1st tube amp: newbie questions
Nikolai Bakschies
Hi,
I'm just starting to built my first tube power amp. Since I'm not so
very experienced I have some questions on how-to (yes, I want to test
and play, but I want to avoid the worst errors.. ;-)
What is planned:
- stereo power amp wit 4 EL84 per channel in PP configuration
(ultra-linear)
- 1 dual triode (ECC81, 82, 83) as input stage and phase splitter
(cathodyne) per channel
- no overall negative feedback, if possible
- full power at 1V(eff) input voltage (CD player will be connected
directly)
Objectives
The amp shall not be designed for much output power but for best audio
performance.
For my first one I want to use standard tubes.
Questions:
1.) what triode to use ? I'm thinking about ECC81 or ECC82 because of
their lower inner resistance compared to the ECC83. Any comments on
those tubes, esp. distortion etc ?
2.) In general, can anybody tell me if the audio quality is affected by
the plate voltage/the plate current (looking at the tube diagrams it
seems that driving them at higher voltage results in more linearity)
3.) are there any DISadvantages when I use very big capacitors for my
(diode) power supply ?
4.) will the amp work without negative feedback from output transformer
to input stage (assumed the tubes are not microphonic and the wiring is
OK)?
5.) what's the use of the serial resistors I've often seen placed before
the grid ? I've read something about HF-compensation. To my opinion
every resistor in the signal path causes "errors". May I omit them ?
adTHANXvance for any input !
Niko ;-)
SBench
>Subject: 1st tube amp: newbie questions
>From: Nikolai Bakschies <ni...@topteam.duedorf.de.ibm.com>
>Date: Mon, Mar 9, 1998 13:01 EST
>Message-id: <35042E92...@topteam.duedorf.de.ibm.com>
Hello,
For the ECC81-3 parts, with no negative feedback,
the 83 may end up with the lowest distortion, depending
on your circuit topology. The good news is that since all
3 types use the same socket, if you bias the circuit for
ECC83, you can probably try all 3 types to see which
sounds best to you.
Generally higher voltage/current on the tube will result in
lower distortion. Higher supply voltage gives you more
potential swing, so at a given signal level, the distortion can
be lower. Higher current results in operating in the more
linear regions too. Be careful not to overdisipate the tube
by combining too high voltage with too high current.
There are only a couple disadvantage too very large caps
in the PS, especially when combined with solid state rectifier.
These are, during turn on/turn off conditions, you will be
causing some increased level of "cathode stripping" which
will cause slightly lower tube life, and VERY large caps
have a tendency to be slightly less effective at high
frequencies. The later may affect the quality of sound,
again depending on your circuit topology.
The amp will work without negative feedback. I generally
put a "pot" in the NFB leg and adjust it to what sounds best
(which varies depending on the source material).
The series grid resistors are called "stopper" resistors
and are intended to de Q the grid circuit to prevent parasitic
oscillations. Usually not needed, but very good preventative
measure. Would recommend keeping them.
Best Regards,
Steve
Sander deWaal
On Mon, 09 Mar 1998 19:01:55 +0100, Nikolai Bakschies
<ni...@topteam.duedorf.de.ibm.com> wrote the following:
>What is planned:
>- stereo power amp wit 4 EL84 per channel in PP configuration
>(ultra-linear)
Good, tho triode is even better.
>- 1 dual triode (ECC81, 82, 83) as input stage and phase splitter
>(cathodyne) per channel
ECC 81 is preferrable, it's the middle ground btw. the low-mu
but powerful 82, and the hi-mu, but incapable of delivering
enough juice 83.
>- no overall negative feedback, if possible
I'm afraid not, in your setup (penthode UL), the output
impedance will be too high, and frequency range perhaps
too low.
Just add some 10 dB of GNB.
>- full power at 1V(eff) input voltage (CD player will be connected
>directly)
Achieveable with ECC 81 and 10 dB of GNB.
>Objectives
>The amp shall not be designed for much output power but for best audio
>performance.
>For my first one I want to use standard tubes.
Good headstart.
>Questions:
>1.) what triode to use ? I'm thinking about ECC81 or ECC82 because of
>their lower inner resistance compared to the ECC83. Any comments on
>those tubes, esp. distortion etc ?
