71A amp
Raymond Koonce
I finally got my 71A amp up and running and in a proper chassis. It's a
push-pull design with transformer coupled 56s driving the 71As. It has
transformer input with a pair of Newcomb microphone transformers.
Really nice sound, zero hum in Klipsch Cornwalls. I used, at Al Marcy's
suggestion (thanks Al) Blumlein Garters. I found the bass to much
improved over a standard bias method. Nine tubes and sixteen
transformers for 2 watts/channel :-). I posted a photo on
alt.binaries.pictures.radio. I will have a web page up tomorrow.
Happy listening!
Raymond
dre7
Congratulations! My server won't let me access binaries anymore. Do you
think you could email a pic? I'm really happy with my transformer input,
interstage coupled 6J5>PP 6B4G. Standard cathode bias, though. Bass is
decent through my T-Line speakers, but I've downloaded the Blumlein circuit
as I admire it's elegance and might put it to use in some other project.
Andrew
Raymond Koonce <rko...@dsityler.com> wrote in message
news:49fe2496$0$23746$bbae...@news.suddenlink.net...
Jeff
That is absolutely beautiful, Raymond. Congratulations!
On Sun, 03 May 2009 19:11:19 -0400, Raymond Koonce <rko...@dsityler.com>
wrote:
--
Jesla
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
Sal Brisindi
Beautiful job Raymond. I like the copper chassis!
Sal
Raymond Koonce
at http://www.timebanditaudio.com/71A/71a.html.
Best regards,
Raymond
Raymond Koonce
Raymond Koonce
Thanks Sal. It's a lot of work, but it sure looks nice when it's done.
Raymond
Raymond Koonce
>
> I'm curious, what technique did you use for the socket blind screw
> mounting?
Thanks,
I built the amp on a sub-chassis and mounted it from underneath. You
can see more details at http://www.timebanditaudio.com/71A/71a.html.
Raymond
GerryE123
SE 71A amp? Thanks.
Gerry
"Raymond Koonce" <rko...@dsityler.com> wrote in message
news:49fe2496$0$23746$bbae...@news.suddenlink.net...
Raymond Koonce
Thanks for the kind words. I've never heard a SE 71A amp, but the
output power is only 800mW or so. I'm getting 2 whole watts from a
pair. The sound is very good, IMO. Maybe the transformer
coupling/splitting? The Blumlein Garter circuit really made a distinct
change in the bass response by eliminating any DC offset in the OPT.
Best,
Raymond
Andre Jute
That's a thing of beauty, Raymond. Congratulations! It must have a
very high Dunker Factor (weight/power ratio).
Andre Jute
Mo' iron is betta iron!
Raymond Koonce
Hi Andre,
Thanks! It was a lot of work getting the chassis built. To me, that's
the most challenging part of a project. The wiring is easy. The
breadboard that I built this on was made from the cases of old computer
power supplies with appropriate holes punched to mount the sockets. I
haven't weighed it yet, but I'll estimate 35 pounds, so 4 watts (2 per
side) divided into 35 pounds gives a Dunker of 8.75. Not too bad, I get
about 6 wpc from my SE 300Bs which weigh in at 60 pounds, a DF of 5 :-)
Raymond
Patrick Turner
Raymond Koonce wrote:
>
> I posted a web page with additional photos and commentary. Take a look
> at http://www.timebanditaudio.com/71A/71a.html.
>
> Best regards,
>
> Raymond
Very nice work Raymond.
I'd probably go broke if I limited myself to making 2W amps, but then if
I only made wooden boats then I'd also go broke.
In 2009, much technology which radiates the skill of its maker is only
for those who fall asleep when they see a mass made article, like say a
fibreglass boat, or a modern house that resembles a plastic chook pen.
I cannot understand why bass would be better with Garter biasing though.
Effectively each output cathode has a cap to 0V. But there is twice the
normal Rk used, so Iadc regulation would be better. Its OK for class A.
And if one tube begins to conduct more Iadc than the other as it ages,
then its rise in Ek causes half that rise on the other tube's grid, and
that raises the Ek and then the Eg of the original tube. I'm not sure
sure the Garter is worth the extra heat and B+ lost because you have a
higher Ek than normal, but in this case the waste is negligible because
its only a 2W amp. The Garter is at least better than having a common Rk
and Ck for both cathodes IMHO, although Williamson didn't think so. But
then he had some adjust pots to trim his amps dc and ac. These went up
in smoke when a KT66 or an 807 decided to slowly die. Leak just used
individual Rk and Ck. I like this the best because its simple, and bias
change in one tube doesn't change anything very much in the other tube.
But because we now have had solid state sinjce 1947, we can use a few
devices to make elegant bias management circuits which stop smoke
occuring and allow visual monitoring of bias balance condition.
Patrick Turner.
pf...@aol.com
Very nicely done - a great deal of effort for those couple of watts,
but if ever any speakers deserved such, those would be the Cornwalls.
Please see my direct e-mail when you get the time.
Thanks!
Peter Wieck
Melrose Park, PA
Staring at old audio chassis as a hobby :)
>
> I cannot understand why bass would be better with Garter biasing though.
Hi RATs!
The idle current balances at a point halfway between the points each
tube would operate at, without garters.
This reduces residual current through the halves of the output
transformer primary - minimizing magnetic corruption of the core.
I try to hear first and understand later ;)
Here is Broski's further musings:
http://www.tubecad.com/2009/04/blog0163.htm
He doesn't think listening first is worthwhile, neither ;) His first
effort was modeling on a software testbench.
It is a big planet. Garters prolly are worthless against swine flu.
Happy Ears!
Al
PS After a quarter century in the AZ sun, this endless TX cloud cover
and thunderstorms are an interesting change.
Many of the locals ignore SAD, but us sunspots see their stress. Not a
happy bit of knowledge.
Thank fate for ears ;)
Staring at old audio chassis as a hobby :)
Some more Blumlein esoterica :)
Al
Patrick Turner
Are not Garters really only good for keeping ladies stockings at the
same hight on each of her thighs?
Meanwhile braces are far more useful, but not nearly so sexy, because
they keep a man's trousers at the correct height for a wholesome and
becoming kind of life. If one trouser leg on the left leg drops a bit,
it pulls at brace on one side which has an effect over the soulder to
the arse side of the trouser leg on right. Approximate symetrical
clothing appearance results.
But if a determined strumpet yanks hard on one trouser leg, and the man
does not defend himself, then one trouser leg followed soon by the other
will be whipped off and a they'll be a twang of braces and the whole
theory of balance control becomes farcical.
And so it is with tubes and Garter Biasing, all very hunky dory when you
look at it on paper or in a simulator but as soon as a tube decides to
seriously misbehave you have a problem.
Maybe a given tube needs 30V of bias. OK, then you can establish a fixed
bias of +30V to apply to the grids then the tube would need double its
normal value of Rk because Ek will try to rise to 60V not a normal 30V
above 0V.
The double Rk merely regulates Ek and hence Ia better; not perfectly,
but much better. All is well until you move into class AB operation from
class A....
And what do you do when you have 4 or more output tubes?????????
How do ya balance things then?
How does a sheila with 8 legs keep her garters all at the same height?
( Careful with such shielas, they tend to eat men they have just bonked.
)
Patrick Turner.
Raymond Koonce
Thanks Patrick,
I build these things because I like to do something with my hands and I
love to listen to music. Triodes do it best, IMO. I sure couldn't make
any money building these. This one took months of Saturdays to get
right (the chassis, not the circuit, it was fairly easy).
The small output transformers I used apparently saturate easily with any
unbalanced DC. Since 71As are getting like hen's teeth, matched pairs
are even harder to find. The Blumlein circuit eliminates the DC offset
and prevents the core saturation, allowing the bass response to improve.
I like the use of separate filament transformers for each 71A, which
allows me to use the CT of each one for the bias network.
Your electronic knowledge as a professional is much greater than mine as
a hobbyist, so I can't pretend to design these solid state bias
circuits. I'm sure they work better than what I've done, but this works
for me.
Best regards,
Raymond
Raymond Koonce
Thanks Peter. I appreciate your private email.
Raymond
Andre Jute
<tubegar...@aol.com> wrote:
> It is a big planet. Garters prolly are worthless against swine flu.
Heretic! Clown! Troll! Screech!
(signed) Lottie Offah-Lotta
Raymond Koonce
> On Mon, 04 May 2009 09:10:35 -0500, Raymond Koonce
> mounted is what threw me off.
I found some pan-head screws which allowed me to mount the sub chassis
closer to the external chassis.
Raymond
Ian Iveson
pathetic old man sex delusions deleted.
> And so it is with tubes and Garter Biasing, all very hunky
> dory when you
> look at it on paper or in a simulator but as soon as a
> tube decides to
> seriously misbehave you have a problem.
It will look on a simulator the same as it does in reality.
If you simulate with perfect and identical valves, then it
is true that you may not see what happens with one strong
and one weak. The same is true if you test the real circuit
only with similar good valves.
If you want to test for what happens in your kind of
scenario, then you can build a real circuit using whatever
extreme examples of valves you can find, or you can simulate
with exactly the kind of extreme circumstances you have in
mind. Or, in either case, you might add resistors or
otherwise modify the circuit so a good valve behaves badly,
for the purpose of the test.
Or you can predict using fag packet algebra.
Until you actually try simulating you won't know what
simulating is and it could be that you should stop
pronouncing on a subject about which you know absolutely SFA
or ppl will god forbid assume you know bugger all about all
the other things you pronounce upon without bothering to
read your stuff and work that out longhand.
So there.
Ian
Staring at old audio chassis as a hobby :)
Mr. Broski addresses these very concerns.
Happy Ears!
Al
Patrick Turner
People simulated economics processes and look at the mess the world has
gotten into.
Now they are trying to simulate world climate change, and that looks
like teaching us all to do too little too late about fixing climate
change due to CO2 rise.
The garter biasing scheme looks swell but if a tube malfunctiuons your'e
up shite creek in a barb wire canoe with no paddle.
I don't need to simulate anything using a PC. I do have a brain Ian.
And I can try something out using a soldering iron and find out all
about it in less time and with more accuracy than relying on simulation.
Patrick Turner.
Staring at old audio chassis as a hobby :)
>
> The garter biasing scheme looks swell but if a tube malfunctiuons your'e
> up shite creek in a barb wire canoe with no paddle.
>
Hi RATs!
If a tube malfunctions?
If the tube goes open or shorts, all is well. What fiendish evil are
you anticipating?
Happy Ears!
Al
Ian Iveson
>
> People simulated economics processes and look at the mess
> the world has
> gotten into.
That syllogism is missing a major premise, and the remaining
premise is false. The world's not in a mess, it's just that
the folk in some places have pampered themselves stupid on
heaps of "cheap" Chinese goods on credit and now they can't
pay until they pull their socks up, stop stuffing
themselves, and pull their weight. Simulation's a tool, and
in the case of macro economics lots of boffins got carried
away with particular avenues of investigation, and forgot
the basics. Ultimately, however, it was messing with the
real economy, not simulation, that actually led us into what
you perceive to be a mess.
> Now they are trying to simulate world climate change, and
> that looks
> like teaching us all to do too little too late about
> fixing climate
> change due to CO2 rise.
You can't blame simulation for that, obviously. Without it
we would have realised much later, if at all.
It's not just CO2 either, there's a heap of stuff to be
concerned about.
Weather forecasts have been done by simulation for
ages...that's what supercomputers were for. You can't do it
on the back of an envelope, and experimenting with the real
thing is problematic because there's only one earth to play
with. For big complicated things like global weather and
macro-economics, simulation is the best we can hope for. If
only we could unplug the political gobshites from the back
end of the economic forecasts, we might know better what's
happened to all that money, and where it came from in the
first place.
It's not necessary for valve circuit design, of course, but
it's a very useful tool if used wisely. That's plain for all
but you to see, it seems.
You can sketch a circuit, complete design or fragment, and
then actually make the sketch run like the real thing, and
test it with any instrument and signal you can imagine, all
in a jiffy. How can anyone be daft enough to say that's not
useful?
> The garter biasing scheme looks swell but if a tube
> malfunctiuons your'e
> up shite creek in a barb wire canoe with no paddle.
>
> I don't need to simulate anything using a PC. I do have a
> brain Ian.
Then use it to:
* Stop writing bollocks about simulation.
* Stop writing drivel about sex.
* Stop writing nonsense about China.
* Answer Al's question.
In that order, please.
> And I can try something out using a soldering iron and
> find out all
> about it in less time and with more accuracy than relying
> on simulation.
Rubbish. Expensive timewasting won't help you to start
charging fair prices for your products and services.
Ian
John Byrns
Patrick Turner <in...@turneraudio.com.au> wrote:
I don't care for the garter circuit at all, the reason is that it is
basically an undesirable form of positive feedback.
Consider what happens when the current in one tube increases due to some
internal change. The voltage across the lower cathode resistor then
increases which in turn increases the grid voltage on the other tube in
an attempt to equalize the currents through the two tubes. This action
causes the voltage across the lower cathode resistor of the second tube
to increase and that in turn unfortunately increases the grid voltage on
the first tube, the original one with the problem, further increasing
the current through the first tube, hardly a desirable result
considering that its current was too high to start with. Perhaps they
will both turn red and melt.
As an alternative to the garter let me propose the following scheme.
First use separate cathode resistors for each tube of twice the value
normally used for separate cathode resistors, the same as in the garter
circuit. This will result in lower than the desired current through the
tubes, so if the normal grid bias was 30 volts grid to cathode,
disconnect the grid resistors from ground and connect them to a divider
from B+ that will elevate the grids to 30 volts above ground. This will
restore the desired current through the tubes, while the doubled cathode
resistors will increase the stability of the tube currents with respect
to variations in the tube parameters, without the undesirable
interactions of the garter circuit.
Perhaps one of the simulators in the group could simulate this
configuration and compare it with the results from the garter circuit to
see which scheme provides better balance as tube characteristics vary.
--
Regards,
John Byrns
Surf my web pages at, http://fmamradios.com/
ixtar...@yahoo.com
solid supply is the best, overwhelmingly, asfety relay can be
provided if you like.
I wonder how this amp measures. Without repeatable measurements I
have no opinion on any amp because very terrible amps can " sound
good" to some people who like infidelious sound. Especially through
Cornwalls, which are a rude speaker in many situations.
Proper tweaking of the Cornwall can improve things quite a bit=the
Avedon T35 EV tweeter needs to go-but still it's raw.
ixtar...@yahoo.com
>
> I build these things because I like to do something with my hands and I
> love to listen to music. Triodes do it best, IMO.
Yes but if it doesn't do it accurately why bother? Building time
intensive but bad things is a waste of time. Build good things.
I sure couldn't make
> any money building these. This one took months of Saturdays to get
> right (the chassis, not the circuit, it was fairly easy).
>
> The small output transformers I used apparently saturate easily with any
> unbalanced DC.
Wouldn't buying a transformer made for a single ended circuit have
been much more sensible? It isn't as though they are terrifically
expensive.
Patrick Turner
"Staring at old audio chassis as a hobby :)" wrote:
>
They often don't suddenly go open, (unless a heater becomes
disconnected), or go to a short.
They dwell between the two states when they finally decide to quit being
nice, and they sit there forawhile conducting far more Ia than they
should.
I built an integrated stereo amp last Xmas for a guy with 2 x Sovtek WXT
El34 per channel. Now the protect circuit I put in there comes alive
because 2 have begun to go dodgy, with higher Ia and even a positive
grid when they should be at 0V.
I find another pair of EL34 which bias much better than the the ones
which malfunctioned so badly.
The 4 tubes now each have 570 ohm Rk plus 470uF bypasses and Ek = 30V
+/- 0.5V for all four after 10 minutes and after an hour.
Lets hope I have found the cause of the problem - dud tubes from Sovtek.
I don't get many duds, but sometimes one does, and you need to have a
circuit that lets you know about it.
Ea = 400V, Ia < 55mA, so the tubes don't work hard.
Tubes have a bit of the devil in them, and are thus capable of fiendish
evil.
The devil also brings us many delights otherwise forbidden, so we may as
well accept a bit of devil unless you plan on dying of boredom.
Patrick Turner.
>
> Happy Ears!
> Al
Patrick Turner
Ian Iveson wrote:
>
> Patrick Turner wrote:
> >
> > People simulated economics processes and look at the mess
> > the world has
> > gotten into.
>
> That syllogism is missing a major premise, and the remaining
> premise is false. The world's not in a mess, it's just that
> the folk in some places have pampered themselves stupid on
> heaps of "cheap" Chinese goods on credit and now they can't
> pay until they pull their socks up, stop stuffing
> themselves, and pull their weight. Simulation's a tool, and
> in the case of macro economics lots of boffins got carried
> away with particular avenues of investigation, and forgot
> the basics. Ultimately, however, it was messing with the
> real economy, not simulation, that actually led us into what
> you perceive to be a mess.
