a power
grid tube tube (that had no gas in the start but now does) will support a
large sustance current from external photons.
You claim in most PA's a sudden burst of VHF energy damages grids,
bandswitches, and arcs capacitors over, yet we are to believe a Chinese
3-500 has glass failure from the same cause.
I measured the very same style capacitor used in the SB-220, AL-80 series,
Vectronics PA's and later Drake PA's, and found no high impedance points
up to several hundred MHz, yet you claim these PA's can arc that capacitor
over from VHF energy.
You insist a tiny grid rated at 25 watts, will dissipate 500-700 watts.
Yet we know the massive 3-500Z anode will become incandescent long before
dissipating 500 watts.
None of your claims make technical sense, so you have now reverted to
personal slander of me and even NASA. This act of desperation is pitiful.
73 Tom
R. L. Measures
In article <19961125035...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> In article <measures-ya023180...@news.vcnet.com>,
> meas...@mail.vcnet.com (R. L. Measures ) writes:
>
> > With average-gain 3-500Zs, putting a VHF-dampening device in either the
> >input or output circuit may produce stability. However with
> above-average
> >gain 3-500Zs, using a VHF-dampening device in both the input and in the
> >output is indicated.
>
> This comment puzzles me Rich.
>
> You attribute a wide gain variation to tube differences, indicting it is
> an input to output feedback loop.
Gain-variation has nothing to do with feedback capacitance. Perhaps
this will decrease your level puzzlement, Tom: The SB-220 has a VHF
resonance in the input and a VHF resoance in the output. Both resonances
happen to be around 110MHz. Approximately 0.2pF of capacitance exists
between the input and the output. At 110MHz, 0.2pF has 8000 ohms of
reactance. I believe that occasional oscillations at 110MHz can occur in
the SB-220. The TL-922 has a similar problem at 140MHz, for the same
reason. // According to the notes from our telephone conversation in
1990, you stated that you had worked on over 400 SB-220s, and that many of
them had VHF parasitic-oscillation damage. Are you now saying that VLF is
what you meant to say?
> The worse engineering mistake in the SB-220 was the inclusion of chokes
> in paralle with capacitors from the grids to ground in the SB-220. It
> makes that PA subject to VLF parasitics, and destroys gain flatness. It
> makes the tubes less stable at VHF also. The single most important mod to
> an SB-220 is to ground the grids right at the tube pins with very very
> short leads.
>
Tom: What is the grid-resonant frequency in a SB-220 with the
factory-stock grid circuitry?
Tom: What is the grid-resonant frequency in a SB-220 with "very very short
leads" to ground the grids?
Tom: What is the frequency of the anode circuit resonance in an AL-80...?
Tom: Who is the anonymous Mr. XXX? Why does he list a ficticious Eimac
e-mail address?.
Tom: Who are/is "they"?.
Tom: What are the names of the people at Eimac, to whom you refer to as
"it's the
entire staff", who allegedly say that Mr. Foote and I are wrong about
gold-sputtering being caused by an oscillation condition?
Tom: Does the AL-1500 8877 amplifier, which you designed, use a VHF parasitic
oscillation suppressor?
Tom: You proposed calculating the L/R VHF current-distribution for a typical
3-500Z suppressor. I calculated the VHF current-distribution, and posted
the results here. Would you prefer to drop this matter?
----------------------------------
e-mail copy to Mr. Rauch
--
--Rich-- ag6k, 805-386-3734
R. L. Measures
> You allude to Vectronics using your suppressor, yet you failed to mention
> after they installed your suppression system field service history did not
> improve. Your suppression kit was removed from the Vectronics PA's within
> several months of the initial test.
>
Vectronics built LA30 amplifiers for AEA. The field service was done by
AEA---and, in November of 1991, AEA began retrofitting low VHF-Q parasitic
suppressors into LA30 amplifiers that were returned because of
parasitic-related damage. I have the sales records if you would like to
see them, Mr. Rauch. Mr. Al Chandler was the person I delt with at AEA.
As i recall, Vectronics recommended that Chandler contact me about
retrofitting lower Q suppressors into LA30 amplifiers that were returned
for service. .
Tom: who said that low Q suppressors were removed from Vectronics'
amplifiers within several months of the initial test? I provided the names
of the people I spoke with at Vectronics: engineer Jim Riach and owner Paul
Rivnak. Who did you talk to at Vectronics?
It seems a bit curious that you would be skeptical about the glass envelope
melting problem at Vectronics. I spoke to both the owner and the engineer
about this problem, as well as to Rus Healy at QST. It also seems a bit
curious that Raich and Rivnak would conclude that higher VHF-Q suppressors
were better than lower VHF-Q suppressors if they recommended that AEA
contact me about retrofitting lower VHF-Q suppressors into LA30s that were
already in the field.
Tom: Have you posted your mathematical solution to the following, yet?
Tom: Are you suggesting that the primary path of VHF current is through
the paralleled 50 ohm resistor? If you maintain that most of the VHF
current passes through the 50 ohm resistor, Tom, please show us the
mathematical solution that led you to your conclusion.
------------------------------------------------------------------
e-mail copies to Mr. Rauch, and the Good Source...... e-mailing list
--
--Rich-- ag6k, 805-386-3734
w8j...@aol.com
I have the flu, and may be a bit hazy today, but let me try to reply.
In article <measures-ya023180...@news.vcnet.com>,
meas...@mail.vcnet.com (R. L. Measures ) writes:
> Gain-variation has nothing to do with feedback capacitance.
I never said it did. It is a complex function at VHF in a large tube.
>this will decrease your level puzzlement, Tom: The SB-220 has a VHF
>resonance in the input and a VHF resoance in the output. Both resonances
>happen to be around 110MHz. Approximately 0.2pF of capacitance exists
>between the input and the output. At 110MHz, 0.2pF has 8000 ohms of
>reactance. I believe that occasional oscillations at 110MHz can occur in
>the SB-220. The TL-922 has a similar problem at 140MHz, for the same
>reason.
Rich, I think you've looked at the feedback path incompletely or
incorrectly. The input capacitance each tube is 8.3 pF, the feedback
capacitance .07 pF.
With the filament floating perfectly, and the grid grounded, capacitive
feedback is through a 118:1 voltage divider. This of course neglects the
driving impedance of the parallel tubes, perhaps 50 ohms.
The reactance of the feedback path at 100 MHz is through 11,365 ohms Xc,
and the shunt capacitance is 95.5 ohms Xc. Even if the cathode was
resonated by a parallel inductance of 95.5 ohms Xl, the 50 ohm driving
impedance would remain. Voltage division would be through a ~11,000 to 50
ohm divider (with about 90 degrees phase shift).
So if we perfectly resonated the anode and cathode at 100 MHz, we would
have over 46 dB path loss.
Does the gain of a 3-500 tube exceed 46 dB?
What you have failed to see is the oscillation path is not via feedback to
the cathode, it is to the grid. The grid simply looses it's "grounding"
ability due to the shunt capacitance and series inductance of the grid
leads and structure. The cathode is NO part of the oscillation path except
as it couples to the grid. The dominant effect is the anode grid
capacitance, and that's why the grid needs to be compact and the grid
leads need to be very short. We want to move that effect up away from the
operating frequency.
// According to the notes from our telephone conversation in
>1990, you stated that you had worked on over 400 SB-220s, and that many
of
>them had VHF parasitic-oscillation damage. Are you now saying that VLF
is
>what you meant to say?
No, I never said they had VHF damage. I think you have the same tendency
we all do, you hear what you like to hear. I only would have said I
suspected they had, on occasion, VLF parasitic damage. The grid choke
resonates with the parallel capacitors, and the anode blocking capacitor
tunes the anode via incorrect loads. With the correct load, we might have
a simple TPTG oscillator.
That issue aside, placing any parallel resonant hi Q circuit in the grid
leads is idiotic. The grid needs to be grounded with short low impedance
leads to the chassis.
That circuit came from what was a good idea, the Collins 30S1 (I think
that is the model, the PA with the 4CX1000). In that PA, it provided
negative feedback as well as grid current limiting. But it ONLY works in a
PA without grid current. Applied to the SB-220, it was a disastrous
mistake.
>Tom: What is the grid-resonant frequency in a SB-220 with the
>factory-stock grid circuitry?
I am not positive, it has been over 20 years since I looked at the 220. As
I recall it was about 250 kHz with several lower Q resonances below the
BCB, and the VHF resonance moved down to less than 90 MHz, depending on
lead dress.
>Tom: What is the grid-resonant frequency in a SB-220 with "very very
short
>leads" to ground the grids?
Over 100- 120 MHz.
>Tom: What is the frequency of the anode circuit resonance in an
AL-80...?
That is a complex circuit. There are various frequencies where the
reactance crosses zero. Be more specific.
> Tom: You proposed calculating the L/R VHF current-distribution for a
>typical
>3-500Z suppressor. I calculated the VHF current-distribution, and posted
>the results here. Would you prefer to drop this matter?
Current division using simple L/R values is nothing Rich. You used the
extreme low reactance range of the suppressor and did it at the low
frequency end. Even at that, currents were almost equal. The Q of the
non-nichrome suppressor was almost one.
Your claim deals with suppressor Q. My claim was that the resistance
dominated the circuit at VHF in a properly designed suppressor, that it
was the primary path for VHF currents..while the coil was the primary path
for HF currents. You pushed the example to the edge, still the Q was near
ONE with a small conventional suppressor.
Rich, a Q of ~1 is NOT high Q. The conventional suppressor had a Q of 1.14
with a lossless coil. With 5 ohms ESR nichrome it dropped to 1. Is that
significant?
Do you really think that very small VHF Q drop has anything to do with
less arcing and smoother tuning, or do you think you might be accidentally
lowering HF Q?
When you did you suppressor measurements on you Q meter, did you have the
resistor in place? If not, your measurements were totally flawed.
I posted the resulting Q change from adding nichrome. Did you find my
figures flawed in any way? How much Q change does the nichrome produce?
Tell me the wire gauge and I'll do it precisely at two frequencies, at 30
and 100 MHz. Let's see if the nichrome lowers Q more at 100 MHz, or at 30
MHz.
Or would you rather forgo that, and talk about advantages of nichrome. I
can think of specific applications where it would be useful for
stabilizing a PA.
As per the photons that is silly Rich. Let go of it. Photons are used to
explain any energy transfer by electromagnetic disturbances. Your antenna
radiates photons, the tank circuit does, so does the tube's filament and
anode. Photons are everywhere. You might as well blame the arcing on
vibrations. Perhaps a vibration rattles an electron off the cathode and it
hits the anode causing a parasitic. That could be why tubes don't last as
long in mobile applications, the vibrations shake electrons off and start
a parasitic, and the parasitic causes high currents that break the
elements of the tube.
When I drop a tube off a shelf, it breaks into oscillation. Tubes should
be packed in nichrome packing material, then if they are dropped
parasitics won't cause them to fail. That makes just as much sense.
73 Tom
w8j...@aol.com
Rich, I think you overlooked this post.
You arbitrarily assigned a value of 50 ohms to the resistor, and .070 uH
to the inductor. The AL-80 series uses a 100 ohm resistor, not 50. The
inductance of later units was indeed reduced to ~85 nH, mainly due to
> At DC, nichrome alloys are about 57 times more resistive than copper. To
>the best of my knowledge, there is no type of wire that has more VHF
>resistance, and produces a lower VHF-Q on a Boonton Q-Meter, than
>nickle-chromium-iron alloys.
If you measured the coil WITHOUT the loading resistor, as I recall you
saying, your measurement method was flawed. Add the loading resistor, and
measure the CIRCUIT. You will find almost no measureable change in Q until
you lower the Boonton's frequency to HF. If you send me a suppressor, I'll
measure it on my HP or Wiltron, they go up over 1 GHz.
I think though we can safely conclude making the coil out of nichrome
mainly lowers HF Q, NOT VHF Q.
If any of my numbers are wrong, let me know.
73 Tom
R. L. Measures
In article <19961126043...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> I have the flu, and may be a bit hazy today, but let me try to reply.
> In article <measures-ya023180...@news.vcnet.com>,
> meas...@mail.vcnet.com (R. L. Measures ) writes:
> > Gain-variation has nothing to do with feedback capacitance.
>
> I never said it did. It is a complex function at VHF in a large tube.
>
...snip...
>
> Rich, I think you've looked at the feedback path incompletely or
> incorrectly. The input capacitance each tube is 8.3 pF, the feedback
> capacitance .07 pF.
I have seen spec sheets for 0.07pF to 0.1pF. Are you saying that 2 times
0.07pF is 0.07pF? However, it's a moot point because although feedback C
is applicable at HF., at VHF, it's a different ballgame. At 110MHz, add
300nH of lead-inductance to the cathode and the 8.3pF vanishes---and in the
SB-220, the cathode dips at 110MHz. Oops.
Your explanation of the VHF un-oscillation mechanism in the 3-500Z is too
hazy for me to understand. /// Did anyone understand it?
> // According to the notes from our telephone conversation in
> >1990, you stated that you had worked on over 400 SB-220s, and that many
> of
> >them had VHF parasitic-oscillation damage. Are you now saying that VLF
> is
> >what you meant to say?
>
> No, I never said they had VHF damage.
That raises an interesting possibility, Tom. .......After the caller who
identified himself as Tom Rauch agreed with me that Heath SB-220s have
parasitic oscillation problems which damages components, I told the caller
that I had some bad news. I said: Tom Rauch uses virtually the same
suppressor-design that Heath uses in the SB-220. At this point, the caller
appeared to "loose it"---rather convincingly I might add. The caller began
speaking so loudly that I had to remove the receiver from my ear. I laid
the receiver down on the table. A third party, who was across the room
could hear the caller's voice plainly.
But of course----if the person who telephoned me was an imposter, then it
is absolutely, positively, 100% true that the real Tom Rauch did NOT say
that SB-220s suffer from damage due to parasitic oscillations. Since I
mentioned the telephone caller loosing-it incident in my rebuttal, now I
think I understand why you wrote the letter threatening me and QST with a
libel lawsuit if my rebuttal was ever published in QST. ........... OTOH,
the caller was curiously insistent that I NOT tape record the conversation.
> >Tom: What is the frequency of the anode circuit resonance in an
> AL-80...?
>
> That is a complex circuit. There are various frequencies where the
> reactance crosses zero. Be more specific.
>
Nah, it's a simple check with a dipmeter, Tom. I'm beginning to get the
impression that someone is trying to complicate simple things in order to
lay a smokescreen.
Tom: Does the AL-1500 amplifier, which you designed, use a VHF
parasitic-oscillation suppressor?
------------------------------------------
e-mail copies to Mr. Rauch and the Good Source...... mailing list.
--
--Rich-- ag6k, 805-386-3734
R. L. Measures
In article <19961126144...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> Rich, I think you overlooked this post.
>
> You arbitrarily assigned a value of 50 ohms to the resistor, and .070 uH
> to the inductor. The AL-80 series uses a 100 ohm resistor, not 50. The
> inductance of later units was indeed reduced to ~85 nH,
...snip...
What is the anode resonant frequency in the AL80? You find this with a
dipmeter coupled to either side of the dc blocking capacitor. I can not
calculate the current distribution without knowing the frequency of the
anode resonance. If you do not have access to a dipmeter and an AL80,
please let me know and I will call around and see if I can find someone who
does.
>
> Your calculations are exact for the values "pulled from your hat", but the
> actual shift in current varies with the actual supperssor.
The values of L and R is the actual values we use in the SB-220, TL-922,
2K-4, and similar 2, 3-500Z amplifiers which use a 3000v anode supply.
> A second factor enters the equation, and this is the critical area. Adding
> nichrome to the primary HF path does almost NOTHING to system Q.
...snip...
First things first. I need to calculate the current distribution for 85nH
and 100 ohms, but I can not do it until someone uses a dipmeter and
measures the anode resonant frequency in the amplifier that this suppressor
is used in.
w8j...@aol.com
>Tom: Who is the anonymous Mr. XXX? Why does he list a ficticious Eimac
>e-mail address?.
>
>Tom: Who are/is "they"?.
>
>Tom: What are the names of the people at Eimac, to whom you refer to as
>"it's the
>entire staff", who allegedly say that Mr. Foote and I are wrong about
>gold-sputtering being caused by an oscillation condition?
Rich,
I tried to avoid this, because it would leave you no way out. I really
don't enjoy this, but I see no alternative. You've pushed too hard on this
issue and impugned my credibility with "stories", so you can have what you
demanded.
The former *R+D engineering manager*, Buzz Miklos, of the Salt Lake City
Varian plant has agreed to speak to one or two people on the telephone
about your theories. He said he will also come on line, but since he just
took a position as engineering manager with another tube manufacturer in
PA, it will be sometime in January before he is able to connect directly
to this newsgroup.
But if you want a quick answer, let's find someone to talk to Mr. Miklos.
A second person still at Varian Salt Lake, John Button, will confirm Mr.
Miklos' employment status with Varian, and the fact that Mr. Miklos was
indeed R+D engineering manager for Varian and a long term employee in
power grid tube design and research. Mr. Brandon will also conform Mr.
Miklos' qualifications. Incidentally, both Mr. Button and Reid disagree
with you completely also.
Mr. Miklos feels, and says his Eimac engineering staff felt, your theories
have no merit and display a complete misunderstanding of how a power grid
tube operates and the mechanisms of tube failures.
He prefers you not receive his phone number, since he understands you have
more time to argue than he does. He will, however, call you if you always
promise to include his views whenever you mention the person assigned to
answer customer letters, Mr. Foote.
You wanted to know who at Varian disagrees with you. You now have what you
have asked for...and perhaps more than you wanted.
Mr. Miklos said my interpretation of grid damage was correct, that the
8877 can only dissipate perhaps two times the rated dissipation (50 watts)
before grid damage occurs. He further stated a major cause of gold
sputtering is poor alignment between the grid and cathode bands in the
tube, or grid dissipations above the 35-50 watt range.
He also says, as I did, gold migrates slowly on its' own accord from
normal electron bombardment. Excessive grid current speeds this process.
Excessive dissipation makes the gold migration fatal to the cathode in a
very short time.
Let's pick someone to call Mr. Miklos. He will confirm all of this, and as
much more as needed. Or we can drop all this parasitic nonsense and do
better things.
Your move.
73 Tom
R. L. Measures
In article <19961127002...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> In article <measures-ya023180...@news.vcnet.com>,
> meas...@mail.vcnet.com (R. L. Measures ) writes:
>
> >Tom: Who is the anonymous Mr. XXX? Why does he list a ficticious Eimac
> >e-mail address?.
> >
> >Tom: Who are/is "they"?.
> >
> >Tom: What are the names of the people at Eimac, to whom you refer to as
> >"it's the
> >entire staff", who allegedly say that Mr. Foote and I are wrong about
> >gold-sputtering being caused by an oscillation condition?
>
> Rich,
>
> I tried to avoid this, because it would leave you no way out. I really
> don't enjoy this, but I see no alternative.
...snip...
OK. I will deal with Mr. Miklos when he comes on the Internet in January.
What is Mr. 'Buzz' Miklos' given first name? Who was Mr. Miklos'
immediate supervisor when Mr. Miklos was allegedly employed as"R+D
engineering manager" at Varian-Eimac's Salt Lake City division?
Tom:
> Mr. Miklos said my interpretation of grid damage was correct, ... ...
Congratulations, Tom. However, if Mr. Miklos won't be on the Internet
until January, how would he be reading this debate? ........ Oops.. ...