The 82 is a nice driver tube, the 83 has lots of gain,
and the 81 is the tube for you :-)
>2.) In general, can anybody tell me if the audio quality is affected by
>the plate voltage/the plate current (looking at the tube diagrams it
>seems that driving them at higher voltage results in more linearity)
This is a simple question, which requires a difficult answer!
In general, more supply voltage will result in the ability
to give bigger output swing, especially in the phase splitter
and driver stages (if present).
The common voltage, used for a pair of EL 84's, is 300 V or so,
accordingly to the tubes' specs.
Plate current is also important, if you have access to a data
book (or webpage), you can see that a given tube will be most
linear in a certain portion of the Ia/Va characteristic.
If you're designing from scratch, it's probably best to look
up some formulas in the RDH, and have the tube characteristics
at hand.
>3.) are there any DISadvantages when I use very big capacitors for my
>(diode) power supply ?
Could be.
If you're using a tube rectifier, the cap directly connected to the
cathode, must not exceed 50 uF (for most common rectifiers).
If you're using silicon diodes, the charging current when empty,
may reach very high values.
The wise thing to do, is to put some resistors in series with these
diodes (like 10...100 ohm/5W), in order to prevent this current peak.
IMO, a value bigger than 500 uF is useless for a stereo EL84 PP amp.
Remember to rate the cap's voltage for an unloaded power supply!
>4.) will the amp work without negative feedback from output transformer
>to input stage (assumed the tubes are not microphonic and the wiring is
>OK)?
It will work, but the amp will act like a current source,
more or less.
Due to the configuration (UL-PP), the output impedence will be
fairly high, and the distortion, too.
It takes very impedence-friendly speakers to obtain good sound.
>5.) what's the use of the serial resistors I've often seen placed before
>the grid ? I've read something about HF-compensation. To my opinion
>every resistor in the signal path causes "errors". May I omit them ?
They are mounted for anti-oscillation purposes.
Usually, the value is 1...10 K, and since they don't carry
any current, no power will be dissipated in them (unless grid
current is drawn, unlikely in this setup).
Mount them as close as possible to the tube's grid.
Do NOT omit them, especially with high-mu penthodes (like the EL84).
For the same reason, resistors may be found in the screen grid
circuit, tho lower in value (usually about 100 ohms)..
_
Sander deWaal
postm...@pegasus.demon.nl
www.pegasus.demon.nl
_______________________________________________
peter kalinkov
Nikolai Bakschies <ni...@topteam.duedorf.de.ibm.com> wrote:
>Questions:
>1.) what triode to use ? I'm thinking about ECC81 or ECC82 because of
>their lower inner resistance compared to the ECC83. Any comments on
>those tubes, esp. distortion etc ?
There are hundreds of well documented circuits. A Leak Stereo 20
sounds very close to your thinking: ECC83 i/p, ECC83 phase splitter,
EL84 output. This will get you a decent amp if you use decent iron but
don't expect miracles. The Leak is (or nearly) class A and in my
opinion is difficult to significantly better the circuit using the
same valve train. You can also try to hook the EL84 as triodes; power
is significantly reduced but some prefer the sound.
>2.) In general, can anybody tell me if the audio quality is affected by
>the plate voltage/the plate current (looking at the tube diagrams it
>seems that driving them at higher voltage results in more linearity)
>3.) are there any DISadvantages when I use very big capacitors for my
>(diode) power supply ?
Subjectively, maybe. Large caps may add substance to the bass but
often give the impression of slowing down the pace and rducing the
air.
>4.) will the amp work without negative feedback from output transformer
>to input stage (assumed the tubes are not microphonic and the wiring is
>OK)?
Out of the question unless you're happy with double digit distorsion
figures. Not with the EL84.
>5.) what's the use of the serial resistors I've often seen placed before
>the grid ? I've read something about HF-compensation. To my opinion
>every resistor in the signal path causes "errors". May I omit them ?
They probably reduce the likelihood of HF oscillation acting as low
pass filters together with the grid capacitance. They certainly change
the sound and for my ears to the better. You have to compare and
decide.
Have fun,
Peter
Max or Paul
> Good, tho triode is even better.
It depends on how much power he needs, i.e. what type of speakers he
has.