Blah blah blah.
Its a bit difficult to correctly simulate the world economy.
Or even to play around with the real one.
The COWPAT effect kicks in,
Chance Of Working Any Time = 1 / N squared where N is the number of
variable factors eg, things just forgotten, left out due to politics, or
some math error, or some typo, or a dropped spanner.
>
> > Now they are trying to simulate world climate change, and
> > that looks
> > like teaching us all to do too little too late about
> > fixing climate
> > change due to CO2 rise.
>
> You can't blame simulation for that, obviously. Without it
> we would have realised much later, if at all.
I ain't so sure. Trouble with Greenhouse Warming is that it appears to
be creeping up on us and therefore nobody notices it until its too late.
But cast your mind to the records indicated by geological discoveries.
The records for sea level changes over the last the 30 million years are
all in the rocks. The sea levels allowed Tasmania to be a part of
mainland Oz when the levels were up to 130metres lower. At other times
the the sea levels were very many metres higher than now and Oz had a
vast inland area coverd by a sea extending up the Murray Darling river
basins. Land area was a whopping 8% less when much of the ice at the
poles melted.
Anyway, even without mankind and his wife and kids it looks certain we
live on achanging planet, but now WE are the damn thing changing things
MUCH faster than occurs in nature. Humans are acting like a giant
volcanoe spewing CO2 24/7. I expect a big rise in sea levels before
simulations tell us we are doing too little too late, and therefore
cities like Sydney, Melbourne, Brisbane and Perth will all be well under
water way before we want them to be flooded. The cost of rebuilding the
cities on higher ground will be another huge boost to CO2 emissions. By
then we will have chopped all the forests down and abolished all
wildernesses and made the oceans acidic.
>
> It's not just CO2 either, there's a heap of stuff to be
> concerned about.
Indeed, and far more than we can ever realise.
>
> Weather forecasts have been done by simulation for
> ages...that's what supercomputers were for.
Well they do have many more ways of measuring temp and moisture content
and they are taking in vastly higher numbers of measurements from around
the globe so the increase in data does allow some room to play with
computer models.
But If I had a quid for each time the local weather beureau forcast a
good fall of rain and saw it rain only 1mm, then I'd be rich.
> You can't do it
> on the back of an envelope, and experimenting with the real
> thing is problematic because there's only one earth to play
> with. For big complicated things like global weather and
> macro-economics, simulation is the best we can hope for. If
> only we could unplug the political gobshites from the back
> end of the economic forecasts, we might know better what's
> happened to all that money, and where it came from in the
> first place.
You can't simulate God. Nor the universe. We hardly know about either,
because we have finite brains and the knowledge of God and the Universe
is infinite, and you can't fit infinite knowledge into some finite
computer, or even into a human brain, or all the brains on earth if you
could wire them up in series and parallel.
>
> It's not necessary for valve circuit design, of course, but
> it's a very useful tool if used wisely. That's plain for all
> but you to see, it seems.
>
> You can sketch a circuit, complete design or fragment, and
> then actually make the sketch run like the real thing, and
> test it with any instrument and signal you can imagine, all
> in a jiffy. How can anyone be daft enough to say that's not
> useful?
I do the simulation in my brain.
I don't need the program on the PC.
>
> > The garter biasing scheme looks swell but if a tube
> > malfunctiuons your'e
> > up shite creek in a barb wire canoe with no paddle.
> >
> > I don't need to simulate anything using a PC. I do have a
> > brain Ian.
>
> Then use it to:
>
> * Stop writing bollocks about simulation.
> * Stop writing drivel about sex.
> * Stop writing nonsense about China.
> * Answer Al's question.
I have used it to write about all thse things at times. I give my take
on things.
Would you prefer I would have simulated conversation about these
subjects?
As soon as the number of variable things in some system under
observation reach a critical number, or the interactivity between these
things becomes subject to random variations and quantities, then
everyone disagrees about the input information so the output information
varies very widely.
I try to live a life with minimum complexity. I don't need to simulate
as a result.
>
> In that order, please.
>
> > And I can try something out using a soldering iron and
> > find out all
> > about it in less time and with more accuracy than relying
> > on simulation.
>
> Rubbish. Expensive timewasting won't help you to start
> charging fair prices for your products and services.
Simulation is fine for those who like to spend vast amounts of time
theorizing and building a model they think might fly.
Its very necessary if you wish to build a bridge or a fighter aircaft.
Sooner or later though something has to be built, like the Sydney
Harbour Bridge, or a Spitfire, and people managed with only slide rules
and common sense.
Probably someone simulated making a bicycle frame with carbon fibre
instead of steel. But to find out a whole lot more someone had to make a
carbon fibre bicycle, and then they found out the computer aided design
program didn't get everything.
Anyway, If I had the money, I might buy a bike made with mainly carbon
and resin, because it'd be 3Kg ligher than the one I have now which was
made using steels designed by Reynolds after a long tradition of trial
and error, with very little computer assistance.
One trouble with audio engineering on a computer is that there are no
input buttons or windows to fill and then enter about the required
subjective sound quality.
Patrick Turner.
>
> Ian
Patrick Turner
One tube could lose emission and fade down, and then the other would try
to follow it.
Its the blind leading the blind, no?
>
> As an alternative to the garter let me propose the following scheme.
> First use separate cathode resistors for each tube of twice the value
> normally used for separate cathode resistors, the same as in the garter
> circuit. This will result in lower than the desired current through the
> tubes, so if the normal grid bias was 30 volts grid to cathode,
> disconnect the grid resistors from ground and connect them to a divider
> from B+ that will elevate the grids to 30 volts above ground. This will
> restore the desired current through the tubes, while the doubled cathode
> resistors will increase the stability of the tube currents with respect
> to variations in the tube parameters, without the undesirable
> interactions of the garter circuit.
That is exactly what I said in a recent post.
>
> Perhaps one of the simulators in the group could simulate this
> configuration and compare it with the results from the garter circuit to
> see which scheme provides better balance as tube characteristics vary.
If anything was simulatable, it would be a Garter circuit effect.
Patrick Turner.
Patrick Turner
ixtar...@yahoo.com wrote:
>
> The garter circuit just isn't terribly good. Fixed bias with a good
> solid supply is the best, overwhelmingly, asfety relay can be
> provided if you like.
Certainly, active protection against bias failure in fixed bias amps is
a must, and also needed is some means of visually indicating the Ia
balance in 1/2 primaries of the OPT, ie, a pair of LED.
Fixed bias is good, and can be argued as the best, but in hi-fi amps
which make 97% of their power in pure class A there is simply no need
for fixed biasing. All SE amps can happily use cathode biasing.
Patrick Turner.
Raymond Koonce
>> I build these things because I like to do something with my hands and I
>> love to listen to music. Triodes do it best, IMO.
>
> Yes but if it doesn't do it accurately why bother? Building time
> intensive but bad things is a waste of time. Build good things.
Actually, this is a good thing. It pleases me and I'm the only person
I'm trying to please. Apparently it also pleases some others who have
heard it.
>
>
> I sure couldn't make
>> any money building these. This one took months of Saturdays to get
>> right (the chassis, not the circuit, it was fairly easy).
>
>
>> The small output transformers I used apparently saturate easily with any
>> unbalanced DC.
>
>
> Wouldn't buying a transformer made for a single ended circuit have
> been much more sensible? It isn't as though they are terrifically
> expensive.
Why would it be sensible to buy a single ended transformer for a
push-pull circuit? The ones I used were terrifically free.
Dave Ryan
: ixtar...@yahoo.com wrote:
: >> I build these things because I like to do something with my hands and I
: >> love to listen to music. Triodes do it best, IMO.
: >
: > Yes but if it doesn't do it accurately why bother? Building time
: > intensive but bad things is a waste of time. Build good things.
:
: Actually, this is a good thing. It pleases me and I'm the only person
: I'm trying to please. Apparently it also pleases some others who have
: heard it.
Thats the most important fact. Plus your amp not only sounds great, it's a thing
of beauty. Very impressive chassis work. Keep building and most importantly,
keep listening. Don't let a bunch opf party poopers rain on your parade just
because your circuit displeases them(despite the fact they've never heard it, some
people are amazing hiding behind their keyboards)
Way to go Raymond.
-dave
Jeff
I could care less what they would have to say about my ST70 that I've had
for almost 40yrs. I listen to it everyday and wouldn't trade for anything
else(yes it's been updated and moded over the years). I have piles of SS
equiment in storage because I keep coming back to my Dynaco.
Absolutly beautiful work, Raymond. Wish I had half of your talent.
--
Jesla
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
John Byrns
flipper <fli...@fish.net> wrote:
> On Wed, 06 May 2009 13:16:03 -0500, John Byrns <byr...@sbcglobal.net>
> wrote:
>
> >I don't care for the garter circuit at all, the reason is that it is
> >basically an undesirable form of positive feedback.
> >
> >Consider what happens when the current in one tube increases due to some
> >internal change. The voltage across the lower cathode resistor then
> >increases which in turn increases the grid voltage on the other tube in
> >an attempt to equalize the currents through the two tubes. This action
> >causes the voltage across the lower cathode resistor of the second tube
> >to increase and that in turn unfortunately increases the grid voltage on
> >the first tube, the original one with the problem, further increasing
> >the current through the first tube, hardly a desirable result
> >considering that its current was too high to start with. Perhaps they
> >will both turn red and melt.
>
> Well, with a cursory look it seems that way but I don't think so and
> an equally casual observation is the circuit's symmetry so how does it
> know which tube to 'follow' and if FB is positive why doesn't it trot
> down with the low current tube instead of up?
Your cursory look is way too casual and fails to provide a qualitative
demonstration of what is actually going on. It doesn't know which tube
to follow, it follows the tube or tubes that change. For my example I
assumed that only one of the two tubes changed and that it changed in a
way such that the current through it increased, if the tube that changed
had gone down in current the normal tube would have followed that tube
down.
I left it for the student to figure out what happens when both tubes are
changing, in the same direction or in different directions.
> Let's do a more exhaustive analysis.
>
> First, take 'ideal' equal tubes and note that the voltages on both
> cathodes are equal and our 'half way' point, where the grids are cross
> connected, are also equal. Ok, cool, so let's reconnect the grids to
> the 'half way' point on the *same* side, removing the feedback.
> Voltages and currents remain the same.
>
> Now wave a magic wand over tube A and transform it into a 'more
> conducting' tube. I.E. we have an imbalance. It's cathode V rises and
> the 'half way' point voltage rises. So far this is as you expected.
>
> Now, swap tube B's grid from it's own 'half way' point back to tube
> A's 'half way' point, restoring the first half of the feedback loop.
> Current through tube B will increase from the higher grid voltage.
> again as you expected.
>
> However, it will not conduct as much as tube A because it's the 'less
> conductive' tube. So it's cathode will not rise as high as tube B's
> cathode nor will it's 'half way' point. Or, put another way, we still
> have an imbalance, but not as much because it's grid V was increased.
>
> However, now reconnect tube A's grid to tube B's 'half way' point,
> restoring the second half of the feedback loop. Tube A's grid V
> becomes *less*. I.E. the 'less conducting' tube B's 'half way' point
> is pulling down tube A... and to a current less that the 'no feedback'
> case we started from.
>
> And that part deviates drastically from your analysis because you see
> them both 'raising each other' but the problem is, 'relative to what'?
> And I believe that breaking the feedback, observing the 'balanced
> state', unbalancing the tubes, and then reconnecting answers that
> question. Tube B conducts more than it would 'alone' but tube A
> conducts *less* than it would 'alone' and the circuit does not 'run
> away' chasing the high current tube.
Your last sentence, if true, says that the believers in the garter
circuit are completely wrong in their belief that the garter circuit
improves balance! Fortunately you are wrong and the garter circuit
believers are correct, even though the garter circuit has its problems
and there are better solutions.
I did not say the garter circuit would "run away", what I said was that
with the garter connection, if one tube develops some kind of fault that
causes it to draw more current, with the garter circuit that will cause
the current through the second normal tube to also increase so that
instead of just having one red tube, we will have two red tubes,
assuming their normal operating point was close to the point of redness.
Your thought experiment above is constructed in such a way as to obscure
what is really going on, your conclusion is simply wrong.
So let's do an even more exhaustive analysis.
First, take 'ideal' equal tubes and note that the voltages on both
cathodes are equal and our 'half way' point, where the grids are cross
connected, are also equal. Ok, cool, so let's reconnect both grids to a
fixed voltage source equal to the half way point voltage, removing the
feedback. Voltages and currents remain the same.
Now wave a magic wand over tube A and transform it into a 'more
conducting' tube. I.E. we have an imbalance. It's cathode V rises and
the 'half way' point voltage rises. So far this is as expected.
Now, swap tube B's grid from the fixed 'half way' point voltage back to
tube A's 'half way' point, restoring the first half of the feedback
loop. Current through tube B will increase from the higher grid
voltage. Again this is as expected.
At this point with only half of the feedback restored it is not easily
obvious, at least to me, which tube, A or B, will be conducting more.
However tube B will be conducting more than it was in the normal
condition because its grid voltage, obtained from the half way point on
tube A's cathode resistor will be higher, so the half way point voltage
on its cathode resistor will rise. Put another way, we still have an
imbalance, but not as much because it's grid V was increased.
However, now reconnect tube A's grid to tube B's 'half way' point,
restoring the second half of the feedback loop. Tube A's grid V becomes
*more*. I.E. the 'more conducting' tube B's 'half way' point is pulling
up tube A... and to a current more than the 'no feedback' case we
started from.
This further increase in the current through tube A in turn results in a
further increase in the current through tube B and on and on in ever
decreasing circles until we reach the point of equilibrium.
This is exactly what my original analysis predicted with both tubes
'raising each other' relative to the currents in a normal pair of tubes.
I believe that breaking the feedback, observing the 'balanced state',
unbalancing the tubes, and then reconnecting clearly demonstrates that
the current in both tubes increases when one tube in the garter circuit
becomes more conducting. Tube B conducts more than it would 'alone'.
> >As an alternative to the garter let me propose the following scheme.
> >First use separate cathode resistors for each tube of twice the value
> >normally used for separate cathode resistors, the same as in the garter
> >circuit. This will result in lower than the desired current through the
> >tubes, so if the normal grid bias was 30 volts grid to cathode,
> >disconnect the grid resistors from ground and connect them to a divider
> >from B+ that will elevate the grids to 30 volts above ground. This will
> >restore the desired current through the tubes, while the doubled cathode
> >resistors will increase the stability of the tube currents with respect
> >to variations in the tube parameters, without the undesirable
> >interactions of the garter circuit.
>
> I would think the 'fixed grid' might be less of a restraint and worse
> balance because it doesn't have the other tube pulling the grid in
> opposition.
That is exactly the question that needs quantitive analysis.
> I'm not sure, though, because it does have 'twice' the cathode
> feedback due to double the Rk.
Exactly, that's why quantitive analysis is needed.
> Might turn out they're equivalent with
> the 'half way' point simply providing the grid V you'd otherwise have
> to find a source for.
They can't be exactly equivalent because with the revised circuit the
current through the good tube does not change from normal. It is
conceivable though that they may provide an equivalent balance
improvement, quantitive analysis will answer that question.
Raymond Koonce
Thanks Dave,
Don't worry, I'll keep building. I can take all the bile they have to
spew. In my profession I negotiate with landowners for damages incurred
in oil and gas drilling operations. I've had my ass chewed so many
times it's like leather :-)
Raymond
Raymond Koonce
I too have several SS pieces that I never listen to. Tubes are better.
Best regards,
Raymond
John Byrns
flipper <fli...@fish.net> wrote:
> On Fri, 08 May 2009 10:54:37 -0500, John Byrns <byr...@sbcglobal.net>
> wrote:
>
> >In article <n2r705l23tr4ijskt...@4ax.com>,
> > flipper <fli...@fish.net> wrote:
> >
> >> And that part deviates drastically from your analysis because you see
> >> them both 'raising each other' but the problem is, 'relative to what'?
> >> And I believe that breaking the feedback, observing the 'balanced
> >> state', unbalancing the tubes, and then reconnecting answers that
> >> question. Tube B conducts more than it would 'alone' but tube A
> >> conducts *less* than it would 'alone' and the circuit does not 'run
> >> away' chasing the high current tube.
> >
> >Your last sentence, if true, says that the believers in the garter
> >circuit are completely wrong in their belief that the garter circuit
> >improves balance!
>
> No it doesn't. It says both tubes are pulled toward each other. That
> IS 'balancing'.
It says and I quote "Tube B conducts more than it would 'alone' but tube
A conducts *less* than it would 'alone'", no way can the balance be
improved when the two tubes move in opposite directions.
My original premise which you contradicted, without offering valid
evidence to the contrary, was that if one tube of a pair connected in a
garter circuit started conducting more, then the second tube would also
conduct more. That IS "balancing", the second tube conducting less as
you state above IS NOT "balancing".