You said that Mr. Miklos "further stated a major cause of gold sputtering
is poor alignment between the grid and cathode bands " I was not aware
that Varian-Eimac was selling amplifier tubes with poor grid/cathode
alignment. This is a surprising revelation, Tom. In earlier times,
Eimac-quality was excellent. In fact, all of the gold-sputtered tubes I
cut open and inspected for the article "Parasitics Revisited" appeared to
have correct alignment between the grid and the cathode.
Tom: You said "I tried to avoid this,... ..."---like hang on to your
hats, folks, here it finally comes. (trumpets heard off-stage) However,
after the smoke and fanfare drifted away, it was abundantly clear to us
that you did NOT reveal the name of the anonymous Mr.XXX.
Tom: What is the anode-resonant frequency in the AL-80? Why do you appear
to be avoiding the answer to this question?
Tom Does the AL-1500 8877 amplifier, which you designed, use a VHF
parasitic oscillation suppressor? Why do you appear to be avoiding the
answer to this question?
------------------------------------
e-mail copies to Mr. Rauch and the Good Source... mailing list
--
--Rich-- ag6k, 805-386-3734
Roger Stevens
Let's see here. Is this a 15-rounder, or an unsanctioned non-title fight?
I would have to deduct a point from Mr. Measures for taunting, and award a
point to Mr. Rauch for even-temperedness.
However, on these and one other issue on which I am in need of guidance
(whether or not to short unused turns in a T-match rotary inductor), I must
plead OJ-Jury-Syndrome: it's like the DNA evidence--too complicated--discount
it, effectively throwing it out. All I know for sure is that QRO amplifiers
look great on their web page, and I need an amp. Although they sure have saved
some parts in that HF-2500 with no tuned input circuits.
While Nero fiddles, Svetlana may be getting their act together, which could be
good news for all of us (cheaper amps).
Just one idiot's opinion.
73, Roger WA3FLE
Norristown, PA
(Visit my webpage at http://www.voicenet.com/~rstevens)
In article <19961127002...@ladder01.news.aol.com>, w8j...@aol.com
says...
>
>In article <measures-ya023180...@news.vcnet.com>,
>meas...@mail.vcnet.com (R. L. Measures ) writes:
>
>>Tom: Who is the anonymous Mr. XXX? Why does he list a ficticious Eimac
>>e-mail address?.
>>
>>Tom: Who are/is "they"?.
>>
>>Tom: What are the names of the people at Eimac, to whom you refer to as
>>"it's the
>>entire staff", who allegedly say that Mr. Foote and I are wrong about
>>gold-sputtering being caused by an oscillation condition?
>
>Rich,
>
>I tried to avoid this, because it would leave you no way out. I really
--
-----------------------------------------------------
Roger W. Stevens (rste...@voicenet.com)
amprnet: wa3...@wa3fle.ampr.org [44.80.12.33]
ax25: WA3FLE@WB3JOE.#EPA.PA.USA.NA
"Be a Hero--Be a Donor!"--Mickey Mantle
Visit my homepage at http://www.voicenet.com/~rstevens
-----------------------------------------------------
the...@aol.com
In article <19961127002...@ladder01.news.aol.com>,
w8j...@aol.com writes:
>
>Your move.
>
>73 Tom
Great work Tom!
I follow this thread with anticipation as do many more.
I can't talk 'tubes' like you and Rich do but I do some repair on
amps you all have mentioned.
Rich's grid dissipation rules would sell alot of replacement tubes.
Although I'm sure the warranty would fade away fast.
Enough of that.
Can you tell me why 4 472B PENTA tubes in a Clipperton L lasted
only a few days. I replaced the original tubes and they are still working
but at reduced output which was the reason to buy the new tubes. The
new tubes fell to 100 watts out on 10. Nothing was changed in the
amp..nothing.
Same exciter - FT-767.
Just bad tubes?
Carl The2x4
N9EFJ
w8j...@aol.com
>n article <19961126043...@ladder01.news.aol.com>,
w8j...@aol.com
>wrote:
>
>> I have the flu, and may be a bit hazy today, but let me try to reply.
>> In article <measures-ya023180...@news.vcnet.com>,
>> meas...@mail.vcnet.com (R. L. Measures ) writes:
>> > Gain-variation has nothing to do with feedback capacitance.
Tom said
>> I never said it did. It is a complex function at VHF in a large tube.
>> Rich, I think you've looked at the feedback path incompletely or
>> incorrectly. The input capacitance each tube is 8.3 pF, the feedback
>> capacitance .07 pF.
>
>I have seen spec sheets for 0.07pF to 0.1pF. Are you saying that 2 times
>0.07pF is 0.07pF? However, it's a moot point because although feedback C
>is applicable at HF., at VHF, it's a different ballgame. At 110MHz, add
>300nH of lead-inductance to the cathode and the 8.3pF vanishes---and in
the
>SB-220, the cathode dips at 110MHz. Oops.
That's either technical stupidity or the obvious "distortion of facts
game" you like to play Rich.
1.) First, you indicated the P/K feedthrough C was the problem, not me.
2.) I gave the feedback Xc for 0.14 pF (two tubes) , and the shunt
resistance for the driving impedance of two tubes if the cathode was
resonant.
3.) The feedthrough capacitance is internal to the tube, and so the lead
reactance can not cancel the anode-cathode capacitance.
4.) Resonating the cathode would have little or no effect on stability in
a GG PA, the gain is too low and the feedback of the wrong phase.
5.) I actually intentionally resonate the cathode at the third harmonic of
ten and fifteen meters in some PA's to increase efficiency without hurting
IMD or stability.
As usual, your wild guess falls apart upon examination.
>Your explanation of the VHF un-oscillation mechanism in the 3-500Z is too
>hazy for me to understand. /// Did anyone understand it?
Let me clarify it for you. The cathode to anode capacitance is almost
always an insignificant part of any stability problem. It is primarily a
grid anode problem, but it is a complex problem.
The only case I recall where it is becomes a problem is in the Amp Supply
LA-1000 where the control grid is tied to the cathode directly. In that
PA, it will oscillate on ten and 15 meters if the loading control is
closed and the tuning control rotated. But then, the control grid is tied
to the cathode directly. This is still not a VHF oscillation, but rather
an HF oscillation.
>> // According to the notes from our telephone conversation in
>> >1990, you stated that you had worked on over 400 SB-220s, and that
many
>> of
>> >them had VHF parasitic-oscillation damage. Are you now saying that
VLF
>> is
>> >what you meant to say?
Are these "notes" as accurate and straight forward as your "notes" from
the Foote conversation? Did you save "notes" from the Eimac conversations
where, in long telephone conversations, the staff explained why you were
incorrect?
>> No, I never said they had VHF damage.
>
>That raises an interesting possibility, Tom. .......After the caller
who
>identified himself as Tom Rauch agreed with me that Heath SB-220s have
>parasitic oscillation problems which damages components, I told the
caller
>that I had some bad news.<snip>
Rich, you seem to be losing control. Wild accusations and personal attacks
on my character have no place in technical discussions. Your theories
either stand on their own merit, or they don't.
**The only thing you were ever told, was if you personally slander or tell
malicious lies about anyone it might be cause for legal action. That was
because of your history of inventing or distorting stories to impugn other
persons character.**
Life is quite simple Rich. If you confine yourself, like an intellegent
gentleman, to the facts and issues, you don't have to be threatened. Say
anything you want technically, and anything you say will be fine.
The very reason the person at CPI requested I not allow his name into this
conversation is he felt you have a history of public exaggeration,
personal attack and slander. That is also the very reason people at Eimac
are hesitant to speak to you, they are concerned about being slandered or
personally attacked for merely disagreeing with you. They are afraid that
what they say in private conversation will be distorted to suit your
personal needs and aired publicly.
You, after failing to respond with good technical answers to questions,
have now taken the approach of "if I can't get Tom technically I'll get
him personally".
You may have no idea how that behavior makes you look. Most people are
intellegent enough to realize technical discussions are technical, and not
personal. If I came at you like that, no matter how correct I was
technically, most people would be disgusted and write me off as a complete
ass.
One offer remains, the former R+D engineering manager has agree to talk to
one or two people and explain that your theories are incorrect. He refuses
to allow you his phone number (because of the behavior outlined above),
but will call you.
If you wish to end this at this point, I understand. I have to say I hoped
you would remain a gentleman, I am very disapponted in your behavior. This
is not 40 meters. Calm down and think before you reply, please don't be so
emotional in public. If you want to cuss me out or call me names, you can
do it over the telephone again. But you'll have to pay for the call ! :-)
If you wish to have a technical discussion, get back to the issues and out
of the personal attack mode. Do you still want Mr. Miklos and others
input? It can now be arranged. You have what you asked for, Varian's
engineering and tube designer's final word.
73 Tom
R. L. Measures
In article <57gokd$d...@news1.voicenet.com>, rste...@voicenet.com (Roger
Stevens) wrote:
> Let's see here. Is this a 15-rounder, or an unsanctioned non-title fight?
>
> I would have to deduct a point from Mr. Measures for taunting, and award a
> point to Mr. Rauch for even-temperedness.
Mr. Rauch has repeatedly not answered simple questions like whether or not
he incorporated a VHF parasitic- oscillation suppressor in the design of
his AL-1500 amplifier. The answer is either yes or no.
Mr. Rauch proposed that I calculate the VHF current distribution in the
AL-80's parasitic suppressor. Three things are needed to do the
calculations. L, R and the frequency (of the anode resonance). Mr. Rauch
states that L is 85nH, that R is 100 ohms, but despite an abundance of
requests from me, he has so far not stated what the frequency is.
€ Roger: how many points would you award to each pugilist for evasiveness?
>
> However, on these and one other issue on which I am in need of guidance
> (whether or not to short unused turns in a T-match rotary inductor), I must
> plead OJ-Jury-Syndrome: it's like the DNA evidence--too complicated.
If you enjoy seeing eyebrow-raising, eye-popping, coil-to-frame arcing at
the open end of the inductor, DO NOT short the unused turns. OTOH, if you
would rather loose a few watts, and forgo the arcing, short the unused
turns.
However, I absolutely, positively, 100% have no idea whatsoever how one can
gash one's finger on a cellphone.
Alas, this debate was not my idea, Roger. Since 1990, Mr. Rauch has
indicated that he wanted a public debate with yours truly about the subject
of parasitics. Debating with Mr. Rauch has so far been like trying to play
poker with a guy who proudly declares he has the winning hand, but refuses
to lay his cards on the table.
-----------------------------------------
e-mail copies to Messrs. Clark, Rauch and the Good Source... mailing list.
--
--Rich-- ag6k, 805-386-3734
w8j...@aol.com
Hi Carl,
Would you (or any other person) like to talk to Mr. Miklos about tubes,
and report back to us about his opinions of Measures articles and
suggestions? He hasn't the time to get on now, but he would be willing to
answer questions on the land line.
If you had time, a few other Eimac and ITT design engineers might also be
located. You could tell us in your own words what their opinions of Rich's
theories are.
In article <19961127154...@ladder01.news.aol.com>, the...@aol.com
writes:
>Can you tell me why 4 472B PENTA tubes in a Clipperton L lasted
>only a few days. I replaced the original tubes and they are still working
>but at reduced output which was the reason to buy the new tubes. The
>new tubes fell to 100 watts out on 10. Nothing was changed in the
>amp..nothing.
>Same exciter - FT-767.
>Just bad tubes?
I don't know Carl. Anything I say would be a guess, unless I had much more
information. I am working with some 572B's now in a homebrew PA, but the
circuit I have used is quite different than the Dentron.
I don't think highly of the Dentron design. It has a very poor layout, and
the tubes are not neutralized. I am also not sure of the reliabilty or
service history of Penta 572B's.
I'd like to have any information I could about them. Did the tubes work
correctly initially? Can you describe in detail how the failed? Do the old
tubes still work the same? Who was the vendor, what was the country of
origin?
73 Tom
w8j...@aol.com
>What is the anode resonant frequency in the AL80? You find this with a
>dipmeter coupled to either side of the dc blocking capacitor. I can not
>calculate the current distribution without knowing the frequency of the
>anode resonance. If you do not have access to a dipmeter and an AL80,
>please let me know and I will call around and see if I can find someone
who
>does.
>First things first. I need to calculate the current distribution for
85nH
>and 100 ohms, but I can not do it until someone uses a dipmeter and
>measures the anode resonant frequency in the amplifier that this
suppressor
>is used in.
Resonance tells us nothing about impedance, except the reactance cancels
or crosses zero. Even though you do not understand VAR power, phase shift,
or Q I still expect you should understand a very simple concept like
resonance!
A circuit at resonance could look like zero ohms, infinite impedance, or
anything in between. So even if we know the frequency of resonance, we
know nothing about the impedance of the circuit and can do no
calculations.
Why ask for meaningless data? What is it you want to do, and I will tell
you what numbers you need to do that.
Tom
R. L. Measures
In article <19961127183...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> In article <measures-ya023180...@news.vcnet.com>,
> meas...@mail.vcnet.com (R. L. Measures ) writes:
>
> >What is the anode resonant frequency in the AL80? You find this with a
> >dipmeter coupled to either side of the dc blocking capacitor. I can not
> >calculate the current distribution without knowing the frequency of the
> >anode resonance. If you do not have access to a dipmeter and an AL80,
> >please let me know and I will call around and see if I can find someone
> who
> >does.
>
snip...
> Why ask for meaningless data?...snip....
---------------
At this point in the debate I need a volunteer who owns an AL-80 or
SB-1000, who has access to a dipmeter, to measure the anode resonant
frequency. (Be sure to disconnect the electric-mains plug before
performing the measurement.) The best place to measure the dip is at the
DC blocking capacitor. The resonance is probably between 110MHz and
160MHz. The dip should be sharp and pronounced. Set the tune capacitor to
about 50% of C maximum. This measurement is needed in order to be able to
solve the current division problem for the 85nH inductor and 100 ohm
resistor parasitic suppressor that Mr. Rauch suggested.
Please telephone me collect if you make this measurement.
Tom: Does the AL-1500 8877 amplifier, which you designed, use a VHF
parasitic oscillation suppressor?
-------------------------------
e-mail copies to Mr. Rauch and the Good source....... mailing list.
--
--Rich-- ag6k, 805-386-3734
R. L. Measures
> In article <19961127002...@ladder01.news.aol.com>,
> w8j...@aol.com writes:
>
> >
> >Your move.
> >
> >73 Tom
>
> Great work Tom!
€ Carl: the intro was great, but did you ever figure out what the name of
Mr. XXX is?
...snip...
>
> Can you tell me why 4 472B PENTA tubes in a Clipperton L lasted
> only a few days. I replaced the original tubes and they are still working
> but at reduced output which was the reason to buy the new tubes. The
> new tubes fell to 100 watts out on 10. Nothing was changed in the
> amp..nothing.
> Same exciter - FT-767.
> Just bad tubes?
€ Carl: In the new, Penta 572B tubes that failed, did you see bits of
filament wires and springs at the bottom of the glass envelope?
-------------------------------------------------------
e-mail copies to Messrs. Mottsinger, Rauch and others
--
--Rich-- ag6k, 805-386-3734
w8j...@aol.com
Hi Rich,
You must be having E-mail trouble. Could you look this over an comment. It
seems to indicate you do not understand Q. Could you tell me if I did
something wrong?
>>
A second factor enters the equation, and this is the critical area. Adding
73 Tom<<<
When you answer this, and we resolve any errors I''ll move on to one of
your questions.
By the way, a grid dip meter will not allow you to calculate anything
about impedance in a complex system, surely you understand that much about
RF?
73 Tom
the...@aol.com
>
>€ Carl: In the new, Penta 572B tubes that failed, did you see bits of
>filament wires and springs at the bottom of the glass envelope?
I don't think so Rich. It has been a while. But the owner still has the
Penta tubes
and next time around I will look. I do remmeber that the tubes looked
brand new.
Do these Penta tubes have rusty ol' innards?
Carl
N9EFJ
the...@aol.com
Yes, I have the time to talk to Mr. Miklos and others.
The 572B's were installed and all appeared OK. I was not
there when they failed. The exciter was a FT-767. I did however
put the old tubes back in and they worked at their previous
output before the Pentas were installed which is about 600 watts.
These old tubes are in fact still in use after two years. Upon initial
installation
the 4 Penta's output was a little low but not enough to be concerned with.
My customer still has the old tubes and if I get a chance I'll check them
out.
Rich asked if pieces of the tube innards were floating around. I don't
think so but I'll look again. If fact the tubes looked brand new just
dead.
I wish I could tell you more.
I had a similar case last week with a SB200. In fact it is still here
waiting to
be picked up. A previous owner other than my customer had installed a
standby switch and messed up the bias placing -130 some odd volts on the
bias line at all times
My customer bought it and couldn't get the amp to put out more than 100
watts, so he bought a pair of new Russian tubes from RF PARTS then brought
it to me.
The old tubes were Certron and ended up doing about 350 on 80. Exciter was
an
FT-757. I installed the new, still in the taped up boxes, Svetlana tubes
and only got 100 watts or so output. Same exciter, same amp. What gives? I
made no attempt to delve further in to the situation. Just thought you
might provide an answer from a similar problem. I think these tubes are
terrible! I would hope they are direct replacements and I'm sure everyone
else assumes that. How can they sell them? Why do the offer them for sale?
So they are cheaper. That won't get you across
town with these tubes.
Carl N9EFJ
R. L. Measures
In article <19961128051...@ladder01.news.aol.com>, the...@aol.com
wrote:
Nope, Carl. I haven't spotted any rust. I have heard reports of virtually
new Penta 572Bs that wound up with broken filament wires. Did notice that
the light from the filaments went out in the 572Bs that failed? Were the
grids directly-grounded in the amplifier?
---------------------------------
e-mail copy to Carl, N9EFJ
--
--Rich-- ag6k, 805-386-3734
R. L. Measures
In article <19961128022...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> Hi Rich,
>
> You must be having E-mail trouble. Could you look this over an comment. It
> seems to indicate you do not understand Q. Could you tell me if I did
> something wrong?
>
Hello, Tom---
Nah, you haven't done anything wrong yet--if you are trying to play this
debate for stalemate instead of checkmate. It is true that if you continue
to stonewall embarrassing questions, stall, and smokescreen, I will
eventually walk.
Either you are reluctant to admit that you don't own a dipmeter with which
to make the anode-resonance measurement that we need to proceed, or you
don't have access to an AL-80, or you are trying to misdirect attention
away from the problem that you proposed---i.e., to find the current ratio
in the 85nH (Ls) and100 ohm (Rs) parasitic suppressor which you use in one
version of the AL-80. I am starting to lean toward the latter possibility.
Tom: In order to try and get a handle on what's happening, I will give
you the benefit of the doubt and guesstimate that the anode resonant
frequency in the AL-80 is 160MHz, although I seem to recall hearing a lower
number from AL-80 (and SB-1000) owners. Using the formula XL=2*Pi*f*L, at
160MHz, 85nH has 6.28*1.6*10^5*85*10^-9=85.4513, or roughly 90 ohms of
reactance. To find the current distribution, we can apply 1V to the
paralleled Rs and Ls. The resulting current through Rs is 1v/100 ohms =
10mA. The current through Ls is 1v/90 ohms = 11.1mA. The total current is
10mA plus 11.1mA = 21.1mA. The current through Rs is 10/21.1 = 0.474 of
the total current. The current through Ls is 11.1/21.1 = 0.526 of the
total current. If 52% of the VHF current is flowing through Ls, and 48% of
the VHF current is flowing through Rs, there is obviously a large hole in
the opinion Tom Rauch has held since 1990--i.e., that no significant VHF
current flows through a VHF parasitic-suppressor's L. The bottom-line is
that in the typical VHF parasitic-oscillation suppressor, significant
current flows in Rs and significant current flows in Ls.