> ECC 81 is preferrable, it's the middle ground btw. the low-mu
> but powerful 82, and the hi-mu, but incapable of delivering
> enough juice 83.
>
Its also the least linear of the three -- just look how bent its curves
are. Why does the 'power' of the ECC83 even matter if he's using it to
drive such small output valves (EL84) in class AB1 pentode? Mullard's
5-10 used the '83 quite sucessfully in this position, and may be a good
starting point-- just add more output valves.
> >Questions:
> >1.) what triode to use ? I'm thinking about ECC81 or ECC82 because of
>
> >their lower inner resistance compared to the ECC83. Any comments on
> >those tubes, esp. distortion etc ?
>
> The 82 is a nice driver tube, the 83 has lots of gain,
> and the 81 is the tube for you :-)
Also consider the 6CG7 the 6SN7 and the 5687 all are very linear, and
easily available.
I second your advice about the grid stoppers! Ommiting these will
definitly reduce the chances of the amp being stable when its first put
together.
Max
Wilfred
Hi Niko,
PPP-UL EL84, I like that!
My own EL84 amp uses two per channel, driven by one ECC83 and no negative
feedback.
Well, ehm no feedback...
When I switched from my old blackened Philips EL84's to Sovtek EL84M's, it
became way to harsh, so I used the 8 ohm secondary as cathode windings as
well. The trannys have a 2 ohm tap that can be used as a center tap to
ground for this use. It took the razor edge out of the EL84M's.
Triode mode was too low power for me. Try it. Maybe it will eliminate the
need for feedback alltogether. A friend of mine uses six triode strapped
EL84's per channel, no NFB.
I preferred the sound of the ECC83 over the ECC82 in a loftin-White type
phase inverter. With both types the input sensitivity will be OK for direct
input from a CD player. The ECC83 gives a little more sensitive input, but
not that much.
Good luck,
Wilfred
Nikolai Bakschies heeft geschreven in bericht
<35042E92...@topteam.duedorf.de.ibm.com>...
>Hi,
>I'm just starting to built my first tube power amp. Since I'm not so
>very experienced I have some questions on how-to (yes, I want to test
>and play, but I want to avoid the worst errors.. ;-)
>
>What is planned:
>- stereo power amp wit 4 EL84 per channel in PP configuration
>(ultra-linear)
>- 1 dual triode (ECC81, 82, 83) as input stage and phase splitter
>(cathodyne) per channel
>- no overall negative feedback, if possible
>- full power at 1V(eff) input voltage (CD player will be connected
>directly)
>
>Objectives
>The amp shall not be designed for much output power but for best audio
>performance.
>For my first one I want to use standard tubes.
>Niko ;-)
>
>
André Jute
Nikolai Bakschies <ni...@topteam.duedorf.de.ibm.com> wrote:
> Hi,
> I'm just starting to built my first tube power amp. Since I'm not so
> very experienced I have some questions on how-to (yes, I want to test
> and play, but I want to avoid the worst errors.. ;-)
>
> What is planned:
> - stereo power amp wit 4 EL84 per channel in PP configuration
> (ultra-linear)
> - 1 dual triode (ECC81, 82, 83) as input stage and phase splitter
> (cathodyne) per channel
> - no overall negative feedback, if possible
> - full power at 1V(eff) input voltage (CD player will be connected
> directly)
If you design for 1V out of CD you may not have enough gain. To my way
of thinking an integrated amp should be designed for 300mV sensitivity
or less. You can always cut gain, or simply use only part of the
available gain in the most linear part of the transfer, but it is more
difficult to add on gain. Like in a haircut.
> Objectives
> The amp shall not be designed for much output power but for best audio
> performance.
> For my first one I want to use standard tubes.
>
> Questions:
> 1.) what triode to use ? I'm thinking about ECC81 or ECC82 because of
> their lower inner resistance compared to the ECC83. Any comments on
> those tubes, esp. distortion etc ?
I know they look best with bigger tubes like the EL-34, but have you
considered the 6SN7 and 6SL7 tubes. 6.3V fils and they generally have
better characteristics and sound better than the miniature tubes. To me
anyway.