If you have changed your mind and have now come to the conclusion that
the tubes "pulled toward each other", then that IS "balancing" and is
not at odds with my contention that the garter circuit causes both tubes
to conduct more when something changes in one tube causing it to conduct
more.
Ian Iveson
Bitching aside, each of you appears on the face of it to
have a coherent position. It could be that the inclination
to bicker has unfairly dismissed the possibility that you're
both right. Maybe there is a resolved contradiction in
reality, but you haven't got to the resolution part, and
typically for both of you, all of us maybe, you never will
unless you make an effort to stop the semantic chicanery.
Talking of which, I have a slight semantic problem with the
"if one tube conducts more" approach, because it assumes
from the start that the garter is not effective at balancing
the valves. Perhaps "if one tube is inclined to conduct
more" would be less confusing.
Anyway, I have asked the following two questions quite a few
times throughout history, without great success. Perhaps now
would be an appropriate time to raise them again.
Firstly, what is the best strategy of bias adjustment as a
valve ages? Should constant idle current be maintained, or
what? I guess this might depend on whether it's in AB or A.
Perhaps we could establish what's best for one valve in
class A, then two in PP class A, then two in PP AB1. Seems
to me the ideal would be different in each case. Anyway,
it's not easy to say what's better if we don't know what's
best.
Secondly, in what ways do valve characteristics change as
they age? I can most easily model an aged valve by simply
reducing its perveance. That is, if Ik = K.f(Va, Vg, Vk,
Vs), then I can easily change the value of K, which has the
effect of stretching or squashing the Y axis of the anode
characteristics. Incidentally, how would you simulate ageing
using a good valve in a real circuit? By reducing heater
current?
Anyway, if you can make unambiguous indicative predictions
about the behaviour of a given circuit in particular
circumstances, then I can try to post simulations of same.
I can see a problem that might be interesting to overcome
but I don't know if I have the time. A useful graph would be
one with some measure of current imbalance on the Y axis,
and some measure of difference in valve characteristics
along the X axis. Somehow, total current also needs to be
included, I guess. Trouble is, getting SPICE to do such a
graph directly is a bit of a challenge, coz it would involve
sweeping a parameter (such as K, above). Otherwise it could
be done with a spreadsheet using data from many simulations,
but that would be tiresome.
Finally, has anyone ever thought of using a CCS in parallel
with the OPT primary on one side of a PP circuit, so the
valves can be adjusted for best individual operating points,
with the CCS used to maintain balance by compensating for
the ensuing difference in currents?
Ian
Staring at old audio chassis as a hobby :)
I use the circuit. My amp sounds better than it did. Anyone can try it
for cheap :)
Nothing ventured, nothin gained.
I am to old and sick to argue philosophy ...
Happy Ears!
Al
.
Raymond Koonce
What Al says. I tried it and it improved the sound of the amp. What's
wrong with that?
Go solder.
Raymond
Staring at old audio chassis as a hobby :) wrote:
Patrick Turner
flipper wrote:
> Well, with a cursory look it seems that way but I don't think so and
> an equally casual observation is the circuit's symmetry so how does it
> know which tube to 'follow' and if FB is positive why doesn't it trot
> down with the low current tube instead of up?
>
> Let's do a more exhaustive analysis.
>
> First, take 'ideal' equal tubes and note that the voltages on both
> cathodes are equal and our 'half way' point, where the grids are cross
> connected, are also equal. Ok, cool, so let's reconnect the grids to
> the 'half way' point on the *same* side, removing the feedback.
> Voltages and currents remain the same.
>
> Now wave a magic wand over tube A and transform it into a 'more
> conducting' tube. I.E. we have an imbalance. It's cathode V rises and
> the 'half way' point voltage rises. So far this is as you expected.
>
> Now, swap tube B's grid from it's own 'half way' point back to tube
> A's 'half way' point, restoring the first half of the feedback loop.
> Current through tube B will increase from the higher grid voltage.
> again as you expected.
>
> However, it will not conduct as much as tube A because it's the 'less
> conductive' tube. So it's cathode will not rise as high as tube B's
> cathode nor will it's 'half way' point. Or, put another way, we still
> have an imbalance, but not as much because it's grid V was increased.
>
> However, now reconnect tube A's grid to tube B's 'half way' point,
> restoring the second half of the feedback loop. Tube A's grid V
> becomes *less*. I.E. the 'less conducting' tube B's 'half way' point
> is pulling down tube A... and to a current less that the 'no feedback'
> case we started from.
>
> And that part deviates drastically from your analysis because you see
> them both 'raising each other' but the problem is, 'relative to what'?
> And I believe that breaking the feedback, observing the 'balanced
> state', unbalancing the tubes, and then reconnecting answers that
> question. Tube B conducts more than it would 'alone' but tube A
> conducts *less* than it would 'alone' and the circuit does not 'run
> away' chasing the high current tube.
>
> >As an alternative to the garter let me propose the following scheme.
> >First use separate cathode resistors for each tube of twice the value
> >normally used for separate cathode resistors, the same as in the garter
> >circuit. This will result in lower than the desired current through the
> >tubes, so if the normal grid bias was 30 volts grid to cathode,
> >disconnect the grid resistors from ground and connect them to a divider
> >from B+ that will elevate the grids to 30 volts above ground. This will
> >restore the desired current through the tubes, while the doubled cathode
> >resistors will increase the stability of the tube currents with respect
> >to variations in the tube parameters, without the undesirable
> >interactions of the garter circuit.
>
> I would think the 'fixed grid' might be less of a restraint and worse
> balance because it doesn't have the other tube pulling the grid in
> opposition.
>
> I'm not sure, though, because it does have 'twice' the cathode
> feedback due to double the Rk. Might turn out they're equivalent with
> the 'half way' point simply providing the grid V you'd otherwise have
> to find a source for.
Obviously, quantifying the amount of bias regulation of a pair of garter
biased tubes could be done in terms of Ra, Rk, and gm.
But there is enough complexity in the interactivity between tube A and B
to dis-allow any ease of understanding.
I myself find that plain old cathode biasing works to keep Ek and hence
Ia within +/- 5% for a typical group of 4 tubes even when they have
aged well away from being a matched quad when new.
So I don't see the need for garter biasing.
Now when class AB begins after the inevitable first few watts of class
A1, then plain cathode biasing falls apart as the Ck charges up due to
rectifying currents into the RC cathode biasing network, just like a
power supply rectifier.
What is the situation with garter biasing and class AB?
I doubt its any better than plain cathode biasing.
Possibly, garter biasing offers a four fold increase in cathode bias
regulation, certainly with pure class A working.
First you have double the normal Rk per tube. This gives twice the
regulation. Then if A Ek rises, half that rise is applied to
grid B, so tube B has a rise of Ek, but not as much as tube A, and then
grid A sees half that change which only slightly helps Ia of tube A
increase a little. There is definately some PFB operative, which tends
to keep the two Ia closer than with plain cathode bias.
Consider when A goes open, or dead. Ek goes to 0V. Then so does grid B,
which reduces Ia in tube B a heck of a lot.
So the garter does offer protection against tube failure of one tube
when the other goes open because once the Ia has been drastically
reduced the Ea can rise a lot if the PSU is un-regulated, and the Pda of
the one live tube won't become excessive. And you'll soon hear the
distorton.
If tube A goes into thermal runaway like some naughty tubes do
sometimes, then up goes Ia and A Ek and half that hefty rise is aplied
to grid B, so Ia in B rises, lifting grid A making things worse, and you
can have BOTH tubes running with Ia too high and a then what would then
fail without active protection circuits might be the fragile OPT primary
winding which isn't designed to survive 4 times the idle current for
more than 5 minutes, or else the tube rectifier fails by getting too hot
and arcing over, and maybe an electrolytic goes to a short because of
excessive ripple current. This train of events certainly would have been
the way tube amps may have failed in Blumliens day when tube rectifiers
abd dodgy electros were normal.
Active protection and measures to counter Ek rise in A or AB amps are
dealt with at my website.
In 1935, to have active protection meant that you needed to have an
extra triode fed with a signal derived from the cathodes of each tube
and through a couple of vacuum diodes, like a primitive logic circuit.
When Ek rose to a sufficiently positive voltage, the triode would then
saturate with say 20mA and cause a sensitive and special latching relay
to open, thus turning off the mains supply.
To reset, unlatch the relay with a button press, and turn the amp back
on.
These days two silicon diodes and a sensitive gate SCR, C106D can be
used with resistance dividers formed with the Rk of each cathode bias
circuit.
A 12V auxilliary supply is needed. Much cheaper than the 1935 solution,
and very desirable and necessary imho, and I always fit active
protection in amps I sell because I cannot gurrantee any tubes for
longer than 90 days. I buy tubes in good faith from the sellers, but one
never knows if the girls on the production lines in Russia may have made
a botch of a job one morning after their husbands gave them a beating
the night before.
And then I cannot stop owners replacing tubes I have supplied with
others they think are better, eg, rare NOS Mullard EL34 instead of bog
standard EL34 by Sovtek. And I want my OPT and other parts to *always*
survive failing tubes when they malfunction at any time between now and
20 years time.
I think I am the **only maker** in the world who fits active protection
measures.
So tubes that won't bias properly are spotted and replaced before real
troubles ever begin.
Nobody should mind an occasional failing tube but replacing OPTs because
of such failures is completely unacceptable because of the wasteful
destruction and it antagonises tube amp owners.
>
> >Perhaps one of the simulators in the group could simulate this
> >configuration and compare it with the results from the garter circuit to
> >see which scheme provides better balance as tube characteristics vary.
Or anyone can set up a garter biased pair of EL34, then pull out one
EL34 and replace that with a 6L6GC.
Then compare the difference with standard cathode bias meant for a pair
of EL34, but then remove one EL34, and replace with the same 6L6GC.
This may tell you more than simulation.
Patrick Turner.
Ian Iveson
> I appreciate your attempt at comity but we can't both be
> correct when
> saying opposite things about tube A.
Don't lose heart: you may one day discover the dialectic.
Whatever is, is not.
Everything both is and is not.
Nothing can be, and not not be.
As for the two valves, they could both raise each other and
move towards each other at the same time. They could also
both reduce each other and move towards each other at the
same time. In either case they would be moving in the same
direction relative to their respective starting points, but
in opposite directions relative to each other.
It has not been established what is the best way to adjust
bias as valves age or go berserk, and without that
qualification, how can one balancing circuit be said to be
better than another? Better for what? How much better?
It could be that if you are interested only in the best
sound, then the best strategy would be to allow a miscreant
valve to run away, and even take the other with it, on the
grounds that a short innocent life is better than years of
dementure. If longevity in adverse circumstances is your
objective, as for Patrick, then the garter may not be the
most profitable bet, although I'm sure a half-decent
designer could add another circuit to manage total current.
But why should an amateur builder wish to prolong the agony
of listening to an amp with rubbish valves? Raymond built
it, and he's not going to sit cluelessly and watch it melt.
He'll use decent valves and check total current
occasionally.
OTOH, if you're grovelling for the patronage of congenitally
rich stupid fools who wish only to flaunt their shallow
pretentions of cultural pursuit without the inconvenience of
its actuality, you may feel the need to employ some
idiot-proof method of bias, and bugger the sound. They can't
tell the difference anyway.
>>Talking of which, I have a slight semantic problem with
>>the
>>"if one tube conducts more" approach, because it assumes
>>from the start that the garter is not effective at
>>balancing
>>the valves. Perhaps "if one tube is inclined to conduct
>>more" would be less confusing.
>
> It might help if you thread back and look at the various
> steps in my
> experiment that Byrns ungraciously snipped out. First it
> starts with
> the grid connected to the local 'half way' point and
> 'identical'
> (ideal) tubes. I.E. balanced, same resistances, but with
> no 'garter'
> connection.
>
> Then I wave a magic wand over one tube so that it
> "conducts more."
>
> That is not a failure of the garter circuit because it's
> not connected
> as a garter yet and the purpose is to establish what the
> 'no feedback'
> condition would be as a point of reference to compare with
> what the
> garter subsequently does to it when reconnected.
>
> Later I did use something similar to your choice of
> 'inclined' in
> referring to the 'alone' "preference."
John Byrns
flipper <fli...@fish.net> wrote:
> On Fri, 08 May 2009 18:51:55 -0500, John Byrns <byr...@sbcglobal.net>
> wrote:
>
> >In article <lps8059a251f0tcde...@4ax.com>,
> > flipper <fli...@fish.net> wrote:
> >
> >> On Fri, 08 May 2009 10:54:37 -0500, John Byrns <byr...@sbcglobal.net>
> >> wrote:
> >>
> >> >In article <n2r705l23tr4ijskt...@4ax.com>,
> >> > flipper <fli...@fish.net> wrote:
> >> >
> >> >> And that part deviates drastically from your analysis because you see
> >> >> them both 'raising each other' but the problem is, 'relative to what'?
> >> >> And I believe that breaking the feedback, observing the 'balanced
> >> >> state', unbalancing the tubes, and then reconnecting answers that
> >> >> question. Tube B conducts more than it would 'alone' but tube A
> >> >> conducts *less* than it would 'alone' and the circuit does not 'run
> >> >> away' chasing the high current tube.
> >> >
> >> >Your last sentence, if true, says that the believers in the garter
> >> >circuit are completely wrong in their belief that the garter circuit
> >> >improves balance!
> >>
> >> No it doesn't. It says both tubes are pulled toward each other. That
> >> IS 'balancing'.
> >
> >It says and I quote "Tube B conducts more than it would 'alone' but tube
> >A conducts *less* than it would 'alone'", no way can the balance be
> >improved when the two tubes move in opposite directions.
>
> Not only, of course, it can be, it is.
Yes, I must apologize about that, what you said was correct, I got tube
A and tube B mixed up at that point in my reading and writing, as I knew
was inevitable at some point when writing. I should have read that
sentence of yours more closely because I recognized that my initial
reading of it was at odds with the rest of your writing.
> Imagine the hypothetical 'alone' conductances
>
> Tube A 30mA
> Tube B 28mA
>
> An imbalance of 2mA
>
> Connect the garters and, as per my example,
>
> Tube A moves 'down' to 29.5 mA (I.E. less* than it would 'alone')
> Tube A moves 'up' to 28.5 mA. (I.E. more than it would 'alone')
>
> They are closer in balance due to 'moving in opposite directions'.
Yes, of course, but it depends on which tube is moving which way, as in
my reading error above. I am gratified to see that you can make similar
mistakes in your writing, as in using "Tube A" twice in the above
passage while neglecting "Tube B" entirely.
Now consider my proposed alternative to the garter circuit, which I call
the Faux CCS.
Imagine the hypothetical 'alone' conductances in the reference circuit.
Tube A 30mA
Tube B 28mA
An imbalance of 2mA
Connect the Faux CCSs and, as per my example,
Tube A moves 'down' to 28.9 mA. (I.E. less* than it was 'alone')
Tube B remains 'fixed' at 28.0 mA. (I.E. the same as it was 'alone')
They are closer in balance due to the high current tube moving towards
the norm.
> >My original premise which you contradicted, without offering valid
> >evidence to the contrary, was that if one tube of a pair connected in a
> >garter circuit started conducting more, then the second tube would also
> >conduct more.
>
> What I contradicted was your incorrect claim that Tube B then *also*
> raises Tube A even more.
And indeed Tube B does then *also* raise Tube A even more as I stated,
although I suppose it depends on what you consider as the starting or
reference point for the experiment so that we may both be right.
However the bottom line on the "Garter" circuit, which was my original
point, was that it compounds the problem by causing the second tube to
also conduct more when the first tube becomes more conductive for some
reason, your example above demonstrates this perfectly. My alternative
circuit eliminates this undesirable coupling between the two tubes by
driving each tube independently towards the desired operating current.
Patrick Turner
flipper wrote:
>
> On Mon, 11 May 2009 01:48:15 +0100, "Ian Iveson"
> <IanIves...@blueyonder.co.uk> wrote:
>
> >flipper wrote: A few comments below in reply to a long windy post from Ian.
If Ian had simply enaged with his simulator tool and set up a garter
within it he'd have had something useful to tell us.
If anyone else had pulled out their old breadboard amp which they use
for examining the basics and trying things out then we should have been
told by now just how a real Garter behaves and whatever good or bad
resides in the idea.
Its much clearer and better to cease the blather and take action in ones
hobby shed and solder something up to find outabout.
Patrick Turner.
> I am aware of the dialectic and I am also aware it's amusing but
> essentially useless gibberish, if one is trying to actually accomplish
> something, and especially so in engineering.
>
> >As for the two valves, they could both raise each other and
> >move towards each other at the same time. They could also
> >both reduce each other and move towards each other at the
> >same time. In either case they would be moving in the same
> >direction relative to their respective starting points, but
> >in opposite directions relative to each other.