However, in actual practice, the anode-resonance in the typical single
3-500Z amplifier is closer to 130MHz,, so the reactance of Ls would be
lower and Ls would carry a slightly greater share of the current---and the
hole in Tom Rauch's opinion seems to grow larger.. OTOH, it can be argued
that the intrinsic VHF-R in nichrome would slightly decrease the current in
Ls. ...........All things considered, significant current flows in Rs and
significant flows in Ls.
On page 74 in the September, 1994 issue of QST, Tom Rauch states:
"The coil's reactance increases with frequency and at VHF most of the
signal path is through the resistor. It is plainly evident that the
dominant component at VHF is the resistor, not the coil."
(note: Mr. Rauch appears to use the word "coil" in place of the word inductor)
It is PLAINLY EVIDENT that neither Tom Rauch nor QST Editor Mark Wilson
took the humble minute necessary to perform the calculations to check the
validity of Tom's statement before it became a sorrowful chapter in QST
history.
This easily-preventable blunder would NOT have happened in the days when
George Grammer, W1DF, stood watch over the technical reputation of QST
magazine. According to those who knew him, George Grammer placed nothing
above technical accuracy.
--------------------------------------
e-mail copies to Mr. Rauch et al.
--
--Rich-- ag6k, 805-386-3734
R. L. Measures
significant current flows in Ls.
Gary Coffman
In article <measures-ya023180...@news.vcnet.com> meas...@mail.vcnet.com (R. L. Measures ) writes:
>In article <575aig$8...@villa.fc.net>, fil...@paranoia.com (Filip M
>Gieszczykiewicz) wrote:
>> Greetings. I would just like to interject here that as someone who
>> grew up after tubes "died" but is still MOST interested in these
>> discussions... I'm now totally confused and lost.
>------------
>I understand the confusion, Filip. There has been plenty of smoke and
>mirrors in this debate, and I suspect there will be more. Briefly---The
>roots of the debate date from 1926, when Mr. F. E. Handy said on page 72 in
>the *Radio Amateur's Handbook*:
>
> "The combination of both resistance and inductance is very effective in
>limiting parasitic oscillations to a negligible value of current."
>
> In other words, Mr. Handy found that VHF parasitic-oscillation
>suppressors performed measurably better when the suppressor's inductor (L)
>was made from resistance wire instead of copper wire.
That does not appear to be what Mr. Handy was saying at all. The quote
indicates that a combination of an inductor and a resistor is very effective
as a parasitic suppressor. That advice is, of course, correct, and is the
way *every* parasitic suppressor is made today. Even yours. The resistance
needs to be in *parallel* with the reactance, however, not in series with
it, for best results.
>Nichrome has the
>highest volume-resistivity among resistance wire alloys, but it is not
>easily-soldered and is not commonly available..
> After a few years, Handy's observation was somehow left out of the
>Handbook. After WW2, people apparently forgot what Mr. Handy wrote in 1926
>and 1927, and people started making suppressors out of whatever was handy
>and easily soldered---i.e., copper. Some folks even started silver-plating
>suppressor inductors, evidently for cosmetic reasons. In the 1940's and
>1950's, it didn't matter much what the suppressor L was made from because
>large amplifier tubes did not exhibit very much VHF gain. With modern,
>high-gain amplifier tubes, VHF suppressors with a high VHF-dampening
>ability seem to be indicated. High VHF-dampening means low VHF Q.
Indeed, which is why there is a *paralleled* resistor across the inductor,
to lower the suppressor's *VHF Q*, or more to the point, to increase the
suppressor's *dissipation* at VHF. At *HF*, however, we want high Q so that
we don't have attenuation of the *desired* HF energy. That's why the coil
is made of copper (silver plating *is* mostly cosmetic). The reactance
of the coil is *non-dissipative*, but any series resistance in the coil
*is* dissipative. That's bad because we want HF energy to have free
passage through the reactive branch of the suppressor.
Lets look at a couple of circuits. This one is equivalent to your
nichrome coil.
Rs Xs
---------/\/\/\-----))))))---------
Q is just Rs/Xs, and for a given current I, the power dissipated in
the circuit is simply
P = Rs*I^2 + PF*Xs*I^2
Now the PF of a pure reactance is zero, so the attenuation of the
circuit is simply Rs*I^2.
Of course the value of I is dependent on the total impedance of
the branch, IE Z = R +jX, and X is dependent on frequency. The
higher the frequency, the greater the value of X, the lower the
I through the circuit, and the lower the power dissipated in
Rs. So the circuit dissipates maximum power at lower frequencies
and dissipates least at higher frequencies, exactly opposite of
what we want a VHF suppressor to do.
Now lets change things around.
Xp
x---)))))))------x
| |
---------x x-------------
| Rp |
x-----/\/\/\-----x
Now we can use a transform to get the equivalent series resistance
and reactance of this circuit like so
Rs = (Rp*Xp^2)/(Rp^2 + Xp^2)
Xs = (Rp^2*Xp)/(Rp^2 + Xp^2)
The equivalent Q is
Qe = Rs/Xs = (1/Rp) * Xp/(Rp^2 + Xp^2)^2
As Xp increases, Q falls rapidly due to the Xp^4 term in the denominator.
For a current, I, flowing through the circuit, I divides between the Xp
branch and the Rp branch in inverse proportion to their impedances, and
only in the Rp branch is any power dissipated. When Xp is high (VHF) more
current flows in Rp, and dissipation is high. When Xp is low (HF), less
current flows in Rp, and dissipation is low. PF=0 in the Xp branch of
course, so no dissipation occurs there. This is exactly what we want
a VHF suppessor to do. If we were to add series resistance to the coil,
as you do, then our dissipation at low frequencies would also be high,
which we don't want since it adds nothing to stability but does decrease
HF efficiency.
Gary
--
Gary Coffman KE4ZV | You make it, | Due to provider problems
Destructive Testing Systems | we break it. | with previous uucp addresses
534 Shannon Way | Guaranteed! | Email to ke...@radio.org
Lawrenceville, GA 30244 | |
R. L. Measures
In article <1996Nov28.2...@ke4zv.atl.ga.us>, ga...@ke4zv.atl.ga.us
(Gary Coffman) wrote:
> In article <measures-ya023180...@news.vcnet.com>
meas...@mail.vcnet.com (R. L. Measures ) writes:
> >In article <575aig$8...@villa.fc.net>, fil...@paranoia.com (Filip M
> >Gieszczykiewicz) wrote:
> >> Greetings. I would just like to interject here that as someone who
> >> grew up after tubes "died" but is still MOST interested in these
> >> discussions... I'm now totally confused and lost.
> >------------
> >I understand the confusion, Filip. There has been plenty of smoke and
> >mirrors in this debate, and I suspect there will be more. Briefly---The
> >roots of the debate date from 1926, when Mr. F. E. Handy said on page 72 in
> >the *Radio Amateur's Handbook*:
> >
> > "The combination of both resistance and inductance is very effective in
> >limiting parasitic oscillations to a negligible value of current."
> >
> > In other words, Mr. Handy found that VHF parasitic-oscillation
> >suppressors performed measurably better when the suppressor's inductor (L)
> >was made from resistance wire instead of copper wire.
>
> That does not appear to be what Mr. Handy was saying at all. The quote
> indicates that a combination of an inductor and a resistor is very effective
> as a parasitic suppressor.
Mr. Handy was talking about the suppressor inductor itself. I was talking
about the suppressor inductor itself For a grid parasitic-suppressor, F.
E. Handy recommends a100 ohm resistor and 10 ohms per foot resistance wire
to wind the suppressor inductor. Quoting "The grid chokes in the
photograph are most effective of all as a coil of resistance wire is used."
> That advice is, of course, correct, and is the
> way *every* parasitic suppressor is made today. Even yours. The resistance
> needs to be in *parallel* with the reactance, however, not in series with
> it, for best results.
>
This is the position of Mr. Rauch, who insists that insignificant VHF
current flows through the suppressor L. Gary: Do you believe that
insignificant VHF current flows through the suppressor L?
> Indeed, which is why there is a *paralleled* resistor across the inductor,
> to lower the suppressor's *VHF Q*, or more to the point, to increase the
> suppressor's *dissipation* at VHF. At *HF*, however, we want high Q so that
> we don't have attenuation of the *desired* HF energy.
I agree that we don't want HF attenuation. However, the additional loss
created by a low VHF Q suppressor inductor is on the order of 1% at 29MHz
and 0.2% at 14MHz. I found this acceptable because lower VHF Q suppressors
eliminated occasional parasitic arcing in my SB-220 on 40m. If you feel
that a 1% loss is unacceptable at 29MHz, then you should probably not be
using lossy suppressor inductors. // Low VHF Q suppressors also eliminated
the pecuilar double-peak in 40m output during tune-up. Other SB-220 owners
have reported similar results in both of these areas.
Of late, I have seen plenty of seemingly-questionable VHF Q theories
proffered hereabouts. If you have access to a dipmeter, Gary, you may (or
may not) be interested in comparing the dip between a VHF suppressor L made
from copper wire, and a VHF suppressor L made from nichrome resistance
wire. If you are curious, but you need a low VHF Q suppressor to make the
comparison, let me know for which amplifier. I will build a low VHF Q
suppressor and mail it to you--free. The 55-cent stamp is on me. This
offer is good whether or not you report your findings on this newsgroup.
------------------------------
e-mail copy to Mr. Coffman, Mr. Rauch, et al.
--
--Rich-- ag6k, 805-386-3734
R. L. Measures
> Yes, I have the time to talk to Mr. Miklos and others.
Mr. Miklos needs to speak for Mr. Miklos. Mr. Rauch proffers the story
that Mr. Miklos can't get on the Internet until January. I am on ISP
number three. It takes me about an hour to get set up with a new ISP---and
I am no computer genius by any means. Apparently, Mr. Rauch does not even
know Miklos' first name or the name of Miklos' ex-supervisor at
Varian-Eimac's Salt Lake City div. Hmmm. I suspect that the story about
Miklos is a Rauchian ruse to play for time---like the "Tom: does the
AL-1500 amplifier use a VHF suppressor?" issue. Tom knows that if he
answers yes, then I am going to point out that he said a parasitic
suppressor was not needed in an 8877 amplifier. Mr. Rauch knows full well
that eventually I will tire of stalling, bs, and hogwash---whereupon I will
bail out---whereupon Mr.Rauch will proudly proclaim a draw. // BTW,
Carl, would you like to take a wild guess whether or not Rauch's AL-1500
amplifier uses a VHF suppressor?
> The 572B's were installed and all appeared OK. I was not
> there when they failed.
snip
> the 4 Penta's output was a little low but not enough to be concerned with.
> My customer still has the old tubes and if I get a chance I'll check them
> out.
snip
> I wish I could tell you more.
Did the customer report hearing a noise at the time the tubes died? If you
do not recall whether you noticed the filaments lit up in the dead tubes,
Carl, then we will have to wait for the second encounter.
I have heard that the newer 572Bs have a somewhat higher mu than the Cetron
and RCAs. One indication of higher mu is if you see a change in input SWR
when the new tubes are in use. The higher the mu, the less cathode to
filament RF voltage is needed to drive the tubes to the same cathode
current. Thus, the driving Z decreases with higher mu tubes, and that
affects input SWR. With a solid-state driver, higher input SWR means less
RF out, so output from the amplifier should decrease accordingly.
Higher mu also means higher gain, which means more ability to oscillate.
RF Parts is reportedly aware that the failure of Penta 572Bs in SB-200s is
parasitic-related. Cetron tubes are not a sure cure, either. I have heard
of a number of SB-200 owners who had the bandswitch's Wafer A contacts for
the 100pF tune-cap. padder, burn up from parasitic arcing when brand new
Cetrons were installed. The next time you work on a SB-200 with sudden
tube failure or a burned bandswitch, take a dipmeter and check the VHF dip
at the 1000pF, 5kV DC blocking capacitor. Is the dip sharp or broad?
w8j...@aol.com
In article <measures-ya023180...@news.vcnet.com>,
meas...@mail.vcnet.com (R. L. Measures ) writes:
> Nah, you haven't done anything wrong yet--if you are trying to play this
>debate for stalemate instead of checkmate. It is true that if you
continue
>to stonewall embarrassing questions, stall, and smokescreen, I will
>eventually walk.
And while you say that, you avoid answering the question about Q.
Interesting.
>Either you are reluctant to admit that you don't own a dipmeter with
which
>to make the anode-resonance measurement that we need to proceed, or you
>don't have access to an AL-80, or you are trying to misdirect attention
>away from the problem that you proposed---i.e., to find the current ratio
>in the 85nH (Ls) and100 ohm (Rs) parasitic suppressor which you use in
one
>version of the AL-80.
I have network analyzers and other RF impedance measurement devices. And
yes, I even have a grid dip meter.
The point remains what can you possibly do with a grid dip meter Rich? It
won't tell you a thing about the impedance of the system, other than
reactance crosses zero at some frequency near the dip. Big deal.
Why waste time with a useless exercise. You came to circuit analysis
class, put away your drycell and buzzer and get out an instrument.
> Tom: In order to try and get a handle on what's happening, I will give
>you the benefit of the doubt and guesstimate that the anode resonant
>frequency in the AL-80 is 160MHz
You have an uncanny physic ability Rich. You can guess things that
instruments can not measure. I'm sure, when you know the resonant
frequency, you will "guess" the real part of the impedance. And the
imaginary pats distributed in the system as well.
Will this new "guess" be better than your guess about the photons that
cause tube arcs in a tube that has no gas. Or like that guess, and one
that conveniently allows you to change it to a tube WITH some gas.
Is this like your "guess" that Eimac rates the grid of an 8877 at 25
watts, but it is really 500-700 watts?
>although I seem to recall hearing a lower
>number from AL-80 (and SB-1000) owners. Using the formula XL=2*Pi*f*L,
at
>160MHz, 85nH has 6.28*1.6*10^5*85*10^-9=85.4513, or roughly 90 ohms of
>reactance. To find the current distribution, we can apply 1V to the
>paralleled Rs and Ls. The resulting current through Rs is 1v/100 ohms =
>10mA. The current through Ls is 1v/90 ohms = 11.1mA. The total current
is
>10mA plus 11.1mA = 21.1mA. The current through Rs is 10/21.1 = 0.474 of
>the total current. The current through Ls is 11.1/21.1 = 0.526 of the
>total current. If 52% of the VHF current is flowing through Ls, and 48%
of
>the VHF current is flowing through Rs, there is obviously a large hole in
>the opinion Tom Rauch has held since 1990--i.e., that no significant VHF
>current flows through a VHF parasitic-suppressor's L.
You do not even understand the concept of current flow in a circuit with
reactance. All you have shown is the Q is just over one.
You made a mistake in the total current. The coil is an energy storing and
releasing reactance, not a resistance. The amount of energy storage is
given by the Q. Perhaps the lack of understanding of circuit behavior is
at the root of your bad science???
The Q is just barely over one. How much lower can nichrome make the Q go?
What is the lowest practical or possible Q?
You have claimed, before the whole world, that your suppressors lower VHF
and UHF Q. You claim all sorts of system benefits due to VHF and UHF
system changes, yet you just demonstrated the Q is near unity in the
conventional suppressor. If the Q is near one, we know adding nichrome
can't lower it any significant amount. If nichrome can't lower the Q, then
all your claims are wrong.
***The Q is already so low if we remove the coil entirely and make the
suppressor a pure non-inductive and capacitive resistor we can only reach
a Q of zero!!***
If we include the rest of the system, we would find even less change than
that, because the Q could not even reach zero.
> It is PLAINLY EVIDENT that neither Tom Rauch nor QST Editor Mark Wilson
>took the humble minute necessary to perform the calculations to check the
>validity of Tom's statement before it became a sorrowful chapter in QST
>history.
So now it is Mark Wilson also. You list of bad guys is getting pretty
long. How many people are in on this conspiracy, a dozen? Two?
Rich, do a list for us. List all the engineers who have actually said they
support you (not those who you GUESS would) in one group, and all those
that don't in another.
> This easily-preventable blunder would NOT have happened in the days when
>George Grammer, W1DF, stood watch over the technical reputation of QST
>magazine. According to those who knew him, George Grammer placed nothing
>above technical accuracy.
How do you know Grammer would support you? Is it the same way you know
about photons and Q?
Quit blaming everyone else in the world for your mistakes. You should
applaud QST for giving you a moment of glory.
73 Tom
Roger Stevens
A little trivia here--my mother was George Grammer's secretary during WWII (or
was it just prior to), during which time she met my father, Arthur J.
Stevens, Jr., who, fresh out of M.I.T., was doing research at Remington-Rand on
early radar designs. She subsequently became the current Mrs. Stevens sometime
in 1945. Prior to that event, she was Jane Guglielmetti, of Durham, CT.
She has always had the highest regard for Mr. Grammer, by the way, as an
employer, and as an individual. But the, those New Englanders tend to stick
together.
Always wanted to share this with someone.
Now, back to the action.
73, Roger WA3FLE
Norristown, PA
In article <measures-ya023180...@news.vcnet.com>,
meas...@mail.vcnet.com says...
>
>In article <19961128022...@ladder01.news.aol.com>, w8j...@aol.com
>wrote:
>
>> Hi Rich,
>>
>> You must be having E-mail trouble. Could you look this over an comment. It
>> seems to indicate you do not understand Q. Could you tell me if I did
>> something wrong?
>>
>Hello, Tom---
>
> Nah, you haven't done anything wrong yet--if you are trying to play this
>debate for stalemate instead of checkmate. It is true that if you continue
>to stonewall embarrassing questions, stall, and smokescreen, I will
>eventually walk.
>
>Either you are reluctant to admit that you don't own a dipmeter with which
>to make the anode-resonance measurement that we need to proceed, or you
>don't have access to an AL-80, or you are trying to misdirect attention
>away from the problem that you proposed---i.e., to find the current ratio
>in the 85nH (Ls) and100 ohm (Rs) parasitic suppressor which you use in one
>version of the AL-80. I am starting to lean toward the latter possibility.
>
> Tom: In order to try and get a handle on what's happening, I will give
>you the benefit of the doubt and guesstimate that the anode resonant
>frequency in the AL-80 is 160MHz, although I seem to recall hearing a lower
>number from AL-80 (and SB-1000) owners. Using the formula XL=2*Pi*f*L, at
>160MHz, 85nH has 6.28*1.6*10^5*85*10^-9=85.4513, or roughly 90 ohms of
>reactance. To find the current distribution, we can apply 1V to the
>paralleled Rs and Ls. The resulting current through Rs is 1v/100 ohms =
>10mA. The current through Ls is 1v/90 ohms = 11.1mA. The total current is
>10mA plus 11.1mA = 21.1mA. The current through Rs is 10/21.1 = 0.474 of
>the total current. The current through Ls is 11.1/21.1 = 0.526 of the
>total current. If 52% of the VHF current is flowing through Ls, and 48% of
>the VHF current is flowing through Rs, there is obviously a large hole in
>the opinion Tom Rauch has held since 1990--i.e., that no significant VHF
>current flows through a VHF parasitic-suppressor's L. The bottom-line is
>that in the typical VHF parasitic-oscillation suppressor, significant
>current flows in Rs and significant current flows in Ls.
> However, in actual practice, the anode-resonance in the typical single
>3-500Z amplifier is closer to 130MHz,, so the reactance of Ls would be
>lower and Ls would carry a slightly greater share of the current---and the
>hole in Tom Rauch's opinion seems to grow larger.. OTOH, it can be argued
>that the intrinsic VHF-R in nichrome would slightly decrease the current in
>Ls. ...........All things considered, significant current flows in Rs and
>significant current flows in Ls.