Of the miniature tubes, I don't know what the 12AX7 has, except massive
amplification, to make it so common. To my ears it is overblown. 12AU7
is quite a bit better but low gain. The crisp, clean miniature tube is
the 12AT7, which has amplification between the other two, but some
people think it a bit too crisp--some even say "dry". I prefer to think
of it as precise, a tube you want to consider in association with those
before and especially after it. Great with warmish tubes like the
EL84/34.
If you're considering miniatures seriously, the EF86 should not be
dismissed just because this month's fashion is against pentodes. In
pentode or triode mode it has awesome linearity and the built-in hum
shield is worth its weight in gold. In current production from Svetlana.
The sort of amp you have in mind, distortion at listening (as opposed to
measuring) levels is so far under the audibility threshhold, it is of
concern only to the scope-jockeys. If you're building an amp in this
class for yourself you swap tubes not to measure better but to match the
sound better to your expectation and taste.
> 2.) In general, can anybody tell me if the audio quality is affected by
> the plate voltage/the plate current (looking at the tube diagrams it
> seems that driving them at higher voltage results in more linearity)
Hell, yes. High voltages and fat currents add immensely to the clarity
of the sonic experience.
> 3.) are there any DISadvantages when I use very big capacitors for my
> (diode) power supply ?
> 4.) will the amp work without negative feedback from output transformer
> to input stage (assumed the tubes are not microphonic and the wiring is
> OK)?
Try it on test. In any event, the type and amount of feedback, if any is
required, can change the sonic character of your amp. Use a high quality
pot in the loop feedback line when you first build the amp. When you
have determined which setting sounds best, you can replace it with a
fixed resistor. If you're against feedback on principle, try local
feedback around the source of the instability before you try loop
feedback.
> 5.) what's the use of the serial resistors I've often seen placed before
> the grid ? I've read something about HF-compensation. To my opinion
> every resistor in the signal path causes "errors". May I omit them ?
Grid stoppers. You can leave them off if there is no oscillation. I
always leave them in. (I notice now on reading the next post that Steve
describes them as "parasitic". I've always wonder what they feed on.
Imagine them stuck like limpets on the signal and thereby escaping the
amp and passing on the soundwaves through your ears into your brain
where they do terrible things...)
> adTHANXvance for any input !
>
> Niko ;-)
Andre
--
Andre Jute an...@indigo.ie COMMUNICATION JUTE
--see our pages for music lovers, writers and audiophiles at
http://indigo.ie/~andre/ComJuteF1.html
http://www.foundmark.com/ComJute/ComJuteF1.html
Duncan Munro
>When I switched from my old blackened Philips EL84's to Sovtek EL84M's, it
>became way to harsh, so I used the 8 ohm secondary as cathode windings as
>well. The trannys have a 2 ohm tap that can be used as a center tap to
>ground for this use. It took the razor edge out of the EL84M's.
Wilfred, you may care to try the Teslas, they are a little smoother.
>Triode mode was too low power for me. Try it. Maybe it will eliminate the
>need for feedback alltogether. A friend of mine uses six triode strapped
>EL84's per channel, no NFB.
Triode mode gives more NFB due to the lower plate resistance. Kind of
takes the fun out of it!!!!
>I preferred the sound of the ECC83 over the ECC82 in a loftin-White type
>phase inverter. With both types the input sensitivity will be OK for direct
>input from a CD player. The ECC83 gives a little more sensitive input, but
>not that much.
Might be worth trying the ECC81/12AT7 also. I took my ECC83/12AX7
phase splitter out, and replaced with ECC8/12AU7. The amp just died.
It now has ECC81/12AT7 which is the compromise I'm looking for./..
Cordials,
Duncan
--
Duncan Munro
Check out http://www.duncanamps.simplenet.com/
for Homebrew guitar amps and SPICE models
BOF #023
Buffstereo
Oooops!!! I just posted a noste about 12AX7s driving EL84s in a Dynaco SCA-35.
Well, they don't! The EL84s are driven by 7199s. I am sorry for the
obviously boneheaded statement but I was talking to my wife at the same time
and obviously I've difficulty doing two things at once. Walking and chewing
gum at the same time is not an option for me! However removing the feedback
did make the Dynaco amp sound more natural to me and that was what I really
meant to say! Robert Thompson
"Use a transistor, go to jail !!!"