>
> You see, the purpose was to illuminate how the circuit works, not to
> engage in endless speculation about what 'might be'.
>
> >It has not been established what is the best way to adjust
> >bias as valves age or go berserk, and without that
> >qualification, how can one balancing circuit be said to be
> >better than another? Better for what? How much better?
>
> The point of defining 'better' is well and good enough, if that's the
> purpose, but I wasn't engaged in a subjective analysis of 'better'. I
> was explaining how the circuit worked.
>
> >It could be that if you are interested only in the best
> >sound, then the best strategy would be to allow a miscreant
> >valve to run away, and even take the other with it, on the
> >grounds that a short innocent life is better than years of
> >dementure. If longevity in adverse circumstances is your
> >objective, as for Patrick, then the garter may not be the
> >most profitable bet, although I'm sure a half-decent
> >designer could add another circuit to manage total current.
> >
> >But why should an amateur builder wish to prolong the agony
> >of listening to an amp with rubbish valves? Raymond built
> >it, and he's not going to sit cluelessly and watch it melt.
> >He'll use decent valves and check total current
> >occasionally.
> >
> >OTOH, if you're grovelling for the patronage of congenitally
> >rich stupid fools who wish only to flaunt their shallow
> >pretentions of cultural pursuit without the inconvenience of
> >its actuality, you may feel the need to employ some
> >idiot-proof method of bias, and bugger the sound. They can't
> >tell the difference anyway.
>
> All in all a good example of how useless the 'dialectic' is since
> you've rambled on for a page and not added one thing to an
> understanding of how the garter circuit works.
Ian Iveson
> I am aware of the dialectic and I am also aware it's
> amusing but
> essentially useless gibberish, if one is trying to
> actually accomplish
> something, and especially so in engineering.
So on the strength of mere awareness you are able to make
such a sweeping proclamation? That would make you very
great, or very small.
My neighbour said the same, only yesterday funnily enough,
about science. His problem is that he's been brainwashed by
his Mullah.
Which raises the question of what brainwashing is.
He was led into a particular area of thought and then,
ruse-a-bye, the connection between that area and the rest of
thought was cut, a process known as "snapping". He has since
then been constrained to wander round his remote domain
forever, truly believing that the enlightenment it affords
him is all-encompassing: that's all the world there is, and
it all falls within that understanding. Ignorance, as
brainwashing cults cottoned on to, is bliss.
It was a delicate situation. Should I argue with him and
risk the consequences, or should I pander and patronise? His
kids are all messed up inside coz they learn science at
school and anti-science at home. I told him he's entitled to
his views but not to my respect. Next minute he's round my
house waving a big hammer at my head. Now he's gone. Police
turned up and carted him off, and his house is for sale.
Engineering schools should be forced to teach other stuff as
well, just so you don't come out thinking paltry analysis is
all there is.
Thinking that it's up to you what the discussion is about,
and that you have the wit to judge my contribution. Thinking
that the world is bounded by your little list of facts.
But I shan't argue or judge you in return. I've had enough
of hammers this week. I won't call your attention to the OP,
or the totality of the thread which is rather wider than the
breadth of your sad heart. If there is someone who loves
you, perhaps this might help:
http://www.cultinformation.org.uk/pdf/article_1.pdf
Is there anyone who doesn't already know how the garter
works? Is there anyone daft enough to actually go ahead and
simulate it? Or to think that its ability to maintain
balance is a matter of dispute?
I should have said "Well done, Raymond, that's a fine amp
you've built!", but I was held back by my penchant for
verdegris. Shame.
Well done, Raymond, that's a very fine amp you've built.
Ian
>
>>As for the two valves, they could both raise each other
>>and
>>move towards each other at the same time. They could also
>>both reduce each other and move towards each other at the
>>same time. In either case they would be moving in the same
>>direction relative to their respective starting points,
>>but
>>in opposite directions relative to each other.
>
> You see, the purpose was to illuminate how the circuit
> works, not to
> engage in endless speculation about what 'might be'.
>
>
>>It has not been established what is the best way to adjust
>>bias as valves age or go berserk, and without that
>>qualification, how can one balancing circuit be said to be
>>better than another? Better for what? How much better?
>
> The point of defining 'better' is well and good enough, if
> that's the
> purpose, but I wasn't engaged in a subjective analysis of
> 'better'. I
> was explaining how the circuit worked.
>
> All in all a good example of how useless the 'dialectic'
> is since
> you've rambled on for a page and not added one thing to an
> understanding of how the garter circuit works.
>
>>
Patrick Turner
flipper wrote:
>
> On Tue, 12 May 2009 08:44:08 GMT, Patrick Turner
> <in...@turneraudio.com.au> wrote:
>
> >
> >
> >flipper wrote:
> >>
> >> On Mon, 11 May 2009 01:48:15 +0100, "Ian Iveson"
> >> <IanIves...@blueyonder.co.uk> wrote:
> >>
> >> >flipper wrote: A few comments below in reply to a long windy post from Ian.
> >
> >If Ian had simply enaged with his simulator tool and set up a garter
> >within it he'd have had something useful to tell us.
> >
> >If anyone else had pulled out their old breadboard amp which they use
> >for examining the basics and trying things out then we should have been
> >told by now just how a real Garter behaves and whatever good or bad
> >resides in the idea.
> >
> >Its much clearer and better to cease the blather and take action in ones
> >hobby shed and solder something up to find outabout.
> >
> >Patrick Turner.
> >
> >
>
> Well, ok. I took some time and ran simulations for common Rk, separate
> Rks, Byrn's circuit, which I call doubled Rk-Fixed Bias, and the
> Garter.
>
> I used my original 6GK6 amp as the starting schematic with B+ 315V
> for the doubled up Rk case and 304.567V for the singles so idle
> current would be identical. In other works, I compensated for the
> single vs double Rk bias drop..
>
> Common Rk was 135 Ohms and all the others were 270 Ohms.
>
> The doubled Rk-Fixed Bias idle current came out off by 1uA because of
> rounding error when determining the equivalent 'half way' point fixed
> bias voltage.
>
> For imbalance I doubled the tubes on one side.
>
> Currents are anode, no screen. (through OPT)
>
> So, here's the data
> Nominal Bias with 2 Ideal, Equal, Tubes.
>
> Common Rk 35.560mA each, 71.120mA total
> Separate Rk 35.560mA each, 71.120mA total
> Doubled Rk-fixed Bias 35.559mA each, 71.118mA total
> Garter 35.560mA each, 71.120mA total
> 'Broken' garter 35.560mA each, 71.120mA total
>
> With double tube imbalance
>
> Normal Tube 'Double' Tube
>
> Common Rk 25.838mA 51.600mA
> Separate Rk 35.560mA 40.676mA
> Doubled Rk-fixed Bias 35.559mA 38.335mA
> Garter 36.950mA 38.938mA
> 'Broken' garter 35.560mA 40.462mA
>
> Balance Error Total mA
>
> Common Rk 25.762mA 77.438mA
> Separate Rk 5.116mA 76.236mA
> Doubled Rk-fixed Bias 2.776mA 73.894mA
> Garter 1.988mA 75.888mA
> 'Broken' garter 4.902mA 76.022mA
>
> Knowing how lousy the balance is for a common Rk I was sort of taken
> aback when I saw the total current being so close to the others. Makes
> sense when you think about it but no wonder a simple PS fuse doesn't
> do diddle for a runaway tube. Current shoots through the OPT but
> overall current isn't a terribly large bunch more.
Indeed, and the common Rk in Quad-II is one hell of an alchiles heel.
Instead of 70mA per tube, you can have 90mA and 50mA, with one cool tube
and one glowing red, and it can stay like this for years, and the daft
old buggers too lousy to buy new tubes don't seem to notice the terrible
sound.
>
> Another thing of note is there's not a huge difference in total
> current among the lot. The Doubled Rk-fixed Bias is the best in that
> regard, as we surmised, but it's only 2mA, less than 3%.
>
> The Garter was the best balance, even beating the double value Rk with
> fixed bias by a respectable 28%, but just separate Rks are an
> astronomical improvement over the common Rk.
With fixed bias, and adjustable for each side, the balance should be
perfect.
But it might not stay perfect.
>
> 'Broken' Garter is the 'starting point' circuit in the qualitative
> analysis I did. I.E. It is with the grids connected to the local 'half
> way' point so you can see what the tubes do without the garter
> feedback but all other parameters the same. Note that, as I described,
> the 'low current' tube's current increases when the Garter is
> connected while the 'high current' tube's current is less than it was
> without the feedback. They 'move towards each other'.
>
> Interesting to see that total current goes down rather than just
> exchanging one side for the other.
>
> This is probably good enough for a rough relative comparison but I
> wouldn't take it as representative of a real life errant tube because
> doubling the tube doubles gm and that's not the case with grid leak.
>
> As a side note, my current mirror keeps them within 5uA, with 'ideal'
> resistors and transistors. Or, in other words, the balance will be
> determined by the sense resistor precision and, to a lesser degree,
> transistor matching (because the emitter sense resistors reduce
> transistor Vbe dependency).
The trouble with current sinks for each cathode is that in AB operation
the Ek rises dramatiically.
So one might allow a 10% Ek rise then have zeners limit it. But that
isn't always too good because Ek might vary somewhat with equal Ik in
each tube. And if you have multigrids the Ik consists of Ia and Ig2, and
if Ig2 isn't equal between a pair of tubes then Ia won't be although Ik
will be.
I decided the best way was cathode bias with separate Rk&Ck and with a
high % of class A %.
I thn worked out that Ek could be stabilised by dynamically bypassing
the ac cathode currents and the THD/IMD then measueres just as well as a
fixed bias amp when well into class AB1. The ac bypassing network
including any old power transistor does not work at all while the amp
works in pure class A1. See the page at
http://www.turneraudio.com.au/300w-5-bias-stabilizer.htm
Patrick Turner.
Patrick Turner
flipper wrote:
>
> On Wed, 13 May 2009 11:05:17 GMT, Patrick Turner
> Well, that's consistent with the simulation I ran so maybe the 'double
> tube' approach isn't far off.
>
> >> Another thing of note is there's not a huge difference in total
> >> current among the lot. The Doubled Rk-fixed Bias is the best in that
> >> regard, as we surmised, but it's only 2mA, less than 3%.
> >>
> >> The Garter was the best balance, even beating the double value Rk with
> >> fixed bias by a respectable 28%, but just separate Rks are an
> >> astronomical improvement over the common Rk.
> >
> >With fixed bias, and adjustable for each side, the balance should be
> >perfect.
> >
> >But it might not stay perfect.
>
> If "the daft old buggers" don't buy tubes when glowing red what makes
> anyone think they'd bias the thing?
>
> Just saying, that's supposed to be the 'advantage' of self bias.
> Albeit virtually self defeating with a common Rk.
>
> >> 'Broken' Garter is the 'starting point' circuit in the qualitative
> >> analysis I did. I.E. It is with the grids connected to the local 'half
> >> way' point so you can see what the tubes do without the garter
> >> feedback but all other parameters the same. Note that, as I described,
> >> the 'low current' tube's current increases when the Garter is
> >> connected while the 'high current' tube's current is less than it was
> >> without the feedback. They 'move towards each other'.
> >>
> >> Interesting to see that total current goes down rather than just
> >> exchanging one side for the other.
> >>
> >> This is probably good enough for a rough relative comparison but I
> >> wouldn't take it as representative of a real life errant tube because
> >> doubling the tube doubles gm and that's not the case with grid leak.
> >>
> >> As a side note, my current mirror keeps them within 5uA, with 'ideal'
> >> resistors and transistors. Or, in other words, the balance will be
> >> determined by the sense resistor precision and, to a lesser degree,
> >> transistor matching (because the emitter sense resistors reduce
> >> transistor Vbe dependency).
> >
> >The trouble with current sinks for each cathode is that in AB operation
> >the Ek rises dramatiically.
>
> You do this every time I mention the current mirror: go off into a
> discussion of CCS when I keep telling you a current mirror is NOT a
> CCS.
>
> I know CCS doesn't work well AB and it's intuitively obvious when one
> realizes tube current, unless constrained, is not 'constant' going
> into B. That means the CCS forces the tube out of normal bias when
> driven past Class A.
>
> The current mirror doesn't work that way. One side is plain ole
> cathode bias with a current sense transistor inside. The other side is
> a *variable* current sink that sinks the same current as the cathode
> bias side.
>
> That means as the cathode bias side's current increases so does the
> current sink on the second side and if the tubes are matched it would
> be identical. If they are not matched it makes them look matched, as
> if both were plain ole cathode bias. Except they're current matched.
>
> Now, we know that cathode bias is not prefect either in AB but it's
> better than dual current sinks and has the advantage of no adjustment
> by "the daft old buggers" needed. And with the 'cathode bias' current
> mirror balance is dern near perfect.
Cathode bias is sure very imperfect when you have low class A and high
class AB amounts in the power character.
And with a current mirror on one tube to ensure Iadc remains equal no
matter what, then the same problem in class AB exists as with separate
Rk and Ck. Its not as bad as with a CCS on one side with a current
mirror, or two CCS.
I prefer high class A% which means cathode bias is all you'd need, and
then have something to adjust the balance with visual indication.
See the schematics at the page on revised Quad-II amps at
http://turneraudio.com.au/quad2powerampmods.htm
The last effort to instil life into a pair of Quad-II was the one there
with KT90 outputs and a pair of red LED which indicates bias balance and
which also tells you other things.
>
> >So one might allow a 10% Ek rise then have zeners limit it. But that
> >isn't always too good because Ek might vary somewhat with equal Ik in
> >each tube. And if you have multigrids the Ik consists of Ia and Ig2, and
> >if Ig2 isn't equal between a pair of tubes then Ia won't be although Ik
> >will be.
> >I decided the best way was cathode bias with separate Rk&Ck and with a
> >high % of class A %.
> >I thn worked out that Ek could be stabilised by dynamically bypassing
> >the ac cathode currents and the THD/IMD then measueres just as well as a
> >fixed bias amp when well into class AB1. The ac bypassing network
> >including any old power transistor does not work at all while the amp
> >works in pure class A1. See the page at
> >http://www.turneraudio.com.au/300w-5-bias-stabilizer.htm
>
> Yes, I know. I've seen it.
>
> I'm just not convinced the extra complexity is needed as long as
> there's sufficient Class A because the transient Class B doesn't shift
> bias much.
Correct, but with dynamic bias stabilization you can run the tubes with
less Ia, as one can with fixed bias, and still achieve good distortion
figures.
With less bias current, Rk can be a larger value than used for pure
class A, so the larger Rk has more Ia regulating effect than a low value
Rk. Pda in a 6550 need only be 15W instead of 30W.
In my 300W amps there are 12 output tubes.
Having 12 pots for bias adjustment just for one channel is a royal PITA.
But you have to make it fool proof and reliable, so it has to be cathode
bias, and for reliability the Ia is low, and Pda is low.
The tubes last longer this way and the music is excellent.
Patrick Turner.
> >
> >Patrick Turner.
Patrick Turner
flipper wrote:
>
> On Thu, 14 May 2009 08:28:03 GMT, Patrick Turner
> Which is why I specifically said "as long as there's sufficient Class
> A" down below.
>
> Exactly what constitutes the magic "sufficient" value isn't 100% clear
> and it's interesting to observe that a simple idle Ia calculation
> underestimates the transition point because the tubes don't go rigidly
> into cutoff, and that tends to create an average current near idle
> longer than a simple Ia calculation would suggest.
>
> The effect is much more pronounced with triodes.
The usual rule of thumb is to have 50% in pure class A with the rest in
class AB for the rated load.
This means that if you connect a higher load there is less PO total but
more class A and less AB and if you connect a lower load there is more
PO total with less class A and more AB.
I like to set up a PP pair of KT88/6550 in UL to make 60W into 3 ohms,
AB1, absolute max, with only a few watts of class A1. By the time you
get to 8 ohms more than 1/2 the PO is class A, but you only get 34W
total, but the sound is dreamy.
Fuctards at many mass made hi-end factories will never addopt such
quaint ways to set up their amps. Absolute max PO is at 8 ohms when 8
ohms is connected to the 8 ohm OPT tap. Connecting 3 ohms would destroy
their bloody amps.
And when the idle current is low, and class AB amount high with a low RL
value, the tubes are not in their linear region even while in the small
amount of class A. But the current THD is mainly even order and there is
good cancelling, so you get away with it except the THD is 3 times more
than if you stick to the rule of 50% class A out of the possible total.
And the loads each tube see while in class A can be different if the
tubes gn is different, so when you draw the loadline for a single tube
in a class AB amp it is a curved line with a slope starting off at about
1/2 RLa-a then swinging up to 1/4 RELa-a after one tube has cut off. The
transition is a kinked line with beam and pentodes, but more swayed with
UL and triode so the odd order THD due to switching or crossover
distortion becomes worst with pure beam or pentode op in class AB.