>
>On page 74 in the September, 1994 issue of QST, Tom Rauch states:
>
>"The coil's reactance increases with frequency and at VHF most of the
>signal path is through the resistor. It is plainly evident that the
>dominant component at VHF is the resistor, not the coil."
> (note: Mr. Rauch appears to use the word "coil" in place of the word
inductor)
>
> It is PLAINLY EVIDENT that neither Tom Rauch nor QST Editor Mark Wilson
>took the humble minute necessary to perform the calculations to check the
>validity of Tom's statement before it became a sorrowful chapter in QST
>history.
> This easily-preventable blunder would NOT have happened in the days when
>George Grammer, W1DF, stood watch over the technical reputation of QST
>magazine. According to those who knew him, George Grammer placed nothing
>above technical accuracy.
>--------------------------------------
>e-mail copies to Mr. Rauch et al.
>
>--
>--Rich-- ag6k, 805-386-3734
--
R. L. Measures
In article <19961129161...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> In article <measures-ya023180...@news.vcnet.com>,
> meas...@mail.vcnet.com (R. L. Measures ) writes:
>
snip...
> And while you say that, you avoid answering the question about Q.
> Interesting.
>
First things first, Tom. The Q business is coming.
> I have network analyzers and other RF impedance measurement devices. And
> yes, I even have a grid dip meter.
>
> The point remains what can you possibly do with a grid dip meter Rich?
I'm trying, I'm trying, but the guy with the AL-80 keeps stalling and
stalling, and nothing happens. The nice thing about a dipmeter is that it
does not require a physical connection to the circuit under test--as is the
case with other types of test equipment.
> It won't tell you a thing about the impedance of the system, other than
> reactance crosses zero at some frequency near the dip. Big deal.
If it's not a big deal, why not use your dipper to find the anode's
resonance point in the AL-80? If I sink, score one point for Mr. Rauch.
> Why waste time with a useless exercise.
Try me. I will be as brief as possible.
> > Tom: In order to try and get a handle on what's happening, I will give
> >you the benefit of the doubt and guesstimate that the anode resonant
> >frequency in the AL-80 is 160MHz
>
> You have an uncanny physic ability Rich.
snip
Thanks, Tom. BTW, what is a "physic ability".
The reason I had to guess at the 160 figure is because somebody is
withholding the actual anode resonant frequency.
> > Using the formula XL=2*Pi*f*L,
> >at 160MHz, 85nH has 6.28*1.6*10^5*85*10^-9=85.4513, or roughly 90 ohms of
> >reactance. To find the current distribution, we can apply 1V to the
> >paralleled Rs and Ls. The resulting current through Rs is 1v/100 ohms =
> >10mA. The current through Ls is 1v/90 ohms = 11.1mA.
> >The total current is 10mA plus 11.1mA = 21.1mA.
OOPS. Big mistake Rich.
I am surprised that Tom didn't catch that one. The currents DO NOT total
21.1 mA because they are not in phase---as Mario, S53A, pointed out.
Thanks, Mario.
However, the currents in Rs and Ls are not 90 degrees apart either. The L
in the Rs I use is around 18nH. Thus, the current in Rs lags the voltage
by roughly 20 degrees. Also, the intrinsic VHF R in Ls is not known.
Thus, the VHF current in Ls is slightly less than 90 degrees behind the
voltage. Another problem is that the intrinsic R in Ls, and the intrinsic
L in Rs causes slightly less current to flow in each element than my
current calculations indicated. All things considered, however, it seems
logical that similar amounts of VHF current flow in Ls and
Rs.............so there is still a hole in the opinion Tom Rauch has held
since 1990--i.e., that no significant VHF current flows through a VHF
parasitic-suppressor's inductor.
>
> You do not even understand the concept of current flow in a circuit with
> reactance. All you have shown is the Q is just over one. ...snip...
> The Q is just barely over one. How much lower can nichrome make the Q go?
> What is the lowest practical or possible Q?
=========================================
OK, Tom, let's do Q.
A pure inductance with a reactance of 2 ohms is in series with a pure
resistance of 10 ohms. What is the Q? // If the same inductance and
resistance are connected in parallel, what is the Q?
=========================================
> You have claimed, before the whole world,...
Give us a break, Tom. The whole world? With all of the stalling that's
going on, there might only be a few people tuned in. We still don't know
the name of Mr. XXX or whether or not your AL-1500 amplifier design
incorporates a VHF parasitic-oscillation suppressor. OTOH, the mysterious
Mr. Miklos, who has never spoken with me, yet who reportedly refuses to
speak with me on the telephone, perks up the plot a bit---even though it
does sound like something out of a daytime television soap-opera. Are we
supposed to pretend that Mr. Miklos can not hang up the telephone as need
be if he speaks to me?
snip
> So now it is Mark Wilson also.............
Alas, even though Mark Wilson installed a low VHF Q parasitic suppressor
retrofit kit is his once-squirrelly Alpha 77 amplifier, for some
indeterminate reason, on page 74 in the September, 1994 issue of QST, he
allowed Tom Rauch to say:
"The coil's reactance increases with frequency and at VHF most of the
signal path is through the resistor. It is plainly evident that the
dominant component at VHF is the resistor, not the coil."
I suspect that Mark's primary mistake was taking Paul Pagel out of the loop
for the Sept. 1994 Technical Correspondence column. If Mark had given Paul
the opportunity look this over, it seems pretty likely that Paul would have
spotted the problem BEFORE it went to the printer. Another problem that
might have been spotted was Fred Telewski's unenlightened statement about
choke power supply filters being okay for use in linear amplifiers.
Telewski's statement raised the eyebrows of all of the amplifier builders I
know.
The pronouncement that, at VHF, no significant current flows in "coils"
(inductors) is indeed specious. Your continued belief in this
pronouncement is an enigma to me, Tom.
----------------------------------------------
e-mail copy to Mr. Rauch, et al.
--
--Rich-- ag6k, 805-386-3734
Lawrence Stoskopf
Just discussed this supressor bit with an experienced Collins TX
engineer. Found it amusing. They had a 10 KW HF rig that had an inch
wide by three inch aluminum strap from the tube to the tuning capacitor.
Mounted on the strap were two noninductive resistors side by side
(parallel) mounted along the strap. Everyone looking at the amp
wondered why they shorted the resistors.
Apparently you build the amplifier, then figure out where the
parasitics are and do what it takes to get rid of them. Seems a bit
empiric for todays models, but doubtlessly effective. This is the limit
of my information on this.
N0UU
Phil Clements
One of the most inovative suppressors I have ever seen is a current
model made by Henry Radio for the 3CX3000A7. It is a 1 inch wide,
5.5 inch long hard-drawn copper strap bent into a 1 turn coil, 1 inch
in diameter. The strap is then coated with chrome plating (like car
bumpers used to be) until the resistance of the chrome reaches the
necessary R to meet the design specs. for its application. This
eliminates the need for big parallel resistors accross the coil,
nichrome wire, (or strap, in this case) and will certainly handle any
power on H.F. that the tube is capable of! My meager test equipment
(Autec RF-1 and R.S. DVM) show that the L is about .34 uh @ 28 mhz
and the R of the chrome is 25-35 ohms. I have one of these hanging on
my wall to remind me to strive for simplicity and inovation in my
design work. Nice going, Henry!! (((73))) Phil, K5PC
(or strap, in this case)
Wes Stewart
In article <57rtte$e...@dallas1.connect.net> phil...@connect.net (Phil Clements) writes:
>From: phil...@connect.net (Phil Clements)
>Subject: Re: Good source for amplifier tips
>Date: 1 Dec 1996 12:32:46 GMT
Hmmm. Sounds like a Henry 8K. It's been a while since I saw the inside of
one of these, but I doubt that the suppressor is chrome plated. I
think it's polished silver.
Also, your measured inductance seems awfully high. That's probably
about 15-20% of the tank inductance.
BTW, I was trying to help my friend re-engineer this thing to stop the corona
and arcing problems with the homemade relay contacts used to switch the fixed
tank capacitors. Some good old corona reduction techniques like smoothing
corners and increasing spacing did the trick. No nichrome was required.
N7WS
Phil Clements
the plating is chrome..I have been in the plant and seen the process.
BTW Henry has completely changed the solenoids on the 8K...too many were
having the same problem as you experienced. Talk to Paul Johnson for
details. Also, the high L in the suppressor may counteract the high
output C of the tube..the 8K uses no C1 in the pi-l on 10m because of
this.
Phil, K5PC
R. L. Measures
In article <57rtte$e...@dallas1.connect.net>, phil...@connect.net (Phil
Clements) wrote:
> In article <32A0F1...@tri.net>, stos...@tri.NET (Lawrence Stoskopf) says:
> >
> >Just discussed this supressor bit with an experienced Collins TX
> >engineer. Found it amusing. They had a 10 KW HF rig that had an inch
> >wide by three inch aluminum strap from the tube to the tuning capacitor.
> >Mounted on the strap were two noninductive resistors side by side
> >(parallel) mounted along the strap. Everyone looking at the amp
> >wondered why they shorted the resistors.
If the amplifier in question was a Collins 208-U10, at the
8171/4CX10,000D's anode resonance of around 60MHz, the "short" probably
exhibits at least 40 ohms of reactance.
> >Apparently you build the amplifier, then figure out where the
> >parasitics are and do what it takes to get rid of them. Seems a bit
> >empiric for todays models, but doubtlessly effective. This is the limit
> >of my information on this.
> >
> >N0UU
>
> One of the most inovative suppressors I have ever seen is a current
> model made by Henry Radio for the 3CX3000A7. It is a 1 inch wide,
> 5.5 inch long hard-drawn copper strap bent into a 1 turn coil, 1 inch
> in diameter. The strap is then coated with chrome plating (like car
> bumpers used to be) until the resistance of the chrome reaches the
> necessary R to meet the design specs. for its application. This
> eliminates the need for big parallel resistors accross the coil,
> nichrome wire, (or strap, in this case) and will certainly handle any
> power on H.F. that the tube is capable of! My meager test equipment
> (Autec RF-1 and R.S. DVM) show that the L is about .34 uh @ 28 mhz
> and the R of the chrome is 25-35 ohms. I have one of these hanging on
> my wall to remind me to strive for simplicity and inovation in my
> design work. Nice going, Henry!! (((73))) Phil, K5PC
Nice going, Henry, but not fast going by any means, Phil. For years after
the article about intermittent VHF parasitic oscillations was published in
the October 1988 issue of QST, Henry used silver-plating on its VHF
parasitic suppressors. I have been using a resistorless VHF parasitic
oscillation suppressor in my Class AB1 4CX3000A amplifier for 7 years. No
resistor is used---just two UNequal lengths of nickle-chromium-iron alloy
(a.k.a. stainless-steel) shim-stock which are NOT inductively coupled to
each other. At VHF, nickle-chromium alloys have enough intrinsic R to get
the job done. According to my dipmeter, the original silver-plated strap
exhibited a sharp dip at 68MHz. However the nickle-chromium replacement
exhibits a broader and shallower dip at 68MHz. In my mind, better
dampening means a duller dip. Apparently, some people think it does not
matter. // Any comments, Tom? Any comments Reid?
IMO, Henry's newer suppressors would provide even better VHF-dampening if
Henry provided a second VHF current path in the suppressor device with a
different amount of L. Two UNequal L, VHF current current paths creates a
stagger-tuning broadbanding effect like that in IF transformers when the
secondary and primary windings are tuned to different frequencies. Such a
suppressor can be made from entirely from strips of stainless steel
shimstock. However, to minimize VHF-Q, the width of the stainless-steel
strips should be no wider than is necessary to carry the RF circulating
current from the tank at 29MHz. IMO, a VHF parasitic suppressor that does
not get hot at 29MHz is not going to be dampening/absorbing very much
energy above 30MHz.
I still get phone calls from owners of older Henry amplifiers, that use
silver-plated "suppressors" who experience what appears to be VHF parasitic
arcing at the bandswitch. Jim Day, W6DF, who works on Henry amplifiers,
told me that he repeately explained the apparent, intermittent VHF
parasitic arcing problem, and its cure, to the powers that be at Henry
Radio. It took Henry at least 5 years to figure out that Jim Day knew what
he was talking about. ........ . Nice going, Mr. Day. Nice going, Mr.
Handy.
-------------------------------------
e-mail copies to Mr. Clements, Mr. Stoskopf, Mr. Rauch, et al.
--
--Rich-- ag6k, 805-386-3734
R. L. Measures
> In article <57rtte$e...@dallas1.connect.net> phil...@connect.net (Phil
Clements) writes:
> >From: phil...@connect.net (Phil Clements)
> >Subject: Re: Good source for amplifier tips
> >Date: 1 Dec 1996 12:32:46 GMT
>
> >In article <32A0F1...@tri.net>, stos...@tri.NET (Lawrence Stoskopf)
says:
> >>
>
>
>
> > One of the most inovative suppressors I have ever seen is a current
> > model made by Henry Radio for the 3CX3000A7. It is a 1 inch wide,
> > 5.5 inch long hard-drawn copper strap bent into a 1 turn coil, 1 inch
> > in diameter. The strap is then coated with chrome plating (like car
> > bumpers used to be) until the resistance of the chrome reaches the
> > necessary R to meet the design specs. for its application. This
> > eliminates the need for big parallel resistors accross the coil,
> > nichrome wire, (or strap, in this case) and will certainly handle any
> > power on H.F. that the tube is capable of! My meager test equipment
> > (Autec RF-1 and R.S. DVM) show that the L is about .34 uh @ 28 mhz
> > and the R of the chrome is 25-35 ohms. I have one of these hanging on
> > my wall to remind me to strive for simplicity and inovation in my
> > design work. Nice going, Henry!! (((73))) Phil, K5PC
>
>
> Hmmm. Sounds like a Henry 8K. It's been a while since I saw the inside of
> one of these, but I doubt that the suppressor is chrome plated. I
> think it's polished silver.
If the R measures (no pun intended) more than a few ohms at VHF, does it
seem likely that the conductive material is silver, Wes?
> Also, your measured inductance seems awfully high. That's probably
> about 15-20% of the tank inductance.
0.34uH/340nH does seem a bit on the high side.
> BTW, I was trying to help my friend re-engineer this thing to stop the corona
> and arcing problems with the homemade relay contacts used to switch the fixed
> tank capacitors. Some good old corona reduction techniques like smoothing
> corners and increasing spacing did the trick. No nichrome was required.
>
If there is no substanive amount of VHF energy present, VHF-lossy
nickle-chromium alloys serve no purpose. Was your friend trying to switch
the fixed C padders for the tune capacitor while he was transmitting, Wes?
------------------
e-mail copies to Mr. Stewart, et al.
.................... .................. ....................
................................... ...................................
................................... ...................................
................................... ...................................
................................... ...................................
................................... ...................................
................................... ...................................
................................... ...................................
................................... ...................................
................................... ...................................
................................... ...................................
................................... ...................................
................................... ...................................
................................... ...................................
................................... ...................................
................................... ...................................
................................... ...................................
................................... ...................................
................................... ...................................
................................... ...................................
...............
--
--Rich-- ag6k, 805-386-3734
w8j...@aol.com
>Thanks, Tom. BTW, what is a "physic ability".
> The reason I had to guess at the 160 figure is because somebody is
>withholding the actual anode resonant frequency.
>
>
Woops, no speel chek. I meant psychic ability, not the ability to give a
medicine that purges.
Please explain what we would know (with a grid dip meter measurement)
other than the reactance crosses zero at that point.
73 Tom
w8j...@aol.com
>> The point remains what can you possibly do with a grid dip meter Rich?
>
>I'm trying, I'm trying, but the guy with the AL-80 keeps stalling and
>stalling, and nothing happens. The nice thing about a dipmeter is that
it
>does not require a physical connection to the circuit under test--as is
the
>case with other types of test equipment.
I guess you are not familiar with modern test equipment Rich.
The anode circuit of the AL-80B I just measured has two resonances (other
than the ones normally due to HF tank components). One is adjustable over
a range of 63.34 to 22.63 MHz as the plate tuning cap is adjusted. It
moves down in frequency as the capacitor is meshed.
The second resonance is at 152.10 - 161.36 MHz. It moves UP in frequency
as the capacitor is meshed.
With the suppressor removed this resonance moves up to 192.16 MHz.
What can you develop from this data? I'm especially interested in what you
will tell us about Q or impedance, since I also measured those values. I
want to see how close your ESP is to a HP Impedance Analyzer. Also, what
frequency does the PA oscillate on when the suppressor is removed?
Do you want me to E-mail someone the complete data before you reply, so we
can see how close you come without rectangular impedance coordinates for
your calculations?? I can do this, since I already have the Q, impedance,
frequency of oscillation, and so on.
I'm still wondering what you can predict with a GD meter.
73 Tom
R. L. Measures
In article <19961202025...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> In article <measures-ya023180...@news.vcnet.com>,
> meas...@mail.vcnet.com (R. L. Measures ) writes:
>
> >> The point remains what can you possibly do with a grid dip meter Rich?
> >
> >I'm trying, I'm trying, but the guy with the AL-80 keeps stalling and
> >stalling, and nothing happens. The nice thing about a dipmeter is that
> it
> >does not require a physical connection to the circuit under test--as is
> the
> >case with other types of test equipment.
>
> I guess you are not familiar with modern test equipment Rich.
I guess not. I was under the impression that they had to be connected to
the circuit under test.
> The anode circuit of the AL-80B I just measured has two resonances (other
> than the ones normally due to HF tank components). One is adjustable over
> a range of 63.34 to 22.63 MHz as the plate tuning cap is adjusted. It
> moves down in frequency as the capacitor is meshed.
>
> The second resonance is at 152.10 - 161.36 MHz. It moves UP in frequency
> as the capacitor is meshed.
No help. The frequency sounds in the ballpark, however, this resonance
moves DOWN slightly as the tune capacitor is meshed, not UP.
> With the suppressor removed this resonance moves up to 192.16 MHz.
>
> What can you develop from this data? ....snip...
What I am attempting to calculate is how the VHF current distributes in
the suppressor example YOU gave me (85nH/100 ohms). If you (Ameritron) are
really one of the "recognized amplifier experts", you should have nothing
to fear.
Tom: I never saw an answer to the question Jesse asked you:
"Tom, are you absolutely sure it's impossible for a unstable amplifier
to start into oscillation intermittently from switching transcients?
73, Jesse, W6KKT"
---------------------------------------
e-mail copies to Tom Rauch, Jesse et al.
--
--Rich-- ag6k, 805-386-3734
w8j...@aol.com
Hi Rich,
Im think we should finish one thing before starting something else.
In article <measures-ya023180...@news.vcnet.com>,
meas...@mail.vcnet.com (R. L. Measures ) writes:
>No help. The frequency sounds in the ballpark, however, this resonance
>moves DOWN slightly as the tune capacitor is meshed, not UP.
No, the VHF resonance moves UP slightly as the capacitor is meshed. If you
know all the answers, why do you bother asking questions?
> What I am attempting to calculate is how the VHF current distributes in
>the suppressor example YOU gave me (85nH/100 ohms). If you (Ameritron)
are
>really one of the "recognized amplifier experts", you should have nothing
>to fear.
Fear? What has that got to do with parasitics? Does the tube get so afraid
it shakes?
73 Tom
R. L. Measures
In article <19961202144...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> Hi Rich,
>
> Im think we should finish one thing before starting something else.