----------------------------------------
(VTV bumpersticker)
Andy Moss
Nikolai Bakschies wrote:
>
> Hi,
> I'm just starting to built my first tube power amp. Since I'm not so
> very experienced I have some questions on how-to (yes, I want to test
> and play, but I want to avoid the worst errors.. ;-)
I hear ya'!
> What is planned:
> - stereo power amp wit 4 EL84 per channel in PP configuration
> (ultra-linear)
Good choice.
> - 1 dual triode (ECC81, 82, 83) as input stage and phase splitter
> (cathodyne) per channel
So far so good.
> - no overall negative feedback, if possible
Very possible.
> - full power at 1V(eff) input voltage (CD player will be connected
> directly)
No sweat.
> Objectives
> The amp shall not be designed for much output power but for best audio
> performance.
> For my first one I want to use standard tubes.
>
> Questions:
> 1.) what triode to use ? I'm thinking about ECC81 or ECC82 because of
> their lower inner resistance compared to the ECC83. Any comments on
> those tubes, esp. distortion etc ?
I would recommend a 6FQ7/6CG7 as the first voltage amplifier and then
use the other half of the tube as the phase splitter.
> 2.) In general, can anybody tell me if the audio quality is affected by
> the plate voltage/the plate current (looking at the tube diagrams it
> seems that driving them at higher voltage results in more linearity)
Yes. As a rule of thumb in a voltage amplifier (or other config), the
more current you draw through a device (tube, transistor, FET) the more
linear it will be. However, linearity comes with the price of overall
gain.
> 3.) are there any DISadvantages when I use very big capacitors for my
> (diode) power supply ?
Yes, you stand a better chance of blowing out your recitfier! The
benefits of excessive power supply capacitance are lost on my ears.
> 4.) will the amp work without negative feedback from output transformer
> to input stage (assumed the tubes are not microphonic and the wiring is
> OK)?
Yup! It will be more stable that if it had NFB.
> 5.) what's the use of the serial resistors I've often seen placed before
> the grid ? I've read something about HF-compensation. To my opinion
> every resistor in the signal path causes "errors". May I omit them ?
I wouldn't. These are grid-stop resistors. If you remove them you will
likely introduce instabilities. They won't hurt the sound of the
amplifier. And a signal generator and oscilloscope will demonstrate
that to you after you've finished the amplilfier.
Basically you want to build a three stage amplifier. The gain of the
phase splitter is nil - or slightly less than nil. The gain of the
input stage is determined by your choices of cathode resistor, plate
resistor and plate supply voltage. Henry P. discusses load lines in his
FAQ - and does a better job that I can here and now. I highly recommend
you read it if you haven't already.
You will likely want to get a reasonable amount of gain out of your
final stage. To do this you will either have to use a fixed bias
voltage or use self biasing with capacitor bypassing. Chances are your
B+ is fixed so you can't vary that!
This is a good amplifier to experiment with. If done right, it should
sound quite good. The simplicity of the circuit will allow easy
experminentation and the lack of NFB will make every different type of
component sound different! 8-)
Cheers!
Andy
Vince Rhea
Please remember to not loose focus when you have your pinkies in amongst 375 volts.
It may well be the last time you are distracted. Pay attention, lad, pay attention!
Nikolai Bakschies
Hi,
first of all THAKS for the helpful input of you all :-))
Hope further (stupid) questions of mine are answered the same way... ;-)
André Jute wrote:
> Nikolai Bakschies <ni...@topteam.duedorf.de.ibm.com> wrote:
>
> > Hi,
> > I'm just starting to built my first tube power amp. Since I'm not so
> > very experienced I have some questions on how-to (yes, I want to test
> > and play, but I want to avoid the worst errors.. ;-)
> >
> > What is planned:
> > - stereo power amp wit 4 EL84 per channel in PP configuration
> > (ultra-linear)
> > - 1 dual triode (ECC81, 82, 83) as input stage and phase splitter
> > (cathodyne) per channel
> > - no overall negative feedback, if possible
> > - full power at 1V(eff) input voltage (CD player will be connected
> > directly)
>
> If you design for 1V out of CD you may not have enough gain. To my way
> of thinking an integrated amp should be designed for 300mV sensitivity
> or less. You can always cut gain, or simply use only part of the
> available gain in the most linear part of the transfer, but it is more
> difficult to add on gain. Like in a haircut.