>
> >And with a current mirror on one tube to ensure Iadc remains equal no
> >matter what, then the same problem in class AB exists as with separate
> >Rk and Ck.
>
> Yes, it acts like a 'super duper' separate Rk/Ck.
>
> > Its not as bad as with a CCS on one side with a current
> >mirror, or two CCS.
>
> Quite a bit better, in fact.
Yeah, should be.
>
> >I prefer high class A% which means cathode bias is all you'd need, and
> >then have something to adjust the balance with visual indication.
> >See the schematics at the page on revised Quad-II amps at
> >http://turneraudio.com.au/quad2powerampmods.htm
> >The last effort to instil life into a pair of Quad-II was the one there
> >with KT90 outputs and a pair of red LED which indicates bias balance and
> >which also tells you other things.
>
> Why bother with 'indicators' when the current mirror simply 'makes' it
> balanced?
I like to give the owner a visual indication of the bias condition, or
when the Ek of each tube begins to wobble unevenly due to high power
being used, or when a wrong load is connected.
In the sample of Quad-II at my site where I used a pair of KT90 instead
of the original KT66, the bias balance indicator LEDs rarely ever
change brightness even when fairly insensitive 4 ohms speakers are
connected to the 8 ohm tap setting. And I have slightly less Ia in each
output tube than originally specified for the KT66.
There's another sample pair of Quad-II with KT88 in triode and with
fixed bias, and I got a very nice 20W.
The balance had hardly moved since 1998 and the same Sovtek tubes have
been used every day by the owner. He now runs fairly sensitive Chinese
speakers so 20W is more than enough.
>
> >> >So one might allow a 10% Ek rise then have zeners limit it. But that
> >> >isn't always too good because Ek might vary somewhat with equal Ik in
> >> >each tube. And if you have multigrids the Ik consists of Ia and Ig2, and
> >> >if Ig2 isn't equal between a pair of tubes then Ia won't be although Ik
> >> >will be.
> >> >I decided the best way was cathode bias with separate Rk&Ck and with a
> >> >high % of class A %.
> >> >I thn worked out that Ek could be stabilised by dynamically bypassing
> >> >the ac cathode currents and the THD/IMD then measueres just as well as a
> >> >fixed bias amp when well into class AB1. The ac bypassing network
> >> >including any old power transistor does not work at all while the amp
> >> >works in pure class A1. See the page at
> >> >http://www.turneraudio.com.au/300w-5-bias-stabilizer.htm
> >>
> >> Yes, I know. I've seen it.
> >>
> >> I'm just not convinced the extra complexity is needed as long as
> >> there's sufficient Class A because the transient Class B doesn't shift
> >> bias much.
> >
> >Correct, but with dynamic bias stabilization you can run the tubes with
> >less Ia, as one can with fixed bias, and still achieve good distortion
> >figures.
>
> It's been my observation that distortion is less with higher Ia even
> when bias is held rigid (I.E. fixed bias) so I'm not convinced the
> lower Ia with 'bias stabilization' is an improvement in the Class A
> region.
The slightest transient above the class A limit will be bypassed. Ek
then does not move even slightly.
>
> One would think, however, the 'stabilized bias' is 'some degree
> better' during Class B transients but 'how much' is difficult to
> quantify and depends on how much one small shift vs the other small
> shift is. On the other hand, it may be swamped by the lower Ia if the
> purpose is to lower idle.
>
> >With less bias current, Rk can be a larger value than used for pure
> >class A, so the larger Rk has more Ia regulating effect than a low value
> >Rk. Pda in a 6550 need only be 15W instead of 30W.
> >
> >In my 300W amps there are 12 output tubes.
> >
> >Having 12 pots for bias adjustment just for one channel is a royal PITA.
> >
> >But you have to make it fool proof and reliable, so it has to be cathode
> >bias, and for reliability the Ia is low, and Pda is low.
> >
> >The tubes last longer this way and the music is excellent.
>
> I'm sure it is. The question would be if there's another means to
> achieve essentially the same end results.
There isn't any other way to achieve stable cathode biasing when you run
up the signal to clipping with a sine wave in AB.
>
> I did play with your circuit but found it difficult to adjust the
> 'current bypass' for minimum shift, and that was with 'ideal'
> components (simulations).
Build a real circuit and play with values. Then you'll soon find out how
to change values a bit to allow a slight rise in Ek but of less than 10%
of the idle value, and between idle and the +10% the bias rises almost
as a straight line, regardless of output load.
> That doesn't mean it 'fails' but that the
> degree of 'improvement' may not, in real life, be as much as one
> predicts from the basic theory.
Its possible to set up the bypass transistors so that Ek falls a bit
before rising just before clipping and this means your'e bypassing too
much current, ie, the transistors are letting too much current drain out
of each Ck.
If you examine the wave form at the top of each Ck with LF with normal
cathode biasing, its an ugly picture of non-linear waves and a big phase
shift. But with the dynamic bypasser, the waves at the top of the caps
have far less amplitude. The transistors merely provide a path to bypass
the same ac currents that would otherwise charge up the Ck.
The transistors therefore improove the whole AB operation.
>
> One should also note that my circuit and yours are intended to solve
> different problems. I was trying to get near perfect balance
> (succeeded) and 'bias stabilization' doesn't help in that regard.
> There were also serious problems in trying to meld the two so I opted
> for the less complex approach. At least so far.
The bias stablization works with more than 2 output tubes where you wish
to get best AN performance but with no hassles of lots of interactive
bias adjust pots.
As soon as more than two output tubes are used you have to ensure that
Iadc for each tube is fairly well regulated, and the only way to know
all the tubes are OK and happy is to have some visual means where their
Ek across a set Rk is compared to a reference voltage and a read out is
given with a pair of LED of different colour, or a single bi-colour LED.
When I re-wired a bleedin awful VT100 18mths back I got rid of the
terrible direct coupled drive to the 8 output 6550 which had only one
bias voltage per channel of 4 x 6550, and I went to plain RC coupling
from a normal LTP with a 12BH7 each side, then had 8 bias pots. BUT, to
make things easy for the owner there is a green-red bi-colour LED near
each bias pot on each side panel of the box and as long as the tubes
have a very slightly green LED all is well and bias is within +/- 4mA.
This bitch of an amp kept blowing tubes up and blowing mains fuses. ARC
should be ashamed of their idiocy.
But I have not heard from the owner in 18mths so I know all is well
after my drastic mods to entirely re-design these amps.
Someone phoned me the other night about re-enginering a pair of
Reference 600W mono with 16 x 6550 in each channel. Same old story about
terrible reliability and fuses and smoke and bad sound.
If that job comes off, I will dump the entire existing innards into the
bin, and start all over again.
I really fail to see why anyone would ever need to have 600W for each
channel for a domestic hi-fi system.
So it would be prudent of me to try to set up the tubes in cathode bias
which would reduce the effective Ea. The existing loading will become
more suited, and there is provision for various loads, and if anyone
always only ever uses the 4 ohm outlet then the PO total of say 350W
with 16 x 6550 when an 8 ohm speaker is used will have 100W of class A
content at least.
So because so much PO is available, there isn't any need for dynamic
bias stabilization.
Asking 600W from 16 output tubes is the same as asking 75W from just a
pair. McIntosh and others try to do it. Its Bullshit. One should never
ask for more than 60W, and that should only be possible at a load
slightly below the lowest one is allowed to use, ie, about 3.2k a-a when
the Ea and Eg2 for UL is at 500V. But I've seen may makers who like to
have 3.2k : 4, 8, and 16 ohms.
One should never ever use the 8 or 16 connections and ONLY ever use the
4 ohm tap which then makes an 8 ohm speaker give a load to the tube of
6.4k a-a, which is far better than 3.2k.
The only garter that I might see will be on the leg of the dancing girl
after I get paid for the work.
Patrick Turner.
Patrick Turner
flipper wrote:
>
> On Fri, 15 May 2009 11:58:53 GMT, Patrick Turner
> Yes, you've said that before but my datasheets still show 3, or more,
> to 1 in the 'typical' cathode bias applications.
>
> I didn't write 'em. I just read 'em.
I question what I read, and if you have any PP circuit with a certain
Pda total at idle for both tubes, the maximum class A1 possible is about
45% of that power in audio PO if the tubes are beam or pentodes. There
is less PO with triodes unless you do something like set them up in BT.
So even the McIntosh circuit with a pair of 6550 for 75W AB2 will have
Ea at 470V and Ia at say 30mA per tube. Thus Pda total at idle is 28.2W
and with efficiency at 45% you'd get 12.6W of class A maximum but only
if you make the load seen by the tubes far higher than that used to get
the 75W, which is 4k a-a for the 2 x 6550.
>
> >This means that if you connect a higher load there is less PO total but
> >more class A and less AB and if you connect a lower load there is more
> >PO total with less class A and more AB.
> >
> >I like to set up a PP pair of KT88/6550 in UL to make 60W into 3 ohms,
> >AB1, absolute max, with only a few watts of class A1. By the time you
> >get to 8 ohms more than 1/2 the PO is class A, but you only get 34W
> >total, but the sound is dreamy.
> >
> >Fuctards at many mass made hi-end factories will never addopt such
> >quaint ways to set up their amps. Absolute max PO is at 8 ohms when 8
> >ohms is connected to the 8 ohm OPT tap. Connecting 3 ohms would destroy
> >their bloody amps.
>
> Yes, I know. You're the only 'non fucktard' left on the planet.
I dunno, maybe there are actually many non FTs out there lurking.
> Well, your two examples negated the 2 'reasons' you gave.
> That has nothing to do with the comment I made about Class A
> distortion vs Ia.
>
> >> One would think, however, the 'stabilized bias' is 'some degree
> >> better' during Class B transients but 'how much' is difficult to
> >> quantify and depends on how much one small shift vs the other small
> >> shift is. On the other hand, it may be swamped by the lower Ia if the
> >> purpose is to lower idle.
> >>
> >> >With less bias current, Rk can be a larger value than used for pure
> >> >class A, so the larger Rk has more Ia regulating effect than a low value
> >> >Rk. Pda in a 6550 need only be 15W instead of 30W.
> >> >
> >> >In my 300W amps there are 12 output tubes.
> >> >
> >> >Having 12 pots for bias adjustment just for one channel is a royal PITA.
> >> >
> >> >But you have to make it fool proof and reliable, so it has to be cathode
> >> >bias, and for reliability the Ia is low, and Pda is low.
> >> >
> >> >The tubes last longer this way and the music is excellent.
> >>
> >> I'm sure it is. The question would be if there's another means to
> >> achieve essentially the same end results.
> >
> >There isn't any other way to achieve stable cathode biasing when you run
> >up the signal to clipping with a sine wave in AB.
>
> The mentioned "end results" is not to 'achieve stable cathode bias,
> it's to achieve a good sounding amplifier.
Stable bias does lead to better sound though.
>
>
> >> I did play with your circuit but found it difficult to adjust the
> >> 'current bypass' for minimum shift, and that was with 'ideal'
> >> components (simulations).
> >
> >Build a real circuit and play with values. Then you'll soon find out how
> >to change values a bit to allow a slight rise in Ek but of less than 10%
> >of the idle value, and between idle and the +10% the bias rises almost
> >as a straight line, regardless of output load.
>
> Thanks but it doesn't deal with the goal of automatic balance.
Well how does one balance a dozen output tubes?
One way might be to have one tube with a real Rk and the remaining 11
with current mirrors.
But that doesn't stop all the Eks rising together when you move into AB.
So the amp become mis-biased while in AB, or like an underbiased amp, so
there will be MUCH more 3H produced because the amp starts to resemble a
class B amp with heavy crossover distortions.
>
> >> That doesn't mean it 'fails' but that the
> >> degree of 'improvement' may not, in real life, be as much as one
> >> predicts from the basic theory.
> >
> >Its possible to set up the bypass transistors so that Ek falls a bit
> >before rising just before clipping and this means your'e bypassing too
> >much current, ie, the transistors are letting too much current drain out
> >of each Ck.
> >If you examine the wave form at the top of each Ck with LF with normal
> >cathode biasing, its an ugly picture of non-linear waves and a big phase
> >shift. But with the dynamic bypasser, the waves at the top of the caps
> >have far less amplitude. The transistors merely provide a path to bypass
> >the same ac currents that would otherwise charge up the Ck.
> >The transistors therefore improove the whole AB operation.
>
> I understand the goal but simply 'bypassing' current peaks does not
> result in a constant average current into the filter caps because tube
> current also goes to 0.
>
> Moot anyway because it's incompatible with automatic balance.
I think I probably could add some means of having Ik of 2 tubes
equalised, AND have the dynamic as shunting by the bjts and resistors.
>
> >> One should also note that my circuit and yours are intended to solve
> >> different problems. I was trying to get near perfect balance
> >> (succeeded) and 'bias stabilization' doesn't help in that regard.
> >> There were also serious problems in trying to meld the two so I opted
> >> for the less complex approach. At least so far.
> >
> >The bias stablization works with more than 2 output tubes where you wish
> >to get best AN performance but with no hassles of lots of interactive
> >bias adjust pots.
>
> Cathode mirror works with as many tubes as you like but 'bias
> stabilization doesn't auto balance like the current mirror does.
True.
>
> >As soon as more than two output tubes are used you have to ensure that
> >Iadc for each tube is fairly well regulated, and the only way to know
> >all the tubes are OK and happy is to have some visual means where their
> >Ek across a set Rk is compared to a reference voltage and a read out is
> >given with a pair of LED of different colour, or a single bi-colour LED.
>
> Another way is if they are automatically constrained to be balanced,
> like with a current mirror.
>
> >When I re-wired a bleedin awful VT100 18mths back I got rid of the
> >terrible direct coupled drive to the 8 output 6550 which had only one
> >bias voltage per channel of 4 x 6550, and I went to plain RC coupling
> >from a normal LTP with a 12BH7 each side, then had 8 bias pots. BUT, to
> >make things easy for the owner there is a green-red bi-colour LED near
> >each bias pot on each side panel of the box and as long as the tubes
> >have a very slightly green LED all is well and bias is within +/- 4mA.
> >This bitch of an amp kept blowing tubes up and blowing mains fuses. ARC
> >should be ashamed of their idiocy.
> >But I have not heard from the owner in 18mths so I know all is well
> >after my drastic mods to entirely re-design these amps.
> >
> >Someone phoned me the other night about re-enginering a pair of
> >Reference 600W mono with 16 x 6550 in each channel. Same old story about
> >terrible reliability and fuses and smoke and bad sound.
> >
> >If that job comes off, I will dump the entire existing innards into the
> >bin, and start all over again.
> >
> >I really fail to see why anyone would ever need to have 600W for each
> >channel for a domestic hi-fi system.
>
> Well, you could read Phase Linear's white paper on why you "have to
> have" 400 - 800 watts.
I could indeed read white papers about Phlame Linear amps.....
But I'm riding my bicycle tomorrow which is far more interesting.
One reason people keep saying you need 500 times more power instead of
only 50 times more than your average PO is that its profitable for those
promotong hugh power, Tell 'em what they don't need convincingly, and oh
how the money rolls in.
Patrick Turner.
Ian Iveson
>>Is there anyone who doesn't already know how the garter
>>works? Is there anyone daft enough to actually go ahead
>>and
>>simulate it? Or to think that its ability to maintain
>>balance is a matter of dispute?
>>
>>I should have said "Well done, Raymond, that's a fine amp
>>you've built!", but I was held back by my penchant for
>>verdegris. Shame.
>>
>>Well done, Raymond, that's a very fine amp you've built.
>
> While you have been wandering through a dialectic jungle
> pondering the
> metaphysical inferiority of others and curricula you've
> never seen I
> have finished simulations of the garter and 3 other models
> showing the
> relative performance and merits of each.
You're *very* rude, but good for you all the same, although
I can't see what you needed to simulate, and the aspects you
compare are limited, to say the very least. Were you
surprised by your findings?
> In my world this is called using the right tool for the
> job or, in the
> colloquial, there's a time and place for everything.
And the time for everything is always now. Why limit your
understanding to only something?
> To wit, falling through the sky after jumping out of an
> airplane is
> not the time to be pondering the dialectic of parachutes
> and rip
> cords.
But without the dialectic there would be no parachutes, or
aeroplanes. A wise man thinks *before* he jumps. If you
really need to think after you've set off, *only* the
dialectic will help, because you need to do the opposite of
whatever you're doing when you discover your analytical
approach has gone awry.
Ian
Patrick Turner
flipper wrote:
> >> >> >> >>
> >> >> >> >> On Tue, 12 May 2009 08:44:08 GMT, Patrick Turner
> >> >> >> >> <in...@turneraudio.com.au> wrote:
> >Stable bias does lead to better sound though.
>
> Then use fixed bias.
But in an amp with 12 utput tubes, all the bias pots are a real hazard.
People adjust them wrong....