>
> In article <measures-ya023180...@news.vcnet.com>,
> meas...@mail.vcnet.com (R. L. Measures ) writes:
>
> >No help. The frequency sounds in the ballpark, however, this resonance
> >moves DOWN slightly as the tune capacitor is meshed, not UP.
>
> No, the VHF resonance moves UP slightly as the capacitor is meshed. If you
> know all the answers, why do you bother asking questions?
Who told you that I know all the answers? Apparently you haven't read the
copy of the Rebuttal I mailed to you. .
> > What I am attempting to calculate is how the VHF current distributes in
> >the suppressor example YOU gave me (85nH/100 ohms). If you (Ameritron)
> are
> >really one of the "recognized amplifier experts", you should have nothing
> >to fear.
>
> Fear? What has that got to do with parasitics? Does the tube get so afraid
> it shakes?
>
> 73 Tom
---
Hello, Tom:
If you believe that you can prove me wrong, then you should not be afraid
to take your dipper and measure the dip-freq. at the AL-80's DC blocking
capacitor. The anode resonant frequency is highest when the tuning cap. is
at minimum C. The anode resonant frequency is slightly lower when the
tuning cap. is at maximum C. Knowing both frequencies would be helpful.
How are you coming on the Q problems, Tom? That second Q problem is a bit
tricky. Would you prefer to let me post the answers first, or would you
prefer that we simultaneously post the answers at 00:00 hours GMT, or
what??? We need to start digging into the Q issue that you brought up,
Tom.
Tom: Does the AL-1500 amplifier, which you designed, utilize a VHF
suppressor?
Did you answer the question that Jesse asked you:
"Tom, are you absolutely sure it's impossible for a unstable amplifier
to start into oscillation intermittently from switching transcients?"
---73, Jesse, W6KKT
This seems like a fair question, Tom. Roughly 95% of the parasitic
oscillations that I hear about occur when transient currents are present in
the anode circuit. .
===============================
e-mail copies to Mr. Rauch, et al.
--
--Rich-- ag6k, 805-386-3734
R. L. Measures
In article <19961201220...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> In article <measures-ya023180...@news.vcnet.com>,
> meas...@mail.vcnet.com (R. L. Measures ) writes:
>
> >Thanks, Tom. BTW, what is a "physic ability".
> > The reason I had to guess at the 160 figure is because somebody is
> >withholding the actual anode resonant frequency.
> >
> >
>
> Woops, no speel chek. I meant psychic ability, not the ability to give a
> medicine that purges.
Sorry, but I'm still in the dark, Tom. What is 'psychic'?
> Please explain what we would know (with a grid dip meter measurement)
> other than the reactance crosses zero at that point.
>
I need to find the resonant frequency of a circuit in the mid VHF-range,
whose resonating C is 5pF, but to find it without physicallyconnecting a
measurement device to the circuit under test. A dipmeter can do it.
Dipmeters are also useful for comparing the Q of two circuits that resonate
at the same frequency.
--------------------------------
e-mail copies to Mr. Rauch et al.
--
--Rich-- ag6k, 805-386-3734
w8j...@aol.com
hi Rich,
I've been trying to answer your questions, even though you seem to avoid
the ones I ask.
In article <measures-ya023180...@news.vcnet.com>,
meas...@mail.vcnet.com (R. L. Measures ) writes:
>Hello, Tom:
>If you believe that you can prove me wrong, then you should not be afraid
>to take your dipper and measure the dip-freq. at the AL-80's DC blocking
>capacitor. The anode resonant frequency is highest when the tuning cap.
is
>at minimum C. The anode resonant frequency is slightly lower when the
>tuning cap. is at maximum C. Knowing both frequencies would be helpful.
No, that's incorrect. The lower anode resonant frequency follows the
capacitor in the normal manner, but the higher resonant frequency varies
exactly as I posted.
The change of the higher frequency is very slight. You have the data you
asked for.
>How are you coming on the Q problems, Tom? That second Q problem is a
bit
>tricky. Would you prefer to let me post the answers first, or would you
>prefer that we simultaneously post the answers at 00:00 hours GMT, or
>what??? We need to start digging into the Q issue that you brought up,
Coming on what? Digging into what Q issue? Did you mean QST issue?
Let's stick to one thing at a time Rich, when finished we can move on. It
makes no sense to start three other topics when you don't finish the one
at hand.You have the AL-80B anode resonant frequencies. Are you going to
do something with them, or talk about something else?
73 Tom
w8j...@aol.com
Hi Rich,
We absolutely agree on a few things (like exciter power overshoot) but
some things don't make sense.
For example, you suggest someone place a VHF radio near an arcing PA, and
listen for VHF RF, and site that as proof a parasitic is taking place. Yet
any arc has broad bandwidth, and contains energy from dc to light
frequencies. While the arc can be detected at VHF, it can be detected at
VLF also. It can also be detected by ear, and by eye! I can hear it the
arc on an AM BCB portable radio also.
You seem to feel a grid dip meter can be used to measure impedance or Q.
Can you please explain how to do that with a grid dip meter? Can I use a
grid dip meter to measure the impedance of a feedline? Will it tell me if
a circuit is series resonant, or parallel resonant? Can I use a grid dip
meter to measure resistance?
You seem to think the presence of a "grid dip indication" (point where
reactance crosses zero) means a PA will oscillate at that frequency. Yet I
can think of many examples where the PA would actually be LESS prone to
oscillation at a resonant frequency. For example, if I made a series
resonant path to ground at a given frequency, the gain could go to zero at
that frequency. If that occured from the anode to chassis, why would the
PA oscillate on that frequency? If I build a TPTG oscillator and put a
series resonant circuit in the anode, it won't oscillate at the series
resonant frequency.
Finally, I can make a AL-80B oscillate at VHF by removing the suppressor
and increasing the HV. Why doesn't that cause the tuning capacitor to arc,
or the bandswitch to fail? Why don't the components you single out fail
when this parasitic condition is intentionally created? Why, if you are
correct that the feedthrough capacitance has something to do with this
oscillation, doesn't the cathode affect the frequency or level of the
parasitic oscillation? Why do only the anode and grid connections affect
the oscillation?
When a Penta-Labs Chinese 3-500Z was used in the oscillating PA, the glass
does not melt. Why not, since you determined that was the cause of glass
failure?
When I stopped the oscillation and biased the tube into 400 watts of
disipation, the glass failed in less than 15 minutes. Yet a spectrum
analyzer showed complete RF silence. Why did the glass melt with no
RFoscillation, and NOT melt with an oscillation, if the cause was a
parasitic as you claimed?
If you can explain any of this, it would make your theories much more
believable.
73 Tom
Gary Coffman
>meas...@mail.vcnet.com (R. L. Measures ) writes:
>> >I understand the confusion, Filip. There has been plenty of smoke and
>> >mirrors in this debate, and I suspect there will be more. Briefly---The
>> >roots of the debate date from 1926, when Mr. F. E. Handy said on page 72 in
>> >the *Radio Amateur's Handbook*:
>> >
>> > "The combination of both resistance and inductance is very effective in
>> >limiting parasitic oscillations to a negligible value of current."
>> >
>> > In other words, Mr. Handy found that VHF parasitic-oscillation
>> >suppressors performed measurably better when the suppressor's inductor (L)
>> >was made from resistance wire instead of copper wire.
>>
>> That does not appear to be what Mr. Handy was saying at all. The quote
>> indicates that a combination of an inductor and a resistor is very effective
>> as a parasitic suppressor.
>
>about the suppressor inductor itself For a grid parasitic-suppressor, F.
>E. Handy recommends a100 ohm resistor and 10 ohms per foot resistance wire
>to wind the suppressor inductor. Quoting "The grid chokes in the
>photograph are most effective of all as a coil of resistance wire is used."
Ah, then I see why Mr. Handy's advice was dropped from later Handbooks.
Just as you are doing, he fails to consider the suppressor as a unit,
and insists on treating its components as if they were operating in
isolation. That's an incorrect approach.
>> That advice is, of course, correct, and is the
>> way *every* parasitic suppressor is made today. Even yours. The resistance
>> needs to be in *parallel* with the reactance, however, not in series with
>> it, for best results.
>>
>This is the position of Mr. Rauch, who insists that insignificant VHF
>current flows through the suppressor L. Gary: Do you believe that
>insignificant VHF current flows through the suppressor L?
No, not at all. I believe that significant current flows through the
suppressor circuit. Some through the resistance, and some through the
reactance. If the two currents are approximately equal, then the
resistance and the reactance must also be approximately equal, and
hence the Q must be approximately *1*. Now where I grew up, *1* is
considered low Q.
You *cannot* treat the inductor and the resistor in isolation. They
are two parts of a *parallel circuit*. You *must* analyze the behavior
of the *circuit* as a unit to see what effects it may have.
>> Indeed, which is why there is a *paralleled* resistor across the inductor,
>> to lower the suppressor's *VHF Q*, or more to the point, to increase the
>> suppressor's *dissipation* at VHF. At *HF*, however, we want high Q so that
>> we don't have attenuation of the *desired* HF energy.
>
>I agree that we don't want HF attenuation. However, the additional loss
>created by a low VHF Q suppressor inductor is on the order of 1% at 29MHz
>and 0.2% at 14MHz. I found this acceptable because lower VHF Q suppressors
>eliminated occasional parasitic arcing in my SB-220 on 40m. If you feel
>that a 1% loss is unacceptable at 29MHz, then you should probably not be
>using lossy suppressor inductors. // Low VHF Q suppressors also eliminated
>the pecuilar double-peak in 40m output during tune-up. Other SB-220 owners
>have reported similar results in both of these areas.
If any of this is true, it is because the Q at *HF* has been lowered
by the lossy coil. It has *nothing* to do with the Q of the suppressor
at VHF. Calling this "low VHF Q" is extremely misleading.
> Of late, I have seen plenty of seemingly-questionable VHF Q theories
>proffered hereabouts. If you have access to a dipmeter, Gary, you may (or
>may not) be interested in comparing the dip between a VHF suppressor L made
>from copper wire, and a VHF suppressor L made from nichrome resistance
>wire. If you are curious, but you need a low VHF Q suppressor to make the
>comparison, let me know for which amplifier. I will build a low VHF Q
>suppressor and mail it to you--free. The 55-cent stamp is on me. This
>offer is good whether or not you report your findings on this newsgroup.
Rich, measure one of your *suppressors* with your dipmeter, not just the
coil, the whole damn thing. Now replace the nichrome wire with copper wire
and measure the *suppressor* again. If there is a significant difference
at VHF, I will be more than a little surprised.
Tom C. Brown, Jr.
w8j...@aol.com wrote:
[Partial]
> I'm still wondering what you can predict with a GD meter.
>
You can get a pretty good idea of where (frequency wise) the thing might
oscillate. Or at least where it might ring.
I thought Rich's purpose for the GDO measurements was quite obvious.
What do YOU usually use a grid dip oscillator for? It's a really great
tool for finding resonances. And, as you and I both know, if a circuit
is going to break into oscillation, it is liable to do it at (you
guessed it!) the frequency of resonance. And, since we are talking
about parasitic *oscillations* here (among other things), I'd say that
makes Rich's GDO measurements pretty significant.
I would have thought that a "recognized amplifier expert" would have
seen this right away......
73, Tom KJ5IE
w8j...@aol.com
RE: Grid Dip Meter
Hi Gary,
In article <1996Dec3.1...@ke4zv.atl.ga.us>, ga...@ke4zv.atl.ga.us
(Gary Coffman) writes:
>
>Rich, measure one of your *suppressors* with your dipmeter, not just the
>coil, the whole damn thing. Now replace the nichrome wire with copper
wire
>and measure the *suppressor* again. If there is a significant difference
>at VHF, I will be more than a little surprised.
This was my point in asking Richard what he intended to do with the grid
dip meter. It's the single worse "tool" (I certainly wouldn't call it an
instrument) for measuring a complex circuit's RF characteristics.
The dip meter doesn't even measure the reactance zero-crossing reliably,
it simply measures the suck-out of energy from the dip coil. Perhaps a
more correct description is the GDO measures energy absortion via the
coefficient of coupling between the dip meter and the circuit.
NONE of this has much to do with the behavior of the circuit at RF, only a
person with little comprehension of resonant systems would think it did. A
dip does NOT guarantee an oscillation, it very well might do just the
OPPOSITE! It will not tell us the Q or any other important parameter.
For example, the dip meter placed next to a resonant system with poor
(lossy) mutual coupling can show a shallow narrow dip, but if coupling is
increased the dip will get wider and broader. That's why the dip meter's
scale is not calibrated in ohms, Q, or rectangular impedance coordinates.
Secondly, the GD meter couples energy into the system (excites the system)
differently than the normal operating conditions do. You may have a crude
idea something is happening, but you have no idea what the system "looks
like" for normal excitation. All you know is the circuit has higher mutual
coupling at that frequency to some type of loss at that point in the
system. The higher the coupling to the loss, the more pronounced the dip.
My reason for asking what Rich intended to DO with the grid dip meter was
to see if he understood what the meter really measures. I wanted to see,
if as with the photons and 500 watt 8877 grids, he made another giant
unscientific leap of faith in the attempt to support his wild theories.
Gary, you are right on target with the Q issue. The Q of the suppressor
Rich outlined was unity. Adding nichrome lowers the VHF Q an insignificant
amount. If the suppressor became a perfect zero reactance resistance, the
Q would only drop from one to zero. Big deal.
I'm with you on this one Gary.
I consider a Q of one VERY low Q. The nichrome does reduce HF Q, and that
lowers the tendency for tank overshoot when the PA is overdriven or
underloaded. It also will help stabilize the PA at HF, but certainly NOT
at VHF.
I am very suspect of any person who represents himself as an expert on a
technique, and who so clearly fails to understand how to measure things
and how to analyze a simple parallel L/R circuit. This "lower VHF Q thing"
fits right in the same bin as the photons, 500 watt 8877 grids, VHF arcing
tuning caps and switchs, peak tank voltage not being able to exceed the
supply voltage, and so on.
Why did this stuff about nichrome apear in the early years? It's easy to
understand why.
Back then, tubes had long grid and anode leads. Grounded grid operation
was virtually unheard of, everything was grid driven and mostly used high
feedback triodes. Shielding was very poor, even metal chassis' were rare
in homebrew equipment.
Because of all this, PA's oscillated readily. The problem was uncontrolled
feedback below, at, and above the operating frequency. Any resistive
wiring helped stability, because the whole PA was laden with uncontrolled
and "unfixable" feedback paths. The typical PA had as much a tendency to
oscillate at the desired operating frequency as at other frequencies.
Distributed loss at and near the operating frequency is the only universal
cure when layout, components, and circuitry are that bad.
Moving such "cures" to a modern PA should be done only with the
understanding of what is actually being accomplished. Technology was very
different long ago.
73 Tom
R. L. Measures
In article <19961203131...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> hi Rich,
>
> I've been trying to answer your questions, even though you seem to avoid
> the ones I ask.
HI, Tom-------Here is the way I try to operate in a debate: When I read a
reply, at the first statement I come to that does not appear to wash, I
stop reading, and I deal with it. If time and space permit, I continue
reading until I come to another item that does not appear to wash, and I
deal with that. I know that if I do not do this, the debate can become
diffused and turn into a quagmire.
> In article <measures-ya023180...@news.vcnet.com>,
> meas...@mail.vcnet.com (R. L. Measures ) writes:
>
> >Hello, Tom:
> >snip... The anode resonant frequency is highest when the tuning cap.
> is at minimum C. The anode resonant frequency is slightly lower when the
> >tuning cap. is at maximum C. Knowing both frequencies would be helpful.
>
> No, that's incorrect. The lower anode resonant frequency follows the
> capacitor in the normal manner, but the higher resonant frequency varies
> exactly as I posted.
Exactly? Incorrect? I take "meshed" to mean that the tune cap. is at
maximum C. Since the tune C is effectively in series (through the chassis
connection) with the anode's C of 5pF, the net C of the two caps. should be
maximum when the tune C is at maximum. The net C should be at minimum when
the tune C is at minimum, or unmeshed. Since the net C is
parallel-resonant with the anode circuit's total L. (see
<http://www.vcnet.com/measures/D.a.13.GIF>), therefore, Mr. Rauch's
statement (that the resonance moves up in frequency when the tune capacitor
is meshed) sounds backwards.......
I will proceed using Tom's figure of 161MHz in the calculation. // The
problem is: how does the (161MHz) current divide in a paralleled 85nH Ls
and a 100ohm Rs in the AL-80B?......... Using XL=2*Pi*f*L ohms, the
reactance of the 85nH Ls = 86 ohms. Applying 1Vrms to the paralleled Rs
and Ls, the current throuh Rs is 1 volt divided by 100 ohms=10mA,
and the current through the Ls is 1 volt divided by 86 ohms=11.6mA.
Tom Rauch said: "The coil's reactance increases with frequency, and at
VHF most of the signal path is through the resistor. It is plainly evident
that the dominant component at VHF is the resistor, not the coil."
Tom: do you presently stand by this statement?
Tom: are you prepared to discuss Q?
------------------------------
e-mail copies to Mr. Rauch, et al.
--
--Rich-- ag6k, 805-386-3734
R. L. Measures
In article <19961203152...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> Hi Rich,
>
> We absolutely agree on a few things (like exciter power overshoot) but
> some things don't make sense.
Nah, not really, Tom. I have never seen 300+ watt overshoots.
> For example, you suggest someone place a VHF radio near an arcing PA, and
> listen for VHF RF, and site that as proof a parasitic is taking place.
This was an actual case. Someone phoned me and asked why a nearby VHF
field strength meter---about 4 feet away---pegged out whenever he heard/saw
the tune cap. in his amplifier arc? (the perf. cover to the final
compartment was removed so he could observe the arc)
snip.......
> You seem to feel a grid dip meter can be used to measure impedance or Q.
> Can you please explain how to do that with a grid dip meter? Can I use a
> grid dip meter to measure the impedance of a feedline? Will it tell me if
> a circuit is series resonant, or parallel resonant? Can I use a grid dip
> meter to measure resistance?
IMO, a dipmeter can NOT be used to measure impedance or absolute Q'.
A dipmeter is basicly used to measure the frequencies of resonances. Also,
relative Q can be discerned by comparing the dips of two circuits that
resonate on the same frequency. Higher Q is indicated by a sharper dip.
Less Q is indicated by a broader dip. (positioning the dipmeter the same
distance from either resonant circuit is essential). // IMO, a dipmeter
can not be used to find the resonance of a series-resonant circuit unless a
conductive path exists between the ends of the seriesed L and C. If such a
conduction path does not exist, no current can flow in the series-circuit,
and resonace can not be found.. However, a series-resonant circuit whose
ends are connected together through a conductor is a parallel-resonant
circuit.
> You seem to think the presence of a "grid dip indication" (point where
> reactance crosses zero) means a PA will oscillate at that frequency.
Is this something you think you saw in one of my QST articles, or in
e-mail, or from an impression you got during our telephone conversation?
snip......
> Finally, I can make a AL-80B oscillate at VHF by removing the suppressor
> and increasing the HV. Why doesn't that cause the tuning capacitor to arc,
> or the bandswitch to fail? ...snip...
Fortunately, we are about to get into that, Tom. Bob Scott, KF9YH
presently has an AL80 whose bandswitch contacts are arcing. I e-mailed
Bob, so he just might be showing up on this thread. // More good news: I
should be receiving a photocopy of the schematic diagram for the AL1500
8877-ampliier.via snailmail. This will answer the question I have about
whether or not the AL1500 uses a VHF parasitic-oscillation suppressor.