>
I've checked the output of my CD-player and it pushes 2V peak to the outputs.
So sensitivity of 0,5V - 1V should be OK. This could be achived with a ECC82
with an ampl.-factor (BTW what's the engl. abbrev. vor that?) of about 15, I
think.
Regarding the ECC81 I'm afraid the box will get too sensitive.
> > Objectives
> > The amp shall not be designed for much output power but for best audio
> > performance.
> > For my first one I want to use standard tubes.
> >
> > Questions:
> > 1.) what triode to use ? I'm thinking about ECC81 or ECC82 because of
> > their lower inner resistance compared to the ECC83. Any comments on
> > those tubes, esp. distortion etc ?
>
> I know they look best with bigger tubes like the EL-34, but have you
> considered the 6SN7 and 6SL7 tubes. 6.3V fils and they generally have
> better characteristics and sound better than the miniature tubes. To me
> anyway.
>
> Of the miniature tubes, I don't know what the 12AX7 has, except massive
> amplification, to make it so common. To my ears it is overblown. 12AU7
> is quite a bit better but low gain. The crisp, clean miniature tube is
> the 12AT7, which has amplification between the other two, but some
> people think it a bit too crisp--some even say "dry". I prefer to think
> of it as precise, a tube you want to consider in association with those
> before and especially after it. Great with warmish tubes like the
> EL84/34.
>
> If you're considering miniatures seriously, the EF86 should not be
> dismissed just because this month's fashion is against pentodes. In
> pentode or triode mode it has awesome linearity and the built-in hum
> shield is worth its weight in gold. In current production from Svetlana.
>
> The sort of amp you have in mind, distortion at listening (as opposed to
> measuring) levels is so far under the audibility threshhold, it is of
> concern only to the scope-jockeys. If you're building an amp in this
> class for yourself you swap tubes not to measure better but to match the
> sound better to your expectation and taste.
>
> > 2.) In general, can anybody tell me if the audio quality is affected by
> > the plate voltage/the plate current (looking at the tube diagrams it
> > seems that driving them at higher voltage results in more linearity)
>
> Hell, yes. High voltages and fat currents add immensely to the clarity
> of the sonic experience.
What do you think about ECC82 in the following config as a driver stage.Rk =
2k2
Ug = -10,4V (I've read something about a THD minimum at this level)
Ra = 30k
U(+) = 300V
If I'm right this should result in about 5mA Ia .
> > 3.) are there any DISadvantages when I use very big capacitors for my
> > (diode) power supply ?
> > 4.) will the amp work without negative feedback from output transformer
> > to input stage (assumed the tubes are not microphonic and the wiring is
> > OK)?
>
> Try it on test. In any event, the type and amount of feedback, if any is
> required, can change the sonic character of your amp. Use a high quality
> pot in the loop feedback line when you first build the amp. When you
> have determined which setting sounds best, you can replace it with a
> fixed resistor. If you're against feedback on principle, try local
> feedback around the source of the instability before you try loop
> feedback.
>
Is there any (audible) difference between local FB (partial un-coupled Rk) and
loop feedback? I could imagine that local feedback will be better for the
signal-phase, but I'm not sure...
> > 5.) what's the use of the serial resistors I've often seen placed before
> > the grid ? I've read something about HF-compensation. To my opinion
> > every resistor in the signal path causes "errors". May I omit them ?
>
> Grid stoppers. You can leave them off if there is no oscillation. I
> always leave them in. (I notice now on reading the next post that Steve
> describes them as "parasitic". I've always wonder what they feed on.
> Imagine them stuck like limpets on the signal and thereby escaping the
> amp and passing on the soundwaves through your ears into your brain
> where they do terrible things...)
>
> > adTHANXvance for any input !
> >
> > Niko ;-)
>
> Andre
> --
> Andre Jute an...@indigo.ie COMMUNICATION JUTE
> --see our pages for music lovers, writers and audiophiles at
> http://indigo.ie/~andre/ComJuteF1.html
> http://www.foundmark.com/ComJute/ComJuteF1.html
Additional (stupid?) question:
Since I do not have any old tube books (only some new stuff and tables) I do
not know what is the best configuration for the cathodyne PS stage. Since
ampl.-factor is 1 I would like to know if here high current is also
preferrable...