Two tubes in one amp, yeah, no problem, but with 12 tubes? no way man.
>
> Once on the self promotional high horse it's virtually impossible to
> have a rational discussion with you because you never talk about the
> damn topic.
Its been talked to death on the news group before.
>
> To wit, this whole thread is in response to my comparison of the 'self
> balance' in 4 self bias cathode circuits, and a fifth counting the
> current mirror comment, but you've decided to flat ass ignore self
> balance and speak of nothing but your beloved 'bias stabilization'
> circuit.
I raised the idea of dynamic bias reg because all manner of self
balancing systems such as garter and current mirrors and fidling with
grid bias voltages all fail to perform well when inevitably someone uses
a low value load on a PP amp which is designed to work in only pure
class A.
>
> That's nothing against 'bias stabilization' but it WASN'T the blooming
> TOPIC.
Agreed entirely. But signal current shunting as I do it in AB amps works
a treat to prevent Ek bias drift. Balance is only one issue about
biasing any PP amp correctly. The amp should if possible be made immune
to heavy class AB signals cause bias drift. in aclass A amp with a load
that's too low and too much volume, the bias drifts unevenly. Where you
have a pair of LED to indicate bias imbalance of +/-10%, you will see
the LEDs flash one side then the other as the bias currents vary
somewhat a lot. It is almost impossible to stop this sort of thing
unless you stop the cause of it, which is momentary differences in the
charge of each Ck on each side of thr PP circuit. So the ac shunting i
do helps get rid of all the wobblies.
>
> >> >> I did play with your circuit but found it difficult to adjust the
> >> >> 'current bypass' for minimum shift, and that was with 'ideal'
> >> >> components (simulations).
> >> >
> >> >Build a real circuit and play with values. Then you'll soon find out how
> >> >to change values a bit to allow a slight rise in Ek but of less than 10%
> >> >of the idle value, and between idle and the +10% the bias rises almost
> >> >as a straight line, regardless of output load.
> >>
> >> Thanks but it doesn't deal with the goal of automatic balance.
> >
> >Well how does one balance a dozen output tubes?
> >
> >One way might be to have one tube with a real Rk and the remaining 11
> >with current mirrors.
>
> Yep.
>
> >But that doesn't stop all the Eks rising together when you move into AB.
> >So the amp become mis-biased while in AB, or like an underbiased amp, so
> >there will be MUCH more 3H produced because the amp starts to resemble a
> >class B amp with heavy crossover distortions.
>
> Not if, as has been repeated a half dozen times, you run enough Class
> A so transients into Class B become insignificant.
True in most instances, but with real music at high levels the Ek bias
wobbles considerably, especially with low power amps and low sensitivity
speakers at high levels.
>
> >>
> >> I understand the goal but simply 'bypassing' current peaks does not
> >> result in a constant average current into the filter caps because tube
> >> current also goes to 0.
> >>
> >> Moot anyway because it's incompatible with automatic balance.
> >
> >I think I probably could add some means of having Ik of 2 tubes
> >equalised, AND have the dynamic as shunting by the bjts and resistors.
>
> Might be an interesting project. Give it a shot.
When you have 12 output tubes, its difficult to decide what you want to
balance. Its easy to balance a see saw, but much more difficult to
balance a ferris wheel. Too many people.
But one can have 12 equal dc current sinks, one for each cathode except
that you have to make up a board with 12 darlington connected bjts and
their emitter Re, and have a voltage reference for the bases, and all
that's a pain when all you really need is a simple humble resistance. If
there is inequality of Iadc between the 12 tubes its likely it will even
out for the two sides of the PP circuit.
That's what I have found in amps with fixed bias and 6 tubes. The drift
a bit out, or they have different gm, and to eliminate 2H in the output
you sometimes have to swap tubes around after measuring signal currents
in each tube and summing the results for each side of the PP circuit.
The current sink causes no increase in Ik if a tube decides to conduct
more Ia, or Ig2. With an Rk, if Ek rises from say 30V to 60V, at least
you know there's twice the Ik, and most probably twice Ia and Ig2. The
Rk tends to limit Ek rise. But there isn't any limiting if The tube goes
wonky and the bias voltage loses control of the tube *voltages* so you
need to have some good active protection. I decided plain Rk provided
enough regulation and then decided the active protection would stop
runaway tubes, and the dynamic bypass would allow easy carefree biasing
for the owner who could then drive the amp as hard as he wanted to with
all the benefits of fixed bias, ie, very little bias *voltage* drift
caused by rectifying currents charging up Ck, which BTW are each 1,000uF
in my 300wattes with 12 output tubes.
If Ek drifts with such caps you don't want it to stay drifted for too
long, hence the need for the active bypassing.
>
> I've already done a current mirror 'fixed bias' version in which there
> is no bias shift, and it doesn't suffer RC filter ripple like when
> using a differential balance amp, but I haven't been keen on using it
> because fixed bias negates the 'idiot proof' self bias aspect.
I'm not sure what you mean. Fixed bias means tying the cathodes to 0V
without a series parallel R&C network to which you can attatch various
schemes to fiddle with the current and voltages.
>
> I have been wondering, though, if something akin to a Brook bias could
> be added to it since there's no RC time constant associated with the
> current mirror and it's multiple time constants that aggravate bias
> stability problems when trying to 'do both' at the same time.
Hm, the eternal bias problem is to have the Idc balanced and the voltage
bias about equal.
Its impossible to achieve perfection because tubes ain't perfect
especally after they gae a bit. Gm changes, and a different Eg1 is
required for each tube to get the same Ik.
>
> On the wild hair side, I've long been curious about making a little
> bias micro controller that does something seemingly simple like adjust
> fixed bias at turn on and just hold it there till next time it can
> check but that's not a trivial project and "seemingly simple" rarely
> is.
Just connect the amp to a PC via a USB wire and have a program.
Don't let the B+ mix with the mother board. The B+ always wins, easily,
with a complementary puff of smoke.
>
> If a company should read this, hire me and I'll tell you how to do it.
Companies know how to stay well away from here. I think we are alone.
> >One reason people keep saying you need 500 times more power instead of
> >only 50 times more than your average PO is that its profitable for those
> >promotong hugh power, Tell 'em what they don't need convincingly, and oh
> >how the money rolls in.
>
> Yes, of course. It's always a conspiracy and especially when one
> doesn't bother to read the rationale.
>
> I'd tell you but I already have and got tired of hearing you explain
> that the typical 5 Watt amp is being run 50mW average Po.
The typical 5W amp with say a lone 6BQ5 will battle to give a clean
sound into average speakers or 89dB/W/M.
So folks with such amps tend to use old fashioned speakers with
sensitivity over 95dB/W/M, and then the 50mW might be enough if you work
it all out.
I might use about 1 watt average sometimes. It means 89dB at 1W at 1M,
and in aroom with two speakers that's quite loud, too loud for me most
days, so 500mW will be about right. I might also use a 50W amp.
I don't need to be talked into using a 500W amp.
I do know a guy with giant JBL monitors with 2 x 15" bass units and
horned mid/treble units. He has a pair of Yamaha 2200 amps to provide
the power, less than 0.25W most days in his tiny room. Those amps are
capable of 200W per channel, and he has 800W system capability.
That's an excess capability of around 750W at least.
Patrick Turner.
Patrick Turner
flipper wrote:
>
> On Sat, 16 May 2009 10:49:37 -0500, flipper <fli...@fish.net> wrote:
>
> >I've already done a current mirror 'fixed bias' version in which there
> >is no bias shift, and it doesn't suffer RC filter ripple like when
> >using a differential balance amp, but I haven't been keen on using it
> >because fixed bias negates the 'idiot proof' self bias aspect.
> >
> >I have been wondering, though, if something akin to a Brook bias could
> >be added to it since there's no RC time constant associated with the
> >current mirror and it's multiple time constants that aggravate bias
> >stability problems when trying to 'do both' at the same time.
> >
>
> Ok, I got back on the current mirror thing.
>
> To recap, the current mirror is essentially a bypassed RK on one side
> with a current mirror under it so the second side is forced to match.
> Simple an effective for balance.
>
> The above mentioned 'fixed bias' current mirror is simply replacing
> the Rk with an upside down P-MOSFET with 'fixed' gate that holds the
> cathode at that V (plus gate-source bias). Current then flows into the
> current mirror as with the self bias version. Balance is just as good
> with no rising Vk during Class B but it's no longer self biasing.
>
> Re-enter the old Brook bias.
>
> If we recall, Brook bias works by observing that the minimum of the
> summed currents through both tubes equals idle x2 so, in essence, we
> build a negative peak detector, filter, and use that signal to bias.
> This can be easily done using a transistor (or tube) as a detector
> against a constant reference voltage (setting desired Ia) and the
> output then discharges a cap that is also being pulled high. Discharge
> is much quicker than discharge so the cap value settles to whatever
> voltage (plus some ripple) produces the correct bias.
>
> Well, it's a simple matter to feed that bias voltage to the P-MOSFET's
> gate rather than a fixed bias.
>
> So, instead of taking the current mirror 'bottom side' to ground we
> run it through a 10 ohm resistor that senses the current through both
> tubes. That signal goes into the Brook bias and the Brook output goes
> to the P-MOSFET.
>
> It may seem like that adjusts only one tube but since the current
> mirror forces the other to follow it sets both.
>
> One hitch. The current mirror bypass caps also go into the sense
> resistor, as they should so we see the current, but the other side is
> at a fixed voltage. That causes excessive currents as one end bounces
> up and down due to the sense resistor voltage against the fixed bias.
> That's solved by putting another upside down P-MOSFET between the
> sense resistor and current mirror so the current mirror sees a fixed
> voltage on both ends.
>
> I did some rough measurements with the 807 amp I'm thinking of
> building so the numbers are a bit different than the other test since
> Ia is 48mA rather than 30mA.
>
> At any rate, balance with 'ideal' tubes is within 80uA and 300uA in
> the 'double tube' test. At the same time, auto bias increases about
> 700uA. Not bad.
>
> Back to 'ideal' tubes. Bias is about 39.38V static and shifts upward
> about 120mV at 1 Watt Po. At 40 Watts Po it shifts upward about 700mV.
>
> It's not as easy to visualize what bias does in the 'double tube' test
> because, as I previously mentioned, gm is also doubled so the signal
> itself causes a further 'imbalance' as one tube tries to do more work
> than the other, with the result being bias to the 'strong' tube is
> increased while bias to the 'weak' tube is reduced. However, if one
> looks at the two Ia signals you see the bias 'gap' remain constant
> while the 'work' is re balanced between the two.
>
> I wouldn't call this a 'finished' design. For one, I just ran
> simulations and there are some refinements that might be in order,
> like limiting the Brook bias range as there's no reason it should have
> to float up from 0 during a cold start. I mean, we know a working bias
> cannot possibly be 0 so there's no reason to burden bias rise time
> with it.
>
> Looks pretty good, though.
>
> The circuit took 5 bipolars (one being Vbe temp comp) and 2 P-MOSFETs.
>
> An alternative is to connect the Brook bias to the output tube grid
> resistors, which allows the current mirror voltage to be less since
> the grids can be taken negative rather that raising Vk. In that case
> the Brook signal needs an inversion and another supply rail (negative)
> but that's not difficult.
How about posting a schematic?
Patrick Turner.
John Byrns
flipper <fli...@fish.net> wrote:
> On Tue, 19 May 2009 09:01:09 GMT, Patrick Turner
> <in...@turneraudio.com.au> wrote:
>
> >When you have 12 output tubes, its difficult to decide what you want to
> >balance. Its easy to balance a see saw, but much more difficult to
> >balance a ferris wheel. Too many people.
>
> Not really. Ever see a tire balancer?
How does balancing a tire have anything to do with balancing 12 tubes?
I don't think many people looking for close balance in a 12 tube amp
would be satisfied with the results of applying the typical tire
balancing scheme to a 12 tube amp. The typical tire balancer down at
the tire shop only controls two degrees of freedom, at least as
typically used. Balancing a 12 tube amp would require
measuring/controlling at least eleven degrees of freedom.
Patrick Turner
flipper wrote:
>
> On Tue, 19 May 2009 10:17:53 GMT, Patrick Turner
> <in...@turneraudio.com.au> wrote:
> >
> >How about posting a schematic?
> >
> >Patrick Turner.
>
> Well, it sure wasn't easy this time but I did.
>
> http://flipperhome.dyndns.org/Auto%20Bias%20and%20Balance.htm
>
> The schematic looks like crap because Circuitmaker got into it's
> favorite mood of crashing anytime something is moved so I had to
> squeeze everything into whatever existing space I could find. Text is
> essentially that post massaged a bit.
I will have to spend serious time to undertsand what you've done because
it looks like you have at least 5 interactive things going on in that
schematic, and the explanation step by step is a bit difficult for my
simple and old mind.
Many thanks,
Patrick Turner.
Patrick Turner
flipper wrote:
>
> On Tue, 19 May 2009 09:01:09 GMT, Patrick Turner
> <in...@turneraudio.com.au> wrote:
>
> >
> >
> >flipper wrote:
> >
> >> >> >> >> >>
> >> >> >> >> >> On Tue, 12 May 2009 08:44:08 GMT, Patrick Turner
> >> >> >> >> >> <in...@turneraudio.com.au> wrote:
> >
> >> >Stable bias does lead to better sound though.
> >>
> >> Then use fixed bias.
> >
> >But in an amp with 12 utput tubes, all the bias pots are a real hazard.
> >People adjust them wrong....
>
> So there are sometimes other considerations.
>
> Exactly my point.
!
>
> >Two tubes in one amp, yeah, no problem, but with 12 tubes? no way man.
> >>
> >> Once on the self promotional high horse it's virtually impossible to
> >> have a rational discussion with you because you never talk about the
> >> damn topic.
> >
> >Its been talked to death on the news group before.
>
> And the problem persists.
Alas, we are all dead, the walking talking Zombies of Tubedom, from
planet Vacuum, where the religion is to be a Cathode Follower.
>
> >> To wit, this whole thread is in response to my comparison of the 'self
> >> balance' in 4 self bias cathode circuits, and a fifth counting the
> >> current mirror comment, but you've decided to flat ass ignore self
> >> balance and speak of nothing but your beloved 'bias stabilization'
> >> circuit.
> >
> >I raised the idea of dynamic bias reg because all manner of self
> >balancing systems such as garter and current mirrors and fidling with
> >grid bias voltages all fail to perform well when inevitably someone uses
> >a low value load on a PP amp which is designed to work in only pure
> >class A.
>
> Except it completely ignores the purpose and point of the discussion
> and is akin to telling someone who asked about watches how to build a
> bicycle. It might be a terrific bicycle but it isn't what was asked.
Its hard for me to stay entirely on topic of bias balancing because
biasing is a bit personal. Its hard for chess players to just
concentrate on one type of chess opening, when one or two moves
different make the opening into another type.
All types of openings in chess lead to either a loss or a win in 5 or 55
moves' time.....
Be liberal man, be happy, don't worry....
>
> >> That's nothing against 'bias stabilization' but it WASN'T the blooming
> >> TOPIC.
> >
> >Agreed entirely. But signal current shunting as I do it in AB amps works
> >a treat to prevent Ek bias drift. Balance is only one issue about
> >biasing any PP amp correctly. The amp should if possible be made immune
> >to heavy class AB signals cause bias drift. in aclass A amp with a load
> >that's too low and too much volume, the bias drifts unevenly. Where you
> >have a pair of LED to indicate bias imbalance of +/-10%, you will see
> >the LEDs flash one side then the other as the bias currents vary
> >somewhat a lot. It is almost impossible to stop this sort of thing
> >unless you stop the cause of it, which is momentary differences in the
> >charge of each Ck on each side of thr PP circuit. So the ac shunting i
> >do helps get rid of all the wobblies.
>
> You really don't get it do you?
No, not always.
If I actively stabilize the bias it tends to stay better balanced.
I don't know much though.
>
> >> >> >> I did play with your circuit but found it difficult to adjust the
> >> >> >> 'current bypass' for minimum shift, and that was with 'ideal'
> >> >> >> components (simulations).
> >> >> >
> >> >> >Build a real circuit and play with values. Then you'll soon find out how
> >> >> >to change values a bit to allow a slight rise in Ek but of less than 10%
> >> >> >of the idle value, and between idle and the +10% the bias rises almost
> >> >> >as a straight line, regardless of output load.
> >> >>
> >> >> Thanks but it doesn't deal with the goal of automatic balance.
> >> >
> >> >Well how does one balance a dozen output tubes?
> >> >
> >> >One way might be to have one tube with a real Rk and the remaining 11
> >> >with current mirrors.
> >>
> >> Yep.
> >>
> >> >But that doesn't stop all the Eks rising together when you move into AB.
> >> >So the amp become mis-biased while in AB, or like an underbiased amp, so
> >> >there will be MUCH more 3H produced because the amp starts to resemble a
> >> >class B amp with heavy crossover distortions.