>
> When a Penta-Labs Chinese 3-500Z was used in the oscillating PA, the glass
> does not melt. Why not, since you determined that was the cause of glass
> failure?
Chinese tubes have been improved since they first came on the market,
although I have no specifics on the glass-recipe they are currently using.
The tubes that Mr. Raich had the melted glass problem with were
manufactured around1990.
> When I stopped the oscillation and biased the tube into 400 watts of
> disipation, the glass failed in less than 15 minutes. Yet a spectrum
> analyzer showed complete RF silence. Why did the glass melt with no
> RFoscillation, and NOT melt with an oscillation, if the cause was a
> parasitic as you claimed?
> If you can explain any of this, it would make your theories much more
> believable.
How long did the vhf oscillation last?
What was the frequency of oscillation, approximately?
What is the code-date on the tube that you forced to oscillate?
-------------------------------------------
copies to Mr. Rauch et al.
--
--Rich-- ag6k, 805-386-3734
w8j...@aol.com
In article <32A50D...@TECLink.net>, "Tom C. Brown, Jr."
<tbr...@TECLink.net> writes:
>
>[Partial]
>> I'm still wondering what you can predict with a GD meter.
>>
>
>You can get a pretty good idea of where (frequency wise) the thing might
>oscillate. Or at least where it might ring.
I guess the correctness of your statement depends heavily on what we
consider "a pretty good idea". The GDO is certainly better than looking at
the system with a conventional ohm meter, or a battery and a light bulb,
but it doesn't indicate anything definitive about ringing or oscillation.
The GDO may dip at a frequency where the PA has the lowest anode impedance
(the least gain) or the highest impedance (the most gain). Does your grid
dip meter measure impedance? Does it have a scale that indicates Q, R, or
X ? If not, it tells you almost nothing of major importance in a complex
system.
>I thought Rich's purpose for the GDO measurements was quite obvious.
Not to me. It's no more obvious than how a 25 watt rated grid can
dissipate 500 watts. Does anyone want to call Mr. Miklos about this or the
other claims like photons??
>What do YOU usually use a grid dip oscillator for?
I don't. I did in the 60's when I was building transmitters and receivers.
The last time I took mine off the shelf was to measure the the resonant
frequency of a parallel tuned circuit. In that case, I didn't need to know
the Q, R, or X. I just needed to see where it was resonant at. Even a
MFJ-259 will tell you much more than a grid meter, so why use a GDO in
general applications like this?
>It's a really great
>tool for finding resonances.
Well, kind of. I would agree if you left the word "great" out. Simple
would be OK, crude would be better. It is useful if we ONLY need to know
the resonant frequency of a simple circuit, like a parallel or series
lumped L/C circuit.
>And, as you and I both know, if a circuit
>is going to break into oscillation, it is liable to do it at (you
>guessed it!) the frequency of resonance.
You may know or think that, but I don't. The frequency where a circuit
oscillates is often much more complex than that.
Resonance (as measured by a dip meter) merely indicates the point where
maximum energy is removed from the dip meter's oscillator tank. It tells
us a little bit about resonance in a complex system, but it just as likely
may indicate the frequency where oscillation can NOT occur as the one
where it does. In something as complex as a PA, it would be pure blind
luck if the system oscillated at any particular "dip frequency".
>And, since we are talking
>about parasitic *oscillations* here (among other things), I'd say that
>makes Rich's GDO measurements pretty significant.
>I would have thought that a "recognized amplifier expert" would have
>seen this right away......
Tom, it's bad enough when one person turns this into a personal issue by
constantly throwing barbs and personal digs, and making up stories about
conversations and letters. I fully expect that from some people, because
they view technical disagreement as ego damage or something very personal.
That kind of behavior is childish and has no place in a technical
discussion. If you see me do it, correct me and I will quit doing it...
and I hope you will do the same.
Do you understand how a GDO actually works internally? Do you understand
what makes an oscillator oscillate? If you do, why do you think a dip on
the meter indicates a problematic oscillation may occur?
Imagine I place this circuit from the anode to ground.
anode-----www---] [----((((((((-----ground
res cap ind
With R a low value, this circuit will dip at the resonant frequency of the
L/C combo. In a TPTG oscillator, would you say the stage is more or less
likely to oscillate at that frequency?
What if I have a TPTG oscillator with the grid parallel resonant at 120
MHz and the anode parallel resonant at 190 MHz. Would you say the system
is more likely to oscillate near 120 or 190 MHz?
Suppose the VHF anode Q is one or two, do you think adding nichrome to the
suppressor inductor will lower the VHF Q substantially? If so, can you
explain how?
73 Tom
Tom C. Brown, Jr.
Tom C. Brown, Jr. wrote:
>
> w8j...@aol.com wrote:
>
> [Partial]
> > I'm still wondering what you can predict with a GD meter.
> >
>
> You can get a pretty good idea of where (frequency wise) the thing might
> oscillate. Or at least where it might ring.
>
>
> tool for finding resonances. And, as you and I both know, if a circuit
> is going to break into oscillation, it is liable to do it at (you
> about parasitic *oscillations* here (among other things), I'd say that
> makes Rich's GDO measurements pretty significant.
>
> I would have thought that a "recognized amplifier expert" would have
> seen this right away......
>
Plus:
Don't forget that Rich was not making exact, quantified measurements
with the GDO. He was (1) finding resonances in the circuit *WITHOUT*
attaching anything physically to the circuit and he was (2) able to
compare the Q in a "before and after" situation by just looking at the
response of the dip meter in each case. In other words, he was able to
say that the Q was greater or less after swapping the suppressors. If
*you* are unable to make such judgements with a GDO, then I guess you
really wouldn't have much use for one.....
In one case, he saw that the Q at one particular VHF resonant point was
visibly and substantially less after doing NOTHING but replacing the
suppressor.
Now, in THAT particular case, if the act of changing the suppressor did
not affect the Q at that VHF frequency, then what did? And do we agree
that a change in the Q at a VHF resonant point is indeed important? Or
do you believe the Q at the (or one of the) VHF resonant points in
unimportant when considering an amp's propensity to oscillate?
Have you ever been present during a catastrophic "parasitic event"? If
so, then you should know that such an event is not even close to the
same thing as a controlled oscillation in laboratory conditions. Gary's
comment about the RCA and Harris field guys is very appropriate here.
You're contradicting a lot of high powered folks when you catagorically
dismiss the parasitic problem.
Of course, you maintain that "the whole staff" at Eimac agrees with you,
but you have only identified *one* person at Eimac so far. Surely we
can all agree that one person does not make up "the whole staff".
Oh, and I asked before.....what is Mr. Brandon's official title at
Eimac?
Jumping back, for a second.....
Did you make the analysis with your HP network analyzer *without*
hooking *anything* physically to the circuit under test? If not, then
the data you collected concerning the resonant points is suspect.
Anytime you connect something conductive (like a probe) to a circuit and
start measuring resonant frequencies up in the range where "stray
reactances" can be a significant part of the whole picture, the results
may indeed be affected (possibly substantially) by the very connection
of the probe. I have seen this before, and believe me, it *is* a
concern.
Now, if you *did* make all your measurements with no physical connection
to the circuit, and without overcoupling anything in any way to the
circuit, then please disregard the above. Just a thought......
73, Tom KJ5IE
w8j...@aol.com
Rich,
We need to talk more about how PA's work, and how to measure important
parameters than all these unrelated Edisonian test methods. But first, I
wonder why Vectronics stopped using the Penta Chinese tubes if your
suppressors fixed the problem? I also wonder why they stoped using your
suppressors if they helped PA life?
In article <measures-ya023180...@news.vcnet.com>,
meas...@mail.vcnet.com (R. L. Measures ) writes:
>> For example, you suggest someone place a VHF radio near an arcing PA,
and
>> listen for VHF RF, and site that as proof a parasitic is taking place.
>
> This was an actual case. Someone phoned me and asked why a nearby VHF
>field strength meter---about 4 feet away---pegged out whenever he
heard/saw
>the tune cap. in his amplifier arc? (the perf. cover to the final
>compartment was removed so he could observe the arc)
All that test indicates is there was VHF energy present, which we all know
is present in any arc. For example, my two meter radio's S meter pins and
the squelch opens when my HF antenna arcs. The two meter antenna is over
100 feet from my HF antenna. Does that mean my HF antenna has parasitics?
Of course not.
If I turn on a wall switch, and hold my two meter handheld near the
switch, I hear it click. If I hold the switch almost closed, and make the
switch arc, I hear an S-9 plus noise on two meters. Does that mean my wall
switch needs nichrome suppressors to cure its arcing?
I hear lightning discharges and corona discharges on my two meter radio.
Does that mean lightning has parasitics?
>IMO, a dipmeter can NOT be used to measure impedance or absolute Q'.
>A dipmeter is basicly used to measure the frequencies of resonances.
<snip>
>a series-resonant circuit whose
>ends are connected together through a conductor is a parallel-resonant
>circuit.
At least we agree on this point. The dip meter tells us next to nothing
when it is used to measure a complex circuit. We have no idea if it is
indicating a series or parallel resonance as viewed by the tube element
being "dipped". We have no idea what the Q is, or if the impedance is high
or low at the tube element.
>> When a Penta-Labs Chinese 3-500Z was used in the oscillating PA, the
glass
>> does not melt. Why not, since you determined that was the cause of
glass
>> failure?
>
>Chinese tubes have been improved since they first came on the market,
>although I have no specifics on the glass-recipe they are currently
using.
>The tubes that Mr. Raich had the melted glass problem with were
>manufactured around1990.
This was an old tube, one remaining from a batch purchased from Penta
several months after they were introduced. Here's what I think you are
implying:
1.) If the glass would have melted during the parasitic, it would have
been the bad glass problem (like the kind that was never fixed when
Vectronics installed your kits).
2.) Since the glass melted and sucked a hole in at less than rated
dissipation after a short period of dc operation, it may be different
recipe glass.
Rich, it seems like everything NOT supporting your claims is questionable.
If it supports your claims it is data worth etching in stone.
The Q of the nichrome suppressor at VHF you gave as an example is only
about 1 at 100 MHz. The Q of the regular suppressor at VHF is a tiny bit
larger, 1.14. Do you think a Q reduction from 1.14 to a Q of one will
cure oscillations that melt glass? Do you think it will cure all
oscillations?
Nichrome lowers Q much more on 20 MHz than 200 MHz. Your "low VHF Q"
suppressors decrease the likelyhood of arcs and make tuning slightly
"smoother" because of the lower HF Q, and not because of any "improvement"
at VHF. That much can be proven.
Like Gary said, measure the suppressor with the resistor in place Rich.
Otherwise your test is a waste of time. Your suppressor will help some
problems, but NOT VHF oscillations.
73 Tom
Will, KN6DV
For the interesting party's again.
I have uploaded PART 8 of the on going discussion.
It is still interesting and also educational (for me).
Also you will find the letter from Eimac( to AG6K here) and a few other
links to amplifier related web pages.
Hope to be a help.
Thanks 73 Will, KN6DV
--
http://www.av.qnet.com/~kn6dv
REFORM
Unless the reformer can invent something which substisutes attractive
virtues for atractive vices, he will fail.
(Walter Lippmann)
w8j...@aol.com
>Incorrect? I take "meshed" to mean that the tune cap. is at
>maximum C. Since the tune C is effectively in series (through the
chassis
>connection) with the anode's C of 5pF, the net C of the two caps. should
be
>maximum when the tune C is at maximum. The net C should be at minimum
when
>the tune C is at minimum, or unmeshed. Since the net C is
>parallel-resonant with the anode circuit's total L. (see
><http://www.vcnet.com/measures/D.a.13.GIF>), therefore, Mr. Rauch's
>statement (that the resonance moves up in frequency when the tune
capacitor
>is meshed) sounds backwards......
You only think that because you have no idea what the plate tuning
capacitor looks like, impedance wise, at frequencies above 150 MHz. I
expect that, since you do not have access to more elaborate test
equipment.
When the capacitor (as with almost any capacitor) passes its' series
resonate point, it looks inductive. It does not have a high amount of
inductive reactance, but it has some. Above the series resonat frequency,
this inductive reactance increases as the plates are unmeshed. That is
because the shunt capacitance is far down the plate, and is such a low
impedance the series inductance is the dominating reactance.
This is a common effect in RF components. Large vacuum capacitors exhibit
the same effect (but up in the upper UHF range), as they are unmeshed they
become MORE inductive.
HF roller inductors exhibit a "mirror" effect. As inductance is lowered,
the VHF impedance null moves LOWER in frequency. This can be seen on a
network analyzer. That's why some rollers quit working at the upper end of
HF, and start arcing, overheating, or just won't "tune".
Most experienced RF designers are aware of these effects.
> I will proceed using Tom's figure of 161MHz in the calculation. // The
>problem is: how does the (161MHz) current divide in a paralleled 85nH Ls
>and a 100ohm Rs in the AL-80B?
Sorry Rich, wrong frequency. The AL80B I tested did not oscillate there.
It oscillated around 182 Mhz.
......... Using XL=2*Pi*f*L ohms, the
>reactance of the 85nH Ls = 86 ohms. Applying 1Vrms to the paralleled Rs
>and Ls, the current throuh Rs is 1 volt divided by 100 ohms=10mA,
>and the current through the Ls is 1 volt divided by 86 ohms=11.6mA.
That is for an inductior of infinite Q, and neglects stray coupling and
capaciatance. The AROUND (remember I plainly indicated ~85 nH) 85 nH
inductor, actually measures slightly higher than it calculates or measures
in free air when it is installed over the resistor.
This calculation is a wasted exercise when dealing with "standard"
components at VHF. The actual characteristics are different than the
textbook perfect components, but I will allow the use of perfect
components for arguements sake if you at least use the correct frequency.
> Tom Rauch said: "The coil's reactance increases with frequency, and at
>VHF most of the signal path is through the resistor. It is plainly
evident
>that the dominant component at VHF is the resistor, not the coil."
To put my statement in the context given, one of describing the merits of
your nichrome at VHF, let's calculate the Q (it measures 0.8 on my HP) at
180 MHz. Lets calculate the change in Q when nichrome is added. Tell me
what gauge nichrome you use and the length, or mail me some, and I'll
calculate or measure the 180 MHz Q.
>Tom: do you presently stand by this statement?
Absolutely. The dominant component controling Q is the resistor, not the
ESR of the inductor. Adding nichrome makes little difference in VHF Q. If
the parasitic suppressor I tested was replaced with a pure 100 ohms
resistance (no distributed inductance or suppressor coil) the q would be
zero.
Is a Q change from 0.8 (by actual measurement) to zero a large
improvement?? What about if the suppressor Q was one? Is a change from one
to zero a large change?
73, Tom
Wes Stewart
In article <32A662...@TECLink.net> "Tom C. Brown, Jr." <tbr...@TECLink.net> writes:
>From: "Tom C. Brown, Jr." <tbr...@TECLink.net>
>Subject: Re: Good source for amplifier tips
>Tom C. Brown, Jr. wrote:
[snip]
>Don't forget that Rich was not making exact, quantified measurements
>with the GDO. He was (1) finding resonances in the circuit *WITHOUT*
>attaching anything physically to the circuit and he was (2) able to
>compare the Q in a "before and after" situation by just looking at the
>response of the dip meter in each case. In other words, he was able to
>say that the Q was greater or less after swapping the suppressors. If
>*you* are unable to make such judgements with a GDO, then I guess you
>really wouldn't have much use for one.....
Please! I'm sure Rich's attempts at determining circuit parameters are
well-meaning, and a GDO has its place, but... This ain't one of them.
>In one case, he saw that the Q at one particular VHF resonant point was
>visibly and substantially less after doing NOTHING but replacing the
>suppressor.
This may be true, but it's not proved with a GDO measurement.
[snip]
>Of course, you maintain that "the whole staff" at Eimac agrees with you,
>but you have only identified *one* person at Eimac so far. Surely we
>can all agree that one person does not make up "the whole staff".
>Oh, and I asked before.....what is Mr. Brandon's official title at
>Eimac?
Spurious arguments. Wy bother?
>Jumping back, for a second.....
>Did you make the analysis with your HP network analyzer *without*
>hooking *anything* physically to the circuit under test? If not, then
>the data you collected concerning the resonant points is suspect.
>Anytime you connect something conductive (like a probe) to a circuit and
>start measuring resonant frequencies up in the range where "stray
>reactances" can be a significant part of the whole picture, the results
>may indeed be affected (possibly substantially) by the very connection
>of the probe. I have seen this before, and believe me, it *is* a
>concern.
Clearly, you have no concept of modern network measurement techniques and I'm
not about to give you a free lesson. I recently retired after 30 years with a
major aerospace firm where I was engaged for most of that time in making r-f
and microwave measurements and doing failure analysis on everything
from components to all-up missiles.
I can assure you, that we "hooked something" to every damn one of these
devices in order to characterize it electrically. Tom has provided data in the
form of |S21|, ie, the magnitude of the forward transmission component. In
layman's terms, this is the "gain" of the device, expressed in dB, vs.
frequency. Viewed graphically, you would (he did) see the peaks and nulls of
transmission through the tube.
A series resonant grid will be a low Z impedance to ground and express a
minimum "gain" (high loss) through the path from cathode to plate. Conversely, a
parallel resonant situation will do the opposite.
I'm not hereby either arguing for or against anybody's opinions on suppresser
performance, indeed I'm formulating my own, but I will certainly defend Tom's
measurement technique against your criticism
>Now, if you *did* make all your measurements with no physical connection
>to the circuit, and without overcoupling anything in any way to the
>circuit, then please disregard the above. Just a thought......
>73, Tom KJ5IE
Wes N7WS
Tom C. Brown, Jr.
w8j...@aol.com wrote:
>
> hi Rich,
>
> I've been trying to answer your questions, even though you seem to avoid
> the ones I ask.
>
You mean like the way you have avoided questions like (1) Who is Mr. XXX
at Eimac? and (2) Who are all those folks at Eimac who agree with you on
the parasitics issue?
I know you are sick of hearing those questions asked, Tom, but don't
forget, YOU brought both issues up in this thread. I didn't. Rich
didn't. We just asked for real information once YOU brought it all up.
And as soon as we asked for real, verifiable information, you dropped
the subject and have pleaded with us to do the same.
Actually, in my opinion, your (1) inability (or flat refusal) to answer
either question, and (2) your intense desire to drop both subjects *IS*
an answer to both questions. Quite a revealing answer, I would say.
73, Tom KJ5IE
R. L. Measures
In article <19961205050...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> In article <measures-ya023180...@news.vcnet.com>,
> meas...@mail.vcnet.com (R. L. Measures ) writes:
snip...
> therefore, Mr. Rauch's
> >statement (that the resonance moves up in frequency when the tune
> capacitor
> >is meshed) sounds backwards......
>
> You only think that because you have no idea what the plate tuning
> capacitor looks like, impedance wise, at frequencies above 150 MHz. I
> expect that, since you do not have access to more elaborate test
> equipment.
>
Dipmeters find resonances, Mr. Rauch.
...snip...
> ......... Using XL=2*Pi*f*L ohms, the
> >reactance of the 85nH Ls = 86 ohms. Applying 1Vrms to the paralleled Rs
> >and Ls, the current throuh Rs is 1 volt divided by 100 ohms=10mA,
> >and the current through the Ls is 1 volt divided by 86 ohms=11.6mA.
Oops. It looks like Mr. Rauch's statement that insignificant current flows
in Ls is in deep feces---so here it comes........
> That is for an inductior of infinite Q, and neglects stray coupling and
> capaciatance. The AROUND (remember I plainly indicated ~85 nH) 85 nH
> inductor, actually......snip..........