Can anybody recommend a good book on calculating tube amp circuits?
Niko ;-)
Nikolai Bakschies
Andy Moss wrote:
> Nikolai Bakschies wrote:
> >
> > Hi,
> > I'm just starting to built my first tube power amp. Since I'm not so
> > very experienced I have some questions on how-to (yes, I want to test
> > and play, but I want to avoid the worst errors.. ;-)
>
> I hear ya'!
>
> > What is planned:
> > - stereo power amp wit 4 EL84 per channel in PP configuration
> > (ultra-linear)
>
> Good choice.
>
> > - 1 dual triode (ECC81, 82, 83) as input stage and phase splitter
> > (cathodyne) per channel
>
> So far so good.
>
> > - no overall negative feedback, if possible
>
> Very possible.
>
> > - full power at 1V(eff) input voltage (CD player will be connected
> > directly)
>
> No sweat.
>
> > Objectives
> > The amp shall not be designed for much output power but for best audio
> > performance.
> > For my first one I want to use standard tubes.
> >
> > Questions:
> > 1.) what triode to use ? I'm thinking about ECC81 or ECC82 because of
> > their lower inner resistance compared to the ECC83. Any comments on
> > those tubes, esp. distortion etc ?
>
> I would recommend a 6FQ7/6CG7 as the first voltage amplifier and then
> use the other half of the tube as the phase splitter.
>
> > 2.) In general, can anybody tell me if the audio quality is affected by
> > the plate voltage/the plate current (looking at the tube diagrams it
> > seems that driving them at higher voltage results in more linearity)
>
> Yes. As a rule of thumb in a voltage amplifier (or other config), the
> more current you draw through a device (tube, transistor, FET) the more
> linear it will be. However, linearity comes with the price of overall
> gain.
>
> > 3.) are there any DISadvantages when I use very big capacitors for my
> > (diode) power supply ?
>
> Yes, you stand a better chance of blowing out your recitfier! The
> benefits of excessive power supply capacitance are lost on my ears.
>
> > 4.) will the amp work without negative feedback from output transformer
> > to input stage (assumed the tubes are not microphonic and the wiring is
> > OK)?
>
> Yup! It will be more stable that if it had NFB.
>
> > 5.) what's the use of the serial resistors I've often seen placed before
> > the grid ? I've read something about HF-compensation. To my opinion
> > every resistor in the signal path causes "errors". May I omit them ?
>
> I wouldn't. These are grid-stop resistors. If you remove them you will
> likely introduce instabilities. They won't hurt the sound of the
> amplifier. And a signal generator and oscilloscope will demonstrate
> that to you after you've finished the amplilfier.
>
> Basically you want to build a three stage amplifier. The gain of the
> phase splitter is nil - or slightly less than nil. The gain of the
> input stage is determined by your choices of cathode resistor, plate
> resistor and plate supply voltage. Henry P. discusses load lines in his
> FAQ - and does a better job that I can here and now. I highly recommend
> you read it if you haven't already.
> You will likely want to get a reasonable amount of gain out of your
> final stage. To do this you will either have to use a fixed bias
> voltage or use self biasing with capacitor bypassing. Chances are your
> B+ is fixed so you can't vary that!
>
Where can I find Henry's FAQ ???Especially the phase splitter config (what
values for Ra and Rk? ) is is something I have some questions about (what
happens if..)
BTW I intend to use fixed bias (Ia at about 36-38mA; have to listen...).
Niko ;-)
Bob C
Buffstereo wrote in message
<19980313074...@ladder01.news.aol.com>...
>Oooops!!! I just posted a noste about 12AX7s driving EL84s in a Dynaco
SCA-35.
> Well, they don't! The EL84s are driven by 7199s. I am sorry for the
The ST-35 uses a 7247, 1/2 12ax7 1/2 12au7, driver and phase splitter.
The Acro 20-20 uses a 12ax7...
Bob
André Jute
Buffstereo <buffs...@aol.com> wrote:
> Walking and chewing gum at the same time is not an option for me!
> "Use a transistor, go to jail !!!"
> ----