> >>
> >> Not if, as has been repeated a half dozen times, you run enough Class
> >> A so transients into Class B become insignificant.
> >
> >True in most instances, but with real music at high levels the Ek bias
> >wobbles considerably, especially with low power amps and low sensitivity
> >speakers at high levels.
>
> In that case rising Ek becomes moot because it's clipped to hell and
> back anyway.
Well, you can have a rising EK without any signal clipping. But running
the amp where clipping does occur frequently does exacerbate the rise in
Ek and its awful effects on the sound.
I find that with cathode bias and with 2 x KT88 set up in UL to make
about 50W with say 20W of max pure class A when the load = 4 x the load
for the 50W of AB1, you don't see a huge amount of Ek bias shift if the
levels are sensible and you have 89dB/W/M sensitivity type of speakers.
This is especially so if the speakers are "8' ohms and they are
connected to a "4 ohm" labelled output terminal, so that the tubes then
see twice the load which gives less PO max but more pre class A, and
therefore less dynamic bias drift or wobble.
>
> >> >> I understand the goal but simply 'bypassing' current peaks does not
> >> >> result in a constant average current into the filter caps because tube
> >> >> current also goes to 0.
> >> >>
> >> >> Moot anyway because it's incompatible with automatic balance.
> >> >
> >> >I think I probably could add some means of having Ik of 2 tubes
> >> >equalised, AND have the dynamic as shunting by the bjts and resistors.
> >>
> >> Might be an interesting project. Give it a shot.
> >
> >When you have 12 output tubes, its difficult to decide what you want to
> >balance. Its easy to balance a see saw, but much more difficult to
> >balance a ferris wheel. Too many people.
>
> Not really. Ever see a tire balancer?
Yes but tubes don't spin around, and you can't add a weight.
The people on the ferris wheel all come in slightly different weights,
and so if you want the ferris wheel to remain motionless if not driven
by a motor you have to carefully place the people, so you have to weigh
them and appoint where they sit.
So it is with 12 output tubes. You can measure the ac current at each
cathode, and add up the 6 currents on each side of the PP circuit then
swap out a few tubes to make the ac currents balance. The dc currents
will usually all be fairly close if you have correctly adjusted fixed
bias or you have separate cathode bias. One could have 6 Garter biased
pairs in parallel though, with 6 pairs of LED to tell you if the bias
currents are equal for that pair.
Some sort of master control to adjust the Idc total of each pair to be
the same might be able to be devised.
Its all quickly becomes complex, and I find its good enough to have
simple cathode biasing with the Pda of the tube at idle = about 1/2 the
rated maximum, ie, 20W for 6550, which is plenty, and Rk then tends to
be a larger value and a better regulator than if you have Pda higher.
>
> Of course, as is usual, you change the topic from the "2 tubes" YOU
> stated to now 12 tubes.
C'mon, its a free world, no? balancing bias is an issue for all amp
builders regardless of the number of tubes used, and also including
users of multiple tubes in SE circuits where it is in the music's
interest to have all parallel SE connected tubes operating with equal
Idc.
It so happens that Idc of each tube in an SE amp with say 4 output tubes
could be +/-10% different amoung themselves without much difference to
maximum PO or THD/IMD. But if people wanted to have some means of
gaining better Idc balance or equality than having just equal Rk on each
of the 4 tubes, then they are free to explore. In the SE case there is
not the slightest need to have dynamic bias stabilization as I have
suggested works well in PP AB designs.
>
> >But one can have 12 equal dc current sinks, one for each cathode except
> >that you have to make up a board with 12 darlington connected bjts and
> >their emitter Re, and have a voltage reference for the bases, and all
> >that's a pain when all you really need is a simple humble resistance. If
> >there is inequality of Iadc between the 12 tubes its likely it will even
> >out for the two sides of the PP circuit.
> >That's what I have found in amps with fixed bias and 6 tubes. The drift
> >a bit out, or they have different gm, and to eliminate 2H in the output
> >you sometimes have to swap tubes around after measuring signal currents
> >in each tube and summing the results for each side of the PP circuit.
> >
> >The current sink causes no increase in Ik if a tube decides to conduct
> >more Ia, or Ig2. With an Rk, if Ek rises from say 30V to 60V, at least
> >you know there's twice the Ik, and most probably twice Ia and Ig2. The
> >Rk tends to limit Ek rise. But there isn't any limiting if The tube goes
> >wonky and the bias voltage loses control of the tube *voltages* so you
> >need to have some good active protection. I decided plain Rk provided
> >enough regulation and then decided the active protection would stop
> >runaway tubes, and the dynamic bypass would allow easy carefree biasing
> >for the owner who could then drive the amp as hard as he wanted to with
> >all the benefits of fixed bias, ie, very little bias *voltage* drift
> >caused by rectifying currents charging up Ck, which BTW are each 1,000uF
> >in my 300wattes with 12 output tubes.
> >If Ek drifts with such caps you don't want it to stay drifted for too
> >long, hence the need for the active bypassing.
>
> None of which has a blessed thing to do with you coming up with a way
> "of having Ik of 2 tubes equalised, AND have the dynamic as shunting
> by the bjts and resistors."
That's what you might say.
But perhaps 2.69 others in the world reading the post might find the
salient issues quite interesting.
>
> >> I've already done a current mirror 'fixed bias' version in which there
> >> is no bias shift, and it doesn't suffer RC filter ripple like when
> >> using a differential balance amp, but I haven't been keen on using it
> >> because fixed bias negates the 'idiot proof' self bias aspect.
> >
> >I'm not sure what you mean. Fixed bias means tying the cathodes to 0V
> >without a series parallel R&C network to which you can attatch various
> >schemes to fiddle with the current and voltages.
>
> No, "fixed bias" means you set bias with a(n) (adjustable) 'fixed'
> voltage and there's more than one way to do it.
OK, so you can indeed have a fixed ( but adjustable ) Vdc source or low
impedance Vdec source applied to the cathode at some fixable voltage
above 0V, then have the grid voltage at 0V.
A PNP BJT darlington pair could be used, and some method derived to
equalise Ik of two ( or more ) output tubes, thus balancing the Idc, and
not permitting the Ek to shift during class AB, thus gaining the
benefits of fixed bias while having automatic biasing.
I have not fully explored this possibility; perhaps it can be simpler
and more effective than my bias stabilizer while balancing actively in a
more effective way than simple resistances used in each cathode circuit.
>
> >> I have been wondering, though, if something akin to a Brook bias could
> >> be added to it since there's no RC time constant associated with the
> >> current mirror and it's multiple time constants that aggravate bias
> >> stability problems when trying to 'do both' at the same time.
> >
> >Hm, the eternal bias problem is to have the Idc balanced and the voltage
> >bias about equal.
>
> The goal is to have balanced Idc and at the right value without user
> adjustment.
Indeed.
And if possible have the amp behave like a fixed bias amp although the
cathodes are at an Ek voltage above 0V, and held there by some low
impedance but current sensing voltage source.
>
> >Its impossible to achieve perfection because tubes ain't perfect
> >especally after they gae a bit. Gm changes, and a different Eg1 is
> >required for each tube to get the same Ik.
>
> That they need different bias is the reason for something 'automatic'
> or else the user has to adjust things.
>
> >> On the wild hair side, I've long been curious about making a little
> >> bias micro controller that does something seemingly simple like adjust
> >> fixed bias at turn on and just hold it there till next time it can
> >> check but that's not a trivial project and "seemingly simple" rarely
> >> is.
> >
> >Just connect the amp to a PC via a USB wire and have a program.
>
> Where's the 'USB port' on your amps?
I was joking.
There is no need to use a PC to control a steam engine, or horse and
buggy, or vacuum tube amp.
But don't let me stop you if you wish to. Its beyond my capabilities to
design a board with a bunch of logic chips and memory chips to control
the bias.
>
> >Don't let the B+ mix with the mother board. The B+ always wins, easily,
> >with a complementary puff of smoke.
>
> Rather moot because simply laying some 'USB wires' around isn't going
> to do anything anyway.
>
> >> If a company should read this, hire me and I'll tell you how to do it.
> >
> >Companies know how to stay well away from here. I think we are alone.
>
> Never hurts to say so 'just in case'.
Could be an occasional ghost reading our posts.
I saw old Harold Leak and David Williamson sitting together in the
corner of my room one evening at 1am with only moonbeans shining through
the windows, and they were discussing my amplifiers, and the CD player
complexity, and the damned PC, and when I peeked around the corner and
said hello they vanished.
Sometimes I wake in fright after dreaming thet Fritz Langford-Smith has
made a personal appearance in a dream and told me firmly, "Patrick, you
are quite WRONG....."
OK, not everyone agrees.....
Even my father thinks I am wasting my old age and being ungratious.
He certainly winged about how I misspent my youth!
>
> >> >One reason people keep saying you need 500 times more power instead of
> >> >only 50 times more than your average PO is that its profitable for those
> >> >promotong hugh power, Tell 'em what they don't need convincingly, and oh
> >> >how the money rolls in.
> >>
> >> Yes, of course. It's always a conspiracy and especially when one
> >> doesn't bother to read the rationale.
> >>
> >> I'd tell you but I already have and got tired of hearing you explain
> >> that the typical 5 Watt amp is being run 50mW average Po.
> >
> >The typical 5W amp with say a lone 6BQ5 will battle to give a clean
> >sound into average speakers or 89dB/W/M.
> >So folks with such amps tend to use old fashioned speakers with
> >sensitivity over 95dB/W/M, and then the 50mW might be enough if you work
> >it all out.
>
> Define 'enough'.
It is having enough power so that clipping rarely ever happens.
>
> >I might use about 1 watt average sometimes. It means 89dB at 1W at 1M,
> >and in aroom with two speakers that's quite loud, too loud for me most
> >days, so 500mW will be about right. I might also use a 50W amp.
> >
> >I don't need to be talked into using a 500W amp.
>
> I didn't say you 'needed' one.
Thank goodness.
> I said you should at least understand
> what their arguments are.
Yes, they say we should have a much higher available dynamic headroom
range.
But recorded music is rarely ever presented to us without it being "pre
limited", or "pre-clipped", or without the drums and percussion
instruments being moved well away from the mics, so prevent the
recording gear clipping.
When music is displayed on a CRO, there are usually many bass waves that
have obviously been clipped or heavily cramped, or compressed.
The speakers people use are unable to reproduce the full natural dynamic
range of some instruments at the volume levels you hear while close to
them.
So all hi-fi is a muse, and an illusion, or a painting, and it just
isn't ever real.
I once stood on stage while someone pounded a grand piano at full tilt.
My ears were taxed, and maybe producing considerable distortion.
But the sound was LOUD, and I thought to reproduce that sound I'd need a
line array set of speakers with 20 drive units per channel and a 300W
amp.
But I don't like standing beside the grand. I prefer being 40 to 100
feet away, and then a 50W amp and normal speakers seem to do OK to
reproduce what I would hear.
Line array speakers are on my list to make, but I never get the time.
>
> >I do know a guy with giant JBL monitors with 2 x 15" bass units and
> >horned mid/treble units. He has a pair of Yamaha 2200 amps to provide
> >the power, less than 0.25W most days in his tiny room. Those amps are
> >capable of 200W per channel, and he has 800W system capability.
> >That's an excess capability of around 750W at least.
>
> Clearly, out of the entire population of earth he is the 'one guy' all
> amplifiers should be designed for.
??
Patrick Turner.
Patrick Turner
flipper wrote:
>
> On Wed, 20 May 2009 09:33:38 GMT, Patrick Turner
> <in...@turneraudio.com.au> wrote:
>
> >
> >
> >flipper wrote:
> >>
> >> On Tue, 19 May 2009 09:01:09 GMT, Patrick Turner
> >> <in...@turneraudio.com.au> wrote:
> >>
> >> >
> >> >
> >> >flipper wrote:
> >> >
> >> >> >> >> >> >>
> >> >> >> >> >> >> On Tue, 12 May 2009 08:44:08 GMT, Patrick Turner
> >> >> >> >> >> >> <in...@turneraudio.com.au> wrote:
> >> >
> >> >> >Stable bias does lead to better sound though.
> >> >>
> >> >> Then use fixed bias.
> >> >
> >> >But in an amp with 12 utput tubes, all the bias pots are a real hazard.
> >> >People adjust them wrong....
> >>
> >> So there are sometimes other considerations.
> >>
> >> Exactly my point.
> >
> >!
> >
> >>
> >> >Two tubes in one amp, yeah, no problem, but with 12 tubes? no way man.
> >> >>
> >> >> Once on the self promotional high horse it's virtually impossible to
> >> >> have a rational discussion with you because you never talk about the
> >> >> damn topic.
> >> >
> >> >Its been talked to death on the news group before.
> >>
> >> And the problem persists.
> >
> >Alas, we are all dead, the walking talking Zombies of Tubedom, from
> >planet Vacuum, where the religion is to be a Cathode Follower.
>
> Not so and even Capt. KIrk was an avid antique fan. The 'real' one
> that is. Lord only knows what 'screwed up time line' Kirk likes.
Meanwhile Dr Who was seen wandering around puzzled because he forgot
where he parked his Tardis.
Maybe a Pentoad gobbled it up...
> >> >> To wit, this whole thread is in response to my comparison of the 'self
> >> >> balance' in 4 self bias cathode circuits, and a fifth counting the
> >> >> current mirror comment, but you've decided to flat ass ignore self
> >> >> balance and speak of nothing but your beloved 'bias stabilization'
> >> >> circuit.
> >> >
> >> >I raised the idea of dynamic bias reg because all manner of self
> >> >balancing systems such as garter and current mirrors and fidling with
> >> >grid bias voltages all fail to perform well when inevitably someone uses
> >> >a low value load on a PP amp which is designed to work in only pure
> >> >class A.
> >>
> >> Except it completely ignores the purpose and point of the discussion
> >> and is akin to telling someone who asked about watches how to build a
> >> bicycle. It might be a terrific bicycle but it isn't what was asked.
> >
> >Its hard for me to stay entirely on topic of bias balancing because
> >biasing is a bit personal. Its hard for chess players to just
> >concentrate on one type of chess opening, when one or two moves
> >different make the opening into another type.
> >All types of openings in chess lead to either a loss or a win in 5 or 55
> >moves' time.....
>
> It's one thing to 'mention' what you 'like' but another to obsess on
> it every time someone speaks.
>
> >Be liberal man, be happy, don't worry....
>
> I am liberal, not a socialist.
!
>
> >> >> That's nothing against 'bias stabilization' but it WASN'T the blooming
> >> >> TOPIC.
> >> >
> >> >Agreed entirely. But signal current shunting as I do it in AB amps works
> >> >a treat to prevent Ek bias drift. Balance is only one issue about
> >> >biasing any PP amp correctly. The amp should if possible be made immune
> >> >to heavy class AB signals cause bias drift. in aclass A amp with a load
> >> >that's too low and too much volume, the bias drifts unevenly. Where you
> >> >have a pair of LED to indicate bias imbalance of +/-10%, you will see
> >> >the LEDs flash one side then the other as the bias currents vary
> >> >somewhat a lot. It is almost impossible to stop this sort of thing
> >> >unless you stop the cause of it, which is momentary differences in the
> >> >charge of each Ck on each side of thr PP circuit. So the ac shunting i
> >> >do helps get rid of all the wobblies.
> >>
> >> You really don't get it do you?
> >
> >No, not always.
> >
> >If I actively stabilize the bias it tends to stay better balanced.
> >
> >I don't know much though.
>
> <rolling eyes>
Well, the universe has infinite knowledge to be known, and I only have a
rather dumb little finite brain....
> It's the other way around. With sufficient Class A the amp will be
> will into clipping, destroying the music, before Ek dramatically rises
> and it's rather esoteric whether a pile of shit 'plus 10%' smells
> significantly worse than the original pile when the goal is to not
> have a pile at all.
Hmm.
>
> >I find that with cathode bias and with 2 x KT88 set up in UL to make
> >about 50W with say 20W of max pure class A when the load = 4 x the load
> >for the 50W of AB1, you don't see a huge amount of Ek bias shift if the
> >levels are sensible and you have 89dB/W/M sensitivity type of speakers.
> >This is especially so if the speakers are "8' ohms and they are
> >connected to a "4 ohm" labelled output terminal, so that the tubes then
> >see twice the load which gives less PO max but more pre class A, and
> >therefore less dynamic bias drift or wobble.
> >
> >
> >>
> >> >> >> I understand the goal but simply 'bypassing' current peaks does not
> >> >> >> result in a constant average current into the filter caps because tube
> >> >> >> current also goes to 0.
> >> >> >>
> >> >> >> Moot anyway because it's incompatible with automatic balance.
> >> >> >
> >> >> >I think I probably could add some means of having Ik of 2 tubes
> >> >> >equalised, AND have the dynamic as shunting by the bjts and resistors.