> This calculation is a wasted exercise.........
....snip,,,,,,,,,,,on and on. ........Cough, cough. Bzzzzzzzzzzzt At this
rate my smoke detector is going to need a new battery by December.
Mr. Rauch proposed that I calculate the currents for an Ls of 85nH, a
parallel Rs of 100 ohms, at a frequency of 161MHz. I complied. Now it
seems that the apparently correct answers are NOT correct.
--------------------------------------
e-mail copies to Mr. Rauch, et al.
--
--Rich-- ag6k, 805-386-3734
w8j...@aol.com
Hi Wes,
Without drawings it's pretty difficult to explain how to measure a PA. But
factually, even if someone "see's a real sharp dip", it means nothing at
all in operation of the PA. It certainly does not indicate the likelihood
of oscillation.
The whole exercise is just another way to support claims that aren't true.
In article <n7ws.208...@azstarnet.com>, n7...@azstarnet.com (Wes
Stewart) writes:
Tom Brown:
>>Of course, you maintain that "the whole staff" at Eimac agrees with you,
>>but you have only identified *one* person at Eimac so far. Surely we
>>can all agree that one person does not make up "the whole staff".
>>Oh, and I asked before.....what is Mr. Brandon's official title at
>>Eimac?
>
>Spurious arguments. Wy bother?
Let me say why we should bother Wes. Rich's intentions may be good, but he
is absolutely giving false information about how a tube operates, and how
it fails. He has taken a situation that was found in one UHF high power
triode, and applied it to other tube types. His descriptions of failures
and the reasons for failures is absolutely incorrect.
The delemma is this, in order to prevent engineers from wasting time with
unimportant calls, or an employee from providing incorrect information,
most companies have a screening process. Engineers channel information
through one person, and that person disperses the information to a
customer service enginner or technician. At Varian (now CPI), the person
in contact with customers is Reid Brandon. The actual engineers are only
available to other engineers who have very specific technical questions
about a tube, or operation of a tube.
Rich refuses to consider Reid Brandon a good source, even though he is the
interface between engineering and the customer base. I did offer to let
any two people call a design engineer that is director of a group of tube
design engineers, a person who was head of Varian's Power Grid tube R and
D group. I also offered John Button, presently a manager at Varian, as
additional confirmation.
The reason for this is Rich has made some claims that, ludicrious as they
are, can cause people to ruin expensive tubes. These claims also causes
people toi falsely believe that parasitics are a major source of PA and PA
tube damage, and divert people from finding or understanding the real
causes and cures of problems. This hurts technology, and costs people time
and money. It can even cost lives.
W2GO was killed installing an "improvement kit" in a Ten Tec Titan. Al had
repeated 3CX800 failures, probably do to the fact Ten Tec did NOT use an
electronic overload grifd protection system in that PA. Had Al been given
accurate information, his repetitious tube failures might have been cured
and he might still be alive. Of course, he might have been killed adding a
grid protection circuit also, but at least he would have been doing
something to actually cure the problem.
I think it is important to get accurate information out, because there is
no reason to put people in harm's way doing needless modifications.
Especially when some of the modifications can actually reduce the life of
expensive tubes.
Rich claims he is Varian's messanger. Yet all of the people I deal with on
a regular basis think Rich's theories are the worst thing that has
happened to PA technology to date. Buzz Miklos, a tube design engineer in
charge of a staff of 15 design engineers, has offered to donate his time
to explaining how a tube fails, and where the parasitic story came from.
But he can not do it on the net.
Apparently no one is interested in the truth, since it is not as exciting
or sinister as some tall stories are.
Tom B wrote:
>>Did you make the analysis with your HP network analyzer *without*
>>hooking *anything* physically to the circuit under test? If not, then
>>the data you collected concerning the resonant points is suspect.
Tom, the data is not suspect at all. It is the only accurate way to
measure the grid's role in shielding. The primary error it is subject to
is the tube being cold. This is the one of the very same tests Eimac uses
to measure the RF characteristics of tubes.
The beauty of a network analyzer is the anode and cathode can be loaded
with the same impedances presented during operation, and the various
parameters displayed accurately. The instrument measures Q and other
parameters with an accuracy of a fraction of a percent, and loads the
circuit with as much loading as desired, from a few ohms to hundreds of
thousands of ohms.
It even can allows placement of dc bias voltages and currents on the
components while testing. A grid dip meter is the poorest method available
to test any of the parameters Rich was claiming to measure, that fact is
indisputable.
Exactly. As I've said all along, with a GDO we have no idea what the
systems impedance or Q is. The user has no idea if the dip indicates a
point of enhanced stability, or reduced stability. There is no indication
of impedance or Q. A GDO is one small step above a battery and buzzer.
Rich avoids technical questions about Q. Very plainly the conventional
parasitic suppressor examples he gave have a Q of about one. Adding
nichrome only reduces the Q a few percent, yet Richard claims it cures all
oscillation problems. While he accuses me of making smoke, he sidesteps
the issue that the parasitic oscillation is at 180 Mhz in the PA I tested.
It's tough to resove something when one of the major players responds to
facts with accusations.
73 Tom
Will, KN6DV
I have made a change on my web page about the "Good source for amplifier
tips" some body gave me a tip to ad the date's to the archive( Thanks
Rich).
This makes it easier to lookup something by date.
Please let me know if there is any other web page with amplifier related
item's so I can make a link to there also.
R. L. Measures
In article <19961205040...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> Rich,
>
> We need to talk more about how PA's work,
.snip......
OK, Tom. Let's do it.
Path:
news.qnet.com!news.he.net!www.nntp.primenet.com!nntp.primenet.com!cs.utexas.
edu!howland.erols.net!portc02.blue.aol.com!audrey01.news.aol.com!not-for-mai
l
From: w8j...@aol.com
Newsgroups: rec.radio.amateur.homebrew
Subject: Re: Good source for amplifier tips
Date: 28 Nov 1996 02:25:20 GMT
snip...
>Hi Rich,
>... ... Could you look this over an comment. It
>seems to indicate you do not understand Q. Could you tell me if I did
>something wrong?
snip
Ok, Tom.. Sorry for the delay, but there were some other issues that didn't
appear to wash, which I dealt with first.
>snip,,,
With 70 nH and 100 ohms:
Q = 2.3
...snip...
........Does Q = 2.3?
This is a parallel circuit, so most things are inverted. (Standing on head
may be useful.) Instead of calculating reactance (X), as would be the case
for a series circuit, we must instead calculate susceptance (B). . In a
parallel circuit, we graph B (with the + and - signs inverted) and
conductance (G), (G=1/R), to calculate admittance (Y) instead of graphing R
and X to calculate impedance (Y). R is in ohms and G is in siemens.
(archaic: mhos) . Using Y=1/Z and Z=1/Y, we can convert a parallel circuit
to a series equivalent circuit, or we can convert a series circuit to a
parallel equivalent circuit.
In order to be able to use the formula Q=X/R, we must first convert the
parallel 70nH/50 ohm circuit (Y) to an equivalent series circuit (Z) by
using Z=1/Y. To do this, most everything inverts. BL @100MHz, is
-1/(2*Pi*100^6 Hz*70^-9 H) = -22.7 milli-siemens. The G of the 50 ohm R is
1/50 = 20 milli-siemens.
To graph this, the -22.7 milli-siemen BL vector heads down and the 10
milli-siemen G vector heads to the right. The resulting vector forms a
right triangle whose tangent is -22.7/10 = -2.27, = -66.22 degrees. The
hypotenuse, Y, can be solved with the square root of the sum of the
squares. Thus, Y = 24.8 siemens@-66.22 degrees
Using Z=1/Y, , we convert this to an equivalent series circuit whose Z
vector can be resolved into its X and R components---the very things we
need to solve Q=X/R. Solving for Z = 1/(24.8 millisiemens @-66.22
degrees), Z = 40.3 ohms @+66.22 degrees. The X vector is sin 66.22 degrees
(0.915) * 40.3 ohms = 36.9 ohms. The R vector is cos 66.22 degrees (0.403)
* 40.3 ohms = 16.2 ohms.
Finally, using Q=X/R, Q = 36.9/16.2 = 2.28. Thus, your answer of Q = 2.3
looks good to me, Tom.
I checked another one of your Q problem's answers, and it looks good too.
Tom: Why did you recently cancel your 28 November post?
------------------------------------------------------
e-mail copy to Mr. Rauch, et al.
--
--Rich-- ag6k, 805-386-3734
Tom C. Brown, Jr.
Wes Stewart wrote:
>
> In article <32A662...@TECLink.net> "Tom C. Brown, Jr." <tbr...@TECLink.net> writes:
> >From: "Tom C. Brown, Jr." <tbr...@TECLink.net>
>
> >Tom C. Brown, Jr. wrote:
> [snip]
>
> >Don't forget that Rich was not making exact, quantified measurements
> >with the GDO. He was (1) finding resonances in the circuit *WITHOUT*
> >attaching anything physically to the circuit and he was (2) able to
> >compare the Q in a "before and after" situation by just looking at the
> >response of the dip meter in each case. In other words, he was able to
> >say that the Q was greater or less after swapping the suppressors. If
> >*you* are unable to make such judgements with a GDO, then I guess you
> >really wouldn't have much use for one.....
>
> Please! I'm sure Rich's attempts at determining circuit parameters are
> well-meaning, and a GDO has its place, but... This ain't one of them.
I disagree, Wes. Finding resonances and potential frequencies of
oscillation are applications where a GDO excels. It did years ago, and
it still does. If Rich's GDO is not indicating the resonant frequencies
in the circuit, then please tell us what it IS indicating?
If the circuit is NOT most likely to oscillate at one of its resonant
points, then please tell us where it WOULD be most likely to oscillate.
And please include an explanation of just why the circuit would be more
liable to oscillate at some place where it is not resonant. You're
contradicting some pretty well known facts and principles there, Wes.
>
> >In one case, he saw that the Q at one particular VHF resonant point was
> >visibly and substantially less after doing NOTHING but replacing the
> >suppressor.
>
> This may be true, but it's not proved with a GDO measurement.
Why not? If the GDO's dip was deep and narrow before replacing the
suppressors, and it is broad and shallow after replacing the
suppressors, then exactly what information do you have that makes you
feel that the GDO is lying to us? I just don't see where you are coming
from on that one, Wes.
>
> [snip]
>
> >Of course, you maintain that "the whole staff" at Eimac agrees with you,
> >but you have only identified *one* person at Eimac so far. Surely we
> >can all agree that one person does not make up "the whole staff".
>
> >Oh, and I asked before.....what is Mr. Brandon's official title at
> >Eimac?
>
> Spurious arguments. Wy bother?
*NOT* spurious arguments. Not by a long shot. And the reason "to
bother" is that Mr. Brandon has been one of Rich's critics, and Tom
seems to put a lot more stock in Mr. Brandon's opinion than he does in
Mr. Foote's letter stating his opinion. We know Mr. Foote's title,
which is the only real indication of his credentials we have. However,
we do not know Mr. Brandon's official title. Therefore we currently
have no information about his qualifications to say that Mr. Foote is
all wet, beyond the fact that he works at Eimac. Well, so did Mr.
Foote. Therefore, logically, we have no reason whatsoever to think Mr.
Brandon knows more about what he is talking about than Mr. Foote did.
Unless, of course, we knew just what Mr. Brandon's lot at Eimac is.
Which is why I asked the question.
>
[Snip]
>
> Clearly, you have no concept of modern network measurement techniques and I'm
> not about to give you a free lesson.
You are starting to sound more and more like Tom. When you start
telling other folks what they do and don't have an understanding of, you
are (to borrow a phrase from Paul Harvey) telling us more than you
know. Not usually a wise practice.
> I recently retired after 30 years with a
> major aerospace firm where I was engaged for most of that time in making r-f
> and microwave measurements and doing failure analysis on everything
> from components to all-up missiles.
>
> I can assure you, that we "hooked something" to every damn one of these
> devices in order to characterize it electrically. Tom has provided data in the
> form of |S21|, ie, the magnitude of the forward transmission component. In
> layman's terms, this is the "gain" of the device, expressed in dB, vs.
> frequency. Viewed graphically, you would (he did) see the peaks and nulls of
> transmission through the tube.
Wes, you have taken my comment that was specifically focused on *this*
case (assessing the resonant frequency and relative Q of the PA cicuits)
and applied it to test and measurement in general. That is not what I
said or what I meant. The fact is that there are some cases with some
circuits where attaching virtually anything conductive to the circuit
will compromise the measurements.
In my opinion, a GDO with no physical connnection, and with care not to
over couple the GDO into the circuit, is one of the least invasive ways
to measure the resonant frequency of the circuit IN THIS CASE.
What kind of network analyzer would be less invasive than that? You are
the expert on modern network analysis (you told me so). Please
enlighten us.
>
[snip]
Would the analyzer that Tom used present an infinite impedance to the
circuit under test AT THE POINT OF CONNECTION?
If yes, then I have no problem with his measurment technique. If not,
then his network analyzer IS affecting the data. If you don't know,
then you are (again) telling us more than you know. (still not a wise
practice)
>
> >Now, if you *did* make all your measurements with no physical connection
> >to the circuit, and without overcoupling anything in any way to the
> >circuit, then please disregard the above. Just a thought......
>
> >73, Tom KJ5IE
>
> Wes N7WS
Tom KJ5IE
w8j...@aol.com
>
>Mr. Rauch proposed that I calculate the currents for an Ls of 85nH, a
>parallel Rs of 100 ohms, at a frequency of 161MHz. I complied. Now it
>seems that the apparently correct answers are NOT correct.
You must like to distort facts Richard.
I never suggested 161 MHz, you picked it. It really isn't my fault the PA
oscillates higher in frequency, is it?? Instead, you right away go into
the well known RM attack mode, and once more claim anything disagreeing
with you must be false.
The whole exercise is a waste of time. You are trying to focus on one
statement I made, that the primary path for VHF currents is through the
resistor, and the primary HF path through the inductor. I'm afraid that
statement was too much of a generalization, but even as such it does not
mean you theory about VHF Q is correct.
The Q of the suppressor in question is about one at that frequency. Adding
the magical nichrome to the inductor lowers the Q and insignificant
amount, especially at the frequency where the PA really oscillates. Why do
you avoid this fact?
Nichrome makes an insignificant change in Q of the suppressor at the
frequency of oscillation.
In article <32A701...@mail.teclink.net>, "Tom C. Brown, Jr."
<mad...@mail.teclink.net> writes:
>I know you are sick of hearing those questions asked, Tom, but don't
>forget, YOU brought both issues up in this thread. I didn't. Rich
>didn't. We just asked for real information once YOU brought it all up.
>And as soon as we asked for real, verifiable information, you dropped
>the subject and have pleaded with us to do the same.
>
>Actually, in my opinion, your (1) inability (or flat refusal) to answer
>either question, and (2) your intense desire to drop both subjects *IS*
>an answer to both questions. Quite a revealing answer, I would say.
>
>73, Tom KJ5IE
Hi Tom,
Say what you like OM. You must have problems receiving messages, or you
don't read everything. Let me repost this one for you.
<19961127002...@ladder01.news.aol.com>, w8j...@aol.com writes:
>The former *R+D engineering manager*, Buzz Miklos, of the Salt Lake City
>Varian plant has agreed to speak to one or two people on the telephone
>about your theories. He said he will also come on line, but since he just
>took a position as engineering manager with another tube manufacturer in
>PA, it will be sometime in January before he is able to connect directly
>to this newsgroup.
>
>But if you want a quick answer, let's find someone to talk to Mr. Miklos.
>A second person still at Varian Salt Lake, John Button, will confirm Mr.
>Miklos' employment status with Varian, and the fact that Mr. Miklos was
>indeed R+D engineering manager for Varian and a long term employee in
>power grid tube design and research. Mr. Brandon will also conform Mr.
>Miklos' qualifications. Incidentally, both Mr. Button and Reid disagree
>with you completely also.
>
>Mr. Miklos feels, and says his Eimac engineering staff felt, your
theories
>have no merit and display a complete misunderstanding of how a power grid
>tube operates and the mechanisms of tube failures.
>
>He prefers you not receive his phone number, since he understands you
have
>more time to argue than he does. He will, however, call you if you always
>promise to include his views whenever you mention the person assigned to
>answer customer letters, Mr. Foote.
>
>You wanted to know who at Varian disagrees with you. You now have what
you
>have asked for...and perhaps more than you wanted.
>
>Mr. Miklos said my interpretation of grid damage was correct, that the
>8877 can only dissipate perhaps two times the rated dissipation (50
watts)
>before grid damage occurs. He further stated a major cause of gold
>sputtering is poor alignment between the grid and cathode bands in the
>tube, or grid dissipations above the 35-50 watt range.
>
>He also says, as I did, gold migrates slowly on its' own accord from
>normal electron bombardment. Excessive grid current speeds this process.
>Excessive dissipation makes the gold migration fatal to the cathode in a
>very short time.
>
>Let's pick someone to call Mr. Miklos. He will confirm all of this, and
as
>much more as needed. Or we can drop all this parasitic nonsense and do
>better things.
As we can plainly see, I did post three names of Varian engineers who
totally disagree with Richard (Reid Brandon, Buzz Miklos, and John
Button). Of the three, Miklos is the most technically inclined, since he
actually designs tubes and oversaw the entire staff in R and D at Varian.
He presently heads a group of about 15 engineers at another tube
manufacturer.
Contrary to your claim, there were names posted and an offer to repute
Measure's claims. Also contrary to your claim, it was explained what
Reid's position is. Rather than taking up room, you can call CPI and ask
THEM what Reid's status is. Let's not make a lot of noise over nothing
Tom.
Many people wrote QST and/or commented negatively on Richard's claims to
QST. According to QST, they had a stack of letters an inch high from
people disagreeing with Measures. Not satisfied with that, QST sent
Richard's work out to at least half a dozen experts in the industry...***
these were NOT the people listed in the QST's technical retraction. *** As
I understood, QST could not find a single person supporting Richard
technically although they looked for months, but that's just hearsay. But
at Rich has you, hi.
Ex-varian employees and present varian employees like Dave Chazen, Bill
Orr, and Jim Aurand have commented strongly against the parasitic
theories, as did industry experts like Warren Bruene and RF Design's
editor Gary Breed. Of course ETO's Dick Erhorn, Collin's Warren Bruene,
and others not named disagreed also.
Enough of that, if you think it's so important, call Buzz, Reid, John, or
someone and check it out yourself. I told you that a month ago. On to
technical issues....
I indicated why a network analyzer is much better, and why the GDO is
almost useless. Would you like me to repeat the reasons? Would you prefer
to talk about the reasons given? For example I mentioned that the GDO does
NOT indicate impedance or feedback in a system. You seem to not receive
answers that are clearly posted, are you having E-mail trouble?
73, Tom
k1...@aol.com
Tom States: "In my opinion, a GDO with no physical connnection, and with
care not to over couple the GDO into the circuit, is one of the least
invasive ways
to measure the resonant frequency of the circuit IN THIS CASE."
I may be stepping into something that is out of my league, but it seems to
me that the "physical connection" argument smacks of religeous fantasy.
There was a time when virtually all significant RF coupling was done
inductively--and even today's most advanced solid state power amplifiers
transfer huge amounts of power in this manner. If I were to connect you
to a powerful 160 MHz source, Tom, would you rather it be through a .5 pF
capacitor with a 10 Meg resistor in series--or through a large inductive
link with a high co-efficient of coupling. Your choice!