> >> >>
> >> >> Might be an interesting project. Give it a shot.
> >> >
> >> >When you have 12 output tubes, its difficult to decide what you want to
> >> >balance. Its easy to balance a see saw, but much more difficult to
> >> >balance a ferris wheel. Too many people.
> >>
> >> Not really. Ever see a tire balancer?
> >
> >Yes but tubes don't spin around, and you can't add a weight.
>
> No, but it's just as 'valid' an example as a Ferris Wheel.
>
> >The people on the ferris wheel all come in slightly different weights,
> >and so if you want the ferris wheel to remain motionless if not driven
> >by a motor you have to carefully place the people, so you have to weigh
> >them and appoint where they sit.
>
> You also can't stick a transistor under them to force all their
> weights to be equal like you can with tubes.
Well no, most would object. Personally, if instead of a transistor,
Jennifer Hawking was placed under the blokes doing the balance
experiment then you'd sure get good project co-operation.
>
> >So it is with 12 output tubes.
>
> No it isn't.
Oh.
>
> > You can measure the ac current at each
> >cathode, and add up the 6 currents on each side of the PP circuit then
> >swap out a few tubes to make the ac currents balance. The dc currents
> >will usually all be fairly close if you have correctly adjusted fixed
> >bias or you have separate cathode bias. One could have 6 Garter biased
> >pairs in parallel though, with 6 pairs of LED to tell you if the bias
> >currents are equal for that pair.
>
> Or you could just stick a current mirror under each one and have them
> balanced to within 100ua.
But class AB would miss bias all the tubes equally.
>
> >Some sort of master control to adjust the Idc total of each pair to be
> >the same might be able to be devised.
> >
> >Its all quickly becomes complex, and I find its good enough to have
> >simple cathode biasing with the Pda of the tube at idle = about 1/2 the
> >rated maximum, ie, 20W for 6550, which is plenty, and Rk then tends to
> >be a larger value and a better regulator than if you have Pda higher.
>
> It's no more 'complex' than 2. It's just repeated n times like the
> tubes are repeated n times.
>
> >> Of course, as is usual, you change the topic from the "2 tubes" YOU
> >> stated to now 12 tubes.
> >
> >C'mon, its a free world, no?
>
> YOU made a statement about being 'sure' you could add balance to your
> TWO tube 'stabilized bias' circuit and whining about how 'complex'
> things are with 12 tubes is irrelevant.
Some complexity for the benefit of greater power is justified, but
whatever is done in addition to reduce the THD/IMD from low to very low
might be unjustifiable, in terms of cost and reliability. However, the
true liberal builds whatever features he likes regardless of the costs.
Probably I will be seen as being inconsistent.
>
> > balancing bias is an issue for all amp
> >builders regardless of the number of tubes used,
>
> True enough but if you're going to whine about 12 tubes then why the
> hell bring it up when the problem statement was to make your circuit
> work with 2?
I'm not whinging. I'm just wondering how all this discussion can be
benefit anyone with more than two output tubes.
>
> > and also including
> >users of multiple tubes in SE circuits where it is in the music's
> >interest to have all parallel SE connected tubes operating with equal
> >Idc.
> >It so happens that Idc of each tube in an SE amp with say 4 output tubes
> >could be +/-10% different amoung themselves without much difference to
> >maximum PO or THD/IMD. But if people wanted to have some means of
> >gaining better Idc balance or equality than having just equal Rk on each
> >of the 4 tubes, then they are free to explore. In the SE case there is
> >not the slightest need to have dynamic bias stabilization as I have
> >suggested works well in PP AB designs.
>
> For heaven's sake, SE has nothing to do with it, as YOU just SAID. So
> WHY the bloody HELL go INTO it?
People are wanting to discuss balance. Or equality of Ia where more than
ONE OP tube is used.
So the discussion can include balance and equalities in SE designs, no?
> No 'might' to it since I just did.
>
> >But perhaps 2.69 others in the world reading the post might find the
> >salient issues quite interesting.
>
> They might find a discussion of the new Star Trek movie quite
> interesting too but that's not on topic either. That's why things have
> topics: so people can read about the things they find interesting.
Lighten up, chill out, take 5, don't obcess, etc.
>
> >> >> I've already done a current mirror 'fixed bias' version in which there
> >> >> is no bias shift, and it doesn't suffer RC filter ripple like when
> >> >> using a differential balance amp, but I haven't been keen on using it
> >> >> because fixed bias negates the 'idiot proof' self bias aspect.
> >> >
> >> >I'm not sure what you mean. Fixed bias means tying the cathodes to 0V
> >> >without a series parallel R&C network to which you can attatch various
> >> >schemes to fiddle with the current and voltages.
> >>
> >> No, "fixed bias" means you set bias with a(n) (adjustable) 'fixed'
> >> voltage and there's more than one way to do it.
> >
> >OK, so you can indeed have a fixed ( but adjustable ) Vdc source or low
> >impedance Vdec source applied to the cathode at some fixable voltage
> >above 0V, then have the grid voltage at 0V.
>
> Right. I use that with the current mirror to make it 'fixed bias'.
>
> >A PNP BJT darlington pair could be used, and some method derived to
> >equalise Ik of two ( or more ) output tubes, thus balancing the Idc, and
> >not permitting the Ek to shift during class AB, thus gaining the
> >benefits of fixed bias while having automatic biasing.
>
> Yes, as the current mirror does. That's only automatic bias 'balance',
> though. Not automatic 'bias', per see.
>
> >I have not fully explored this possibility; perhaps it can be simpler
> >and more effective than my bias stabilizer while balancing actively in a
> >more effective way than simple resistances used in each cathode circuit.
>
> There's pros and cons either way.
>
> With your approach I'd think one could look back at the original, but
> unacceptable, CCS under each since your 'stabilization' might take
> care of the Ek rise. Might even be easier to 'fine tune' since the
> current, being fixed by the CCS, should then be 'known' rather than a
> moving value depending on how the tubes 'self bias'.
>
> Just some speculative thinking aloud.
The CCS under a cathode to 0V can be used to save heat dissipated in the
CCS if you also have a fixed negative grid bias.
So say the wanted bias is 150V as in the case of an 845. The Ik will be
say 75mA, and the Pd in Rk if used would be 11W.
You can have a fixed Eg at -50V and have the voltage across the CCS at
50V, and still have room for the tube to find its own biasing point.
I thought about this whe I built a pair of 845 amps with two 845 per
channel. But after lots of thought I concluded it was simple and easy
enough to have 2k2 resistances for Rk rated at 40W and on a heat sink to
keep things simple. The pair of tubes had nearly identical Ik, even if a
chinese 845 or KR Audio were used together.
But the app was for SE, not PP, where we want equal bias currents
regardless of Ek. I recall discussions some years ago about the benefits
of having regulated Ek AND regulated Ik. I've never found a way to do
both simply.
Ik moves a lot when an amp moves into class AB, ie, there is an
increasing Idc component in the tube current which charges up the Ck.
You can try to force the two Iks to be equal, but what is also wanted is
to maintain the same Ek at high AB power as at idle, lest you have the
amp become miss biased in AB, or biased like a class B or even a class C
amp.
>
> >> >> I have been wondering, though, if something akin to a Brook bias could
> >> >> be added to it since there's no RC time constant associated with the
> >> >> current mirror and it's multiple time constants that aggravate bias
> >> >> stability problems when trying to 'do both' at the same time.
> >> >
> >> >Hm, the eternal bias problem is to have the Idc balanced and the voltage
> >> >bias about equal.
> >>
> >> The goal is to have balanced Idc and at the right value without user
> >> adjustment.
> >
> >Indeed.
> >
> >And if possible have the amp behave like a fixed bias amp although the
> >cathodes are at an Ek voltage above 0V, and held there by some low
> >impedance but current sensing voltage source.
>
> Yeah. Frankly, I'd rather do it by adjusting grid bias because the
> power under Ek is wasted but tinkering with grid bias has, so far at
> least, always involved multiple filter time constants and the
> resultant stability problems.
But if you can maintain Ek, regardless of the tube current, you should
also have Eg held constant.
The bias required for class A or AB should be the same. This is
effectively the case with conventional fixed bias.
The fixed bias use is deliberately chosen to avoid having any Eg to Ek
variation up to clipping. At clipping, and if caused by grid current,
the coupling caps charge up and the charge can linger to cause the amp
to become over biased. Any attempt to actively drain the charge out of
coupling caps by applying a positive voltage to the negative bias ends
of the bias resistors is a nightmare of delayed actions and can make an
amp very unstable. This trick is not for me.
>
>
> >> >Its impossible to achieve perfection because tubes ain't perfect
> >> >especally after they gae a bit. Gm changes, and a different Eg1 is
> >> >required for each tube to get the same Ik.
> >>
> >> That they need different bias is the reason for something 'automatic'
> >> or else the user has to adjust things.
> >>
> >> >> On the wild hair side, I've long been curious about making a little
> >> >> bias micro controller that does something seemingly simple like adjust
> >> >> fixed bias at turn on and just hold it there till next time it can
> >> >> check but that's not a trivial project and "seemingly simple" rarely
> >> >> is.
> >> >
> >> >Just connect the amp to a PC via a USB wire and have a program.
> >>
> >> Where's the 'USB port' on your amps?
> >
> >I was joking.
>
> OK
>
> >There is no need to use a PC to control a steam engine, or horse and
> >buggy, or vacuum tube amp.
>
> 20 years ago you'd have probably included an automobile gasoline
> engine in the list but they got 'em now.
yeah.
A PC is a form of primitive intelligence.
Added electronic intelligence is nearly universal now; in solar heating
and power systems, washing machines and toasters.
>
> With a micro controller you can implement complex functions much
> easier than you can with physical devices, and you can also 'control'
> things. Like, for example, you could hold off audio at power up for a
> 'calibrate' period, measure idle, even determine when it's 'finished
> moving' (within some tolerance), and then 'fix' the bias voltages at
> that point till next turn on.
These things are wonderful until they go phut some years later when
everyone has lost the information on them.
Then you have to remove them and go back to plain old fixed bias, and
then remember to do regular checks.
>
> Dern near trivial with a micro-controller but much more complex to do
> with discrete components. And 'boutique' amps containing a remote
> controlled volume/balance control already have almost enough
> 'computing power' to add it in. Not quite but it's 'less of a more'
> than a standalone implementation.
>
> >But don't let me stop you if you wish to. Its beyond my capabilities to
> >design a board with a bunch of logic chips and memory chips to control
> >the bias.
>
> More complex than I want to bother with right now too, although I do
> have the tools. Primitive, but enough.
Yeah.
>
> Essentially 'one' chip, btw. Micro controllers come with programmable
> ROM, RAM, ADC, D/A, and digital outputs. 'All it takes' is some
> programming. (The 'all it takes' is a joke)
>
> That's why you find them in dern near everything these days.
>
> It's front end heavy development, though, and hard to justify for a
> 'one of' amp.
I've used a pair of bjts and a pair of LEDS to indicate bias balance and
then I've provided a pot for an owner to twiddle bias into balance when
one LED goes a little less bright than the other .
If the LED brightness wobbles a lot on high volume, it indicates volume
is too high and clipping could be happening.
But it also indicates if something else is wrong, like a stuffed tube,
or a shorting speaker cable.
Owners seem to find the twin LED arrangement for manual balance of Ik to
be very easy to live with. They know if things are OK with the OP tubes
at all times.
> Insufficient. Hell, ONE milliwatt is 'enough'.
> If you're using headphones.
I forgot to say enough is enough when clipping rarely happens when using
the desired volume level.
>
> >> >I might use about 1 watt average sometimes. It means 89dB at 1W at 1M,
> >> >and in aroom with two speakers that's quite loud, too loud for me most
> >> >days, so 500mW will be about right. I might also use a 50W amp.
> >> >
> >> >I don't need to be talked into using a 500W amp.
> >>
> >> I didn't say you 'needed' one.
> >
> >Thank goodness.
> >
> >> I said you should at least understand
> >> what their arguments are.
> >
> >Yes, they say we should have a much higher available dynamic headroom
> >range.
>
> And they give you the math for why.
>
> >But recorded music is rarely ever presented to us without it being "pre
> >limited", or "pre-clipped", or without the drums and percussion
> >instruments being moved well away from the mics, so prevent the
> >recording gear clipping.
>
> They knew how music is recorded.
>
> >When music is displayed on a CRO, there are usually many bass waves that
> >have obviously been clipped or heavily cramped, or compressed.
>
> Depends on the recording. It won't be if done properly.
>
> >The speakers people use are unable to reproduce the full natural dynamic
> >range of some instruments at the volume levels you hear while close to
> >them.
> >So all hi-fi is a muse, and an illusion, or a painting, and it just
> >isn't ever real.
> >
> >I once stood on stage while someone pounded a grand piano at full tilt.
> >My ears were taxed, and maybe producing considerable distortion.
> >But the sound was LOUD, and I thought to reproduce that sound I'd need a
> >line array set of speakers with 20 drive units per channel and a 300W
> >amp.
> >But I don't like standing beside the grand. I prefer being 40 to 100
> >feet away, and then a 50W amp and normal speakers seem to do OK to
> >reproduce what I would hear.
>
> It'll do 'OK', depending on how one defines 'OK'.
A guy here has a huge pair of Soundlabs ESL and has found that PP amps
with 8 x 6550 in triode for 100W doesn't work very well.
He likes the level to be quite HIGH. Soundlabs say that a 300W amp
should be used. Lord knows what is the speaker impedance character.
But one customer of mine has 4 x KT90 per channel in a PP CFB arangement
to drive 3 stacked Quad ESL57 per channel.
He has enough power to give a nice clean level he wants.
>
> They made a compelling technical case but whether it's 'noticeable' or
> 'necessary', or even your objective, is a subjective judgment.
>
> >Line array speakers are on my list to make, but I never get the time.
> >
> >
> >>
> >> >I do know a guy with giant JBL monitors with 2 x 15" bass units and
> >> >horned mid/treble units. He has a pair of Yamaha 2200 amps to provide
> >> >the power, less than 0.25W most days in his tiny room. Those amps are
> >> >capable of 200W per channel, and he has 800W system capability.
> >> >That's an excess capability of around 750W at least.
> >>
> >> Clearly, out of the entire population of earth he is the 'one guy' all
> >> amplifiers should be designed for.
> >
> >??
>
> The point was that anecdotal stories only apply to the anecdote and
> not always even then. For example, it could very well be the case that
> "most days" he operates the system for nothing more than soft
> background music and on 'special occasions' gets down to some serious
> jamming. After all, that's why they put volume knobs on 'em.
Yes, but 25 watts could make speakers with 96dB/W/M sensitivity go
painfully loud in a small room.
>
> I don't know but it doesn't matter because there are a few billion
> people who likely have different needs.
Indeed, all those billions out there. Maybe 20% have not yet used a
telephone.
Patrick Turner.
>
> >Patrick Turner.
WB
> I wonder how this amp measures. Without repeatable measurements I
> have no opinion on any amp because very terrible amps can " sound
> good" to some people who like infidelious sound. Especially through
> Cornwalls, which are a rude speaker in many situations.
>
> Proper tweaking of the Cornwall can improve things quite a bit=the
> Avedon T35 EV tweeter needs to go-but still it's raw.
Hi,
That amp is very nice looking, especially the night time shot. As
for the comment above, measurements and a nickel may make one believe an
amp should sound better then another. But, in the real world, many times
an amp, which shows up with poor specs in testing, sounds better to the
human ear when playing actual music. A case in point is the DHSET amp.
Most SET amps, that use minimal or no feedback, have specs that say they
should sound less then perfect when compared to push pull amps. But, I
have heard and built several SET amps that just plain sound amazing. You
need to actually hear an amp before you can judge it. Specifications are
not the last word in an amplifier's performance when playing music. Your
ears are. I just happen to have a radio with push pull 71A's and
transformer coupling. Before I started working with the 2A3 and 300B, I
jury-rigged this radio such that I could drive the power amp section
with a CD player and a reworked Dynaco PAS2. I attached it to a Mirage
190is and then to an 8" Utah Celestia in a custom made cabinet. This amp
sounded quite nice through both of these speakers. It is probably very
similar to this 71A amp we are talking about here. I tested the
frequency response and it did not show up to be very good (about 70hz to
about 8Khz with fast roll off at each end of that spectrum). Still, the
amp sounded really good to my ears as well as other people's ears that I
let hear it. In fact, a couple people wanted to know if I could build
them a stereo version of the radio's power amp. My point is, don't
discount any design until you listen to it. This amp obviously took a
lot of work and looks very professional. This was a very nice job.
Aren't Klipsch speakers great? Man, all you need is a couple of watts
and you can fill a room with sound. My 2A3 parallel amps are single
ended and deliver around 7 watts a channel. I tried them through a pair
of Klipsch speakers and boy were they loud.
Bill