The fact is, with a grid dipper, I don't think I'd have a clue how tightly
I am coupling to the DUT--save by assessing the shapness of the dip
itself. And, any time the GDO is seeing something, it is also loading
something--or the indication wouldn't be happening. I may be stupid, but
I guess I just can't quite grasp where you're coming from!
Rick K1BQT
R. L. Measures
In article <19961206014...@ladder01.news.aol.com>, w8j...@aol.com
wrote:
> Hi Wes,
>
> Without drawings it's pretty difficult to explain how to measure a PA. But
> factually, even if someone "see's a real sharp dip", it means nothing at
> all in operation of the PA. It certainly does not indicate the likelihood
> of oscillation.
I have tested anode circuit VHF resonances that were so sharp, they
sucked the dipmeter out of oscillation. To me, that indicates more than
"nothing at all".
snip
> Rich's intentions may be good, but he
> is absolutely giving false information about how a tube operates, and how
> it fails. He has taken a situation that was found in one UHF high power
> triode, and applied it to other tube types. His descriptions of failures
> and the reasons for failures is absolutely incorrect.
Has Tom ever looked at a gold-sputtered grid through a microscope? It's a
wondrous sight, what with the cathode sprinkled with a hundreds of tiny
gold melt-balls, cooled blisters of once-boiling gold on the grid, and the
grid's base metal stripped of gold plating.
> The delemma is this, in order to prevent engineers from wasting time with
> unimportant calls, snip
Translation: Unimportant Calls: Owners of amplifiers whose bandswitch
contacts mysteriously burned up, phoning amplifier manufacturer.
> Rich refuses to consider Reid Brandon a good source,
I don't consider a person's title. I consider what a person says. So
far, we have not seen a letter from either Mr. Brandon or from Mr. Button.
So far, neither of these gentlemen has shown up on this newsgroup.
> The reason for this is Rich has made some claims that, ludicrious as they
> are, can cause people to ruin expensive tubes. These claims also causes
> people toi falsely believe that parasitics are a major source of PA and PA
> tube damage, and divert people from finding or understanding the real
> causes and cures of problems. This hurts technology, and costs people time
> and money. It can even cost lives.
>
> W2GO was killed installing an "improvement kit" in a Ten Tec Titan. Al had
repeated 3CX800 failures, ......
snip,
Assuming it was an AG6K low vhf Q parasitic suppressor retrofit kit, the
instructions warns customers:--- "Amplifiers that are plugged into the
electric-mains contain lethal voltages, whether they are switched on or
switched off! If you don't pull the electric-mains plug before working on
your amplifier, your amplifier may be for sale at a Śsilent keyą estate
sale!! "
A. Prose Walker, W4BW, had similar problems with his Titan. Ten Tec was
not able to resolve the problems, so Prose contacted me. He described the
problems. I told him such problems might be caused by an intermittent
oscillation condition. Prose installed low VHF-Q suppressor retrofit kit
s/n1451 in Sept. 1990. After a while, Prose contacted me and reported that
the previous problems had not yet reccured. Prose wrote to Ten Tec's
President, Jack Burchfield and told him of the apparent solution to the
problems which Ten Tec had not been able to resolve. Mr. Burchfield
replied and said that even though he had read Mr. Measures' articles in QST
magazine with "great interest", when low Q suppressors were tried, they
became "very hot" during 10m operation. Apparently, Jack did not read the
instructions to adjust the inductance of Ls to obtain moderate heating on
10m. IMO, if a VHF suppressor doesn't get hot at 29MHz, it is not going to
work optimally above 30MHz.
I worked on a Titan in which both 3CX800A7 tubes exhibited gold-sputter
leakage after about 25 hours of operation. The operator had been a ham for
over 20 years, and he knew how to tune up a grounded-grid amplifier. The
glitch that occured at the time the tubes were apparently damaged, did not
occur during tuneup.
snip
> The beauty of a network analyzer is the anode and cathode can be loaded
> with the same impedances presented during operation, and the various
> parameters displayed accurately. The instrument measures Q and other
> parameters with an accuracy of a fraction of a percent, and loads the
> circuit with as much loading as desired, from a few ohms to hundreds of
> thousands of ohms.
>
Eureka. ............ There it is, Tom Brown "as much loading as desired".
Mr. Rauch has been adjusting the NA's resistive termination controls to
whatever HE presumes is appropriate. I wonder if this is how Mr. Rauch
determined that no significant VHF current flows through a VHF suppressor
inductor, Ls.
-----------------------------------------------
e-mail copies to Mr. Rauch, Mr. Brown, et al.
--
--Rich-- ag6k, 805-386-3734
Wes Stewart
In article <measures-ya023180...@news.vcnet.com> meas...@mail.vcnet.com (R. L. Measures ) writes:
>From: meas...@mail.vcnet.com (R. L. Measures )
>Subject: Re: Good source for amplifier tips
[snip]
[more snipping to make my newsreader happy]
Whew!
Why not just use Q=R/X for the parallel case?
73, Wes N7WS
Wes Stewart
In article <32A86A...@mail.teclink.net> "Tom C. Brown, Jr." <mad...@mail.teclink.net> writes:
>From: "Tom C. Brown, Jr." <mad...@mail.teclink.net>
>Subject: Re: Good source for amplifier tips
>Wes Stewart wrote:
>>
>> In article <32A662...@TECLink.net> "Tom C. Brown, Jr."
><tbr...@TECLink.net> writes:
>> >From: "Tom C. Brown, Jr." <tbr...@TECLink.net>
>> >Subject: Re: Good source for amplifier tips
>> >Date: Wed, 04 Dec 1996 23:51:08 -0600
>>
>> >Tom C. Brown, Jr. wrote:
>> [snip]
>>
>> >Don't forget that Rich was not making exact, quantified measurements
>> >with the GDO. He was (1) finding resonances in the circuit *WITHOUT*
>> >attaching anything physically to the circuit and he was (2) able to
>> >compare the Q in a "before and after" situation by just looking at the
>> >response of the dip meter in each case. In other words, he was able to
>> >say that the Q was greater or less after swapping the suppressors. If
>> >*you* are unable to make such judgements with a GDO, then I guess you
>> >really wouldn't have much use for one.....
Actually, I have two Millens and one RF Measurements Corporation GDO, but I
don't use them for Q measurements.
>>
>> Please! I'm sure Rich's attempts at determining circuit parameters are
>> well-meaning, and a GDO has its place, but... This ain't one of them.
>I disagree, Wes. Finding resonances and potential frequencies of
>oscillation are applications where a GDO excels. It did years ago, and
>it still does. If Rich's GDO is not indicating the resonant frequencies
>in the circuit, then please tell us what it IS indicating?
>If the circuit is NOT most likely to oscillate at one of its resonant
>points, then please tell us where it WOULD be most likely to oscillate.
>And please include an explanation of just why the circuit would be more
>liable to oscillate at some place where it is not resonant. You're
>contradicting some pretty well known facts and principles there, Wes.
No, I'm not contradicting any well-known "facts".
The requirements for sustained oscillation are that A*B=1
where A is the amplification factor and
B is the transfer characteristic of the frequency-selective network,
which results in the requirement that the product of A*B must have a magnitude
of 1 and a phase angle of 0 degrees.
(REF: Electronic's Designers' Handbook, 2nd edition, edited by L. J.
Giacoletto, p 16-13.)
NOTE: there is NO requirement that anything be resonant! For example, a phase
shift oscillator can be made with an inverting amplifier and three or more
cascaded RC networks which provide an additional 180 degrees of phase shift.
Please try to find the resonant frequency of this circuit with your GDO.
Even most crystal oscillators do not operate at the EXACT crystal resonance.
Instead, due to non-ideal phase shifts in the active device, the frequency of
oscillation shifts slightly to a point where the phase shift through the
crystal and associated components makes up for the non-ideal device. BTW,
most hams believe that crystals operate at either their minimum (series
mode) or maximum (anti-resonance) impedance. As the foregoing states, this is
not necessarily the case. A crystal oscillator is stable because of the
crystal's very steep phase vs frequency slope, not its point of minimum loss.
But, I digress....
Learning that there is a spurious resonance some place DOES NOT mean that this
is a point of instability. If the loop gain is low enough, the system does
not oscillate. Period. Conversely, it the feedback phase shift is correct
(or incorrect in this case) and the loop gain is greater than one, we have
oscillation.
It appears that you're going to suck me into one.
>You are starting to sound more and more like Tom. When you start
>telling other folks what they do and don't have an understanding of, you
>are (to borrow a phrase from Paul Harvey) telling us more than you
>know. Not usually a wise practice.
>> I recently retired after 30 years with a
>> major aerospace firm where I was engaged for most of that time in making r-f
>> and microwave measurements and doing failure analysis on everything
>> from components to all-up missiles.
>>
>> I can assure you, that we "hooked something" to every damn one of these
>> devices in order to characterize it electrically. Tom has provided data in the
>> form of |S21|, ie, the magnitude of the forward transmission component. In
>> layman's terms, this is the "gain" of the device, expressed in dB, vs.
>> frequency. Viewed graphically, you would (he did) see the peaks and nulls of
>> transmission through the tube.
>Wes, you have taken my comment that was specifically focused on *this*
>case (assessing the resonant frequency and relative Q of the PA cicuits)
>and applied it to test and measurement in general. That is not what I
>said or what I meant. The fact is that there are some cases with some
>circuits where attaching virtually anything conductive to the circuit
>will compromise the measurements.
OH? Tom R posted data from a "specific case" which you challenged and I
defended, both specifically and generally. Indeed, EVERY measurement perturbs
the device under test to some degree, even your GDO. The beauty of modern
network analyzers is that the pertubation is quantifiable. For instance, I
can and have measured transistors that have an input reflection coefficient
greater than one. This means that if you were talking about reflected power,
more is coming out than you are putting in, or expressed another way, the
resistive component is negative.
In my measurement circuit, the device is stable, not oscillating, and
therefore, I have highly "peturbed" it with the measurement system. This does
not mean, however, that I can't predict how it will perform in another
circuit. In fact, with the known data, I can compute how prone to oscillation
an amplifier might be which used this device. Will your GDO predict this?
>In my opinion, a GDO with no physical connnection, and with care not to
>over couple the GDO into the circuit, is one of the least invasive ways
>to measure the resonant frequency of the circuit IN THIS CASE.
>What kind of network analyzer would be less invasive than that? You are
>the expert on modern network analysis (you told me so). Please
>enlighten us.
>>
>[snip]
>Would the analyzer that Tom used present an infinite impedance to the
>circuit under test AT THE POINT OF CONNECTION?
Of course not. Nothing has an infinite impedance. As I stated above, the
"invasiveness" can be accounted for.
>If yes, then I have no problem with his measurment technique. If not,
>then his network analyzer IS affecting the data. If you don't know,
>then you are (again) telling us more than you know. (still not a wise
>practice)
No again. The network analyzer is affecting the circuit but it is not
affecting the data. The data are corrected for the effects of the measurement
system.
>>
>> >Now, if you *did* make all your measurements with no physical connection
>> >to the circuit, and without overcoupling anything in any way to the
>> >circuit, then please disregard the above. Just a thought......
>>
>> >73, Tom KJ5IE
>>
>> Wes N7WS
>Tom KJ5IE
N7WS
w8j...@aol.com
Hi Richard,
Instead of more RM style personal attacks and non-technical comments about
the other Richard we all know of, I'd like to see you answer questions
about Q.
I think your parasitic theories are absolutely wrong. VHF Q does NOT
change significantly with the addition of nichrome, and that is a fact you
are trying to avoid. As a matter of fact, a small adjustment in the
suppressor resistance and inductance makes a larger change than the
addition of "nichrome".
You realize if the nichrome in the suppressor's winding of wire does not
significantly lower VHF Q, then everything you attribute to VHF
oscillation must be a product of vivid imaginations, or directly
attributed to MY claim that HF Q and mistuning / overdriving are the major
causes of arcs and sparks in the tank.
Let's make an agreement.
If I can not lower Q (using coventional components) the same amount or
more at VHF without adding as much loss of Q on ten meters as the nichrome
does, I will concede you have been hawking something of ### great
technical merit ###. I will endorse your product as a good method of
parasitic suppression.
If the system you hawk does NOT significantly reduce VHF Q at or near the
frequency of oscillation (let's say between 140 and 200 MHz) compared to
what I can do with conventional components, will you admit the VHF
parasitic cures and theories you have been making are false?
This sounds fair to me, since you claim to have re-discovered the only
cure for VHF parasitics.
73 Tom
R. L. Measures
In article <n7ws.212...@azstarnet.com>, n7...@azstarnet.com (Wes
Stewart) wrote:
> In article <measures-ya023180...@news.vcnet.com>
meas...@mail.vcnet.com (R. L. Measures ) writes:
> >From: meas...@mail.vcnet.com (R. L. Measures )
> >Subject: Re: Good source for amplifier tips
> >Date: Fri, 06 Dec 1996 10:27:09 -0700
>
>
>
> [snip]
>
> >This is a parallel circuit, so most things are inverted. (Standing on head
> >may be useful.) snip....
> Whew!
>
> Why not just use Q=R/X for the parallel case?
>
> 73, Wes N7WS
Whew is right, Wes. Thanks for the heads up. Sorry. I forgot that for a
parallel circuit, the formula is Q=R/X. So how do we figure total Q when
the parallel Ls contains series R of its own?
I'm still trying to figure out why Tom Rauch cancelled his post of 28
November.
-----------------------------------------------
e-mail copies to Mr. Stewart, Mr. Rauch, et al.
--
--Rich-- ag6k, 805-386-3734
Wes Stewart
In article <measures-ya023180...@news.vcnet.com> meas...@mail.vcnet.com (R. L. Measures ) writes:
>From: meas...@mail.vcnet.com (R. L. Measures )
>Subject: Re: Good source for amplifier tips
>In article <n7ws.212...@azstarnet.com>, n7...@azstarnet.com (Wes
>Stewart) wrote:
>> In article <measures-ya023180...@news.vcnet.com>
>meas...@mail.vcnet.com (R. L. Measures ) writes:
>> >From: meas...@mail.vcnet.com (R. L. Measures )
>> >Subject: Re: Good source for amplifier tips
>> >Date: Fri, 06 Dec 1996 10:27:09 -0700
>>
>>
>>
>> [snip]
>>
>> >This is a parallel circuit, so most things are inverted. (Standing on head
>> >may be useful.) snip....
>> Whew!
>>
>> Why not just use Q=R/X for the parallel case?
>>
>> 73, Wes N7WS
>Whew is right, Wes. Thanks for the heads up. Sorry. I forgot that for a
>parallel circuit, the formula is Q=R/X. So how do we figure total Q when
>the parallel Ls contains series R of its own?
First, compute the parallel equivalent of the series RL alone.
Then add the parallel resistance using the usual formula for such things.
This gives the new parallel equivalent R||L. Q is computed as above.
N7WS
w8j...@aol.com
In article <19961206230...@ladder01.news.aol.com>, k1...@aol.com
writes:
>
>I may be stepping into something that is out of my league, but it seems
to
>me that the "physical connection" argument smacks of religeous fantasy.
>There was a time when virtually all significant RF coupling was done
>inductively--and even today's most advanced solid state power amplifiers
>transfer huge amounts of power in this manner. If I were to connect you
>to a powerful 160 MHz source, Tom, would you rather it be through a .5 pF
>capacitor with a 10 Meg resistor in series--or through a large inductive
>link with a high co-efficient of coupling. Your choice!
This is another area where Mr. Brown and Measures obviously do NOT
understand what they are doing.
The very requiremnt that current is removed from the oscillator, causing a
dip, indicates coupling is VERY high! The oscillator coil in the dip meter
has such high resonant currents that it has a very intense field.
The coupling is so high it actually robs the oscillator of current, cause
a dip to appear on the indicator.
I can measure the field from my GDO almost ten feet away with the probe of
my network analyzer, yet the GDO has "no idea" the NA probe is there!
It's kind of like the difference between using GPS or a compass, isn't it?
Rich, you allude I adjust the resistance "to what I want". That statement
is idotic when we look at the facts.
You couple a GDO to a cold tube that appears as only reactances, when in
operation each element has a very finite resistance. For example, negative
feedback and other effects combine to make the cathode have a low
resistive part in it's terminal impedance and alos greatly reduce the
anodes resistive component (at idle it can be taken as Ep/Ip, although it
varies over the plate current cycle).
I use the mean impedance of the element while the tube is conducting as
the test resistance, while you use an open circuit. Your test is fine if
you never power the tube up. When you adjust the tank components, do you
terminate the anode in a resistance equal to the operating resistance
given in a Chaffe analysis, or do you just leave the anode hang loose?
I find it remarkable you always seem to find fault with correct
proceedures while using the worse methods possible in your tests.
What about Q, how much change does you nichrome make at VHF? Is it a large
cahnge, or insignificant?
73 Tom
R. L. Measures
In article <n7ws.219...@azstarnet.com>, n7...@azstarnet.com (Wes
Stewart) wrote:
snip...
> First, compute the parallel equivalent of the series RL alone.
> Then add the parallel resistance using the usual formula for such things.
> This gives the new parallel equivalent R||L. Q is computed as above.
>
Good thinking, Wes.
--
--Rich-- ag6k, 805-386-3734
R.L. Measures
> Hi Richard,
>
> Instead of more RM style personal attacks and non-technical comments about
> the other Richard we all know of, I'd like to see you answer questions
> about Q.
??? // RE: Q: On 6 Dec. I answered your Nov. 28 post about Q .
> I think your parasitic theories are absolutely wrong. VHF Q does NOT
> change significantly with the addition of nichrome, and that is a fact you
> are trying to avoid. As a matter of fact, a small adjustment in the
> suppressor resistance and inductance makes a larger change than the
> addition of "nichrome".
Since 1990, I have been aware that you have a different opinion, Tom.
However, you appear to base your opinion on the assumption that
insignificant VHF current flows in the suppressor inductor, Ls. When
presented with the calculations (from the very problem you proposed) which
showed that similar VHF currents flow in both Rs and Ls, you proffered
unplausible arguments to the contrary.
> You realize if the nichrome in the suppressor's winding of wire does not
> significantly lower VHF Q, then everything you attribute to VHF
> oscillation must be a product of vivid imaginations, or directly
> attributed to MY claim that HF Q and mistuning / overdriving are the major
> causes of arcs and sparks in the tank.
I do not realize that resistance wire (nichrome) has no effect on VHF-Q. I
have observed that the VHF dip in amplifier anode circuits is rather
different when silver-plated strap suppressor inductors (Ls) are replaced
with a resistance wire Ls. With silver (and copper), the dip is sharper
and deeper. With the resistance wire, the dip is broader and shallower.
With roughly equal currents flowing in Ls and Rs, why not?
This is nothing new, Tom. Mr. F. E. Handy was aware of this technique 70
years ago.. You can read about it in the 1926 Edition of the *Radio
Amateur's Handbook*. You can read about it in the 10/88, 9/90, 10/90 and
1/94 issues of *QST* magazine. There's an article about calculating values
for VHF parasitic-suppressors in the 3/89 issue.
> Let's make an agreement.
I am reluctant to make an agreement with you, Mr. Rauch.
> If I can not lower Q (using coventional components) the same amount or
> more at VHF without adding as much loss of Q on ten meters as the nichrome
> does, I will concede you have been hawking something of ### great
> technical merit ###. I will endorse your product as a good method of
> parasitic suppression.
An endorsement from someone who dismisses well-established AC circuit
analysis , who purports that dipmeters are incapable of finding resonances,
would amount to an UNendorsement.
--------------------------------
e-mail copies to Mr. Rauch, et al.
--
--Rich-- ag6k, 805-386-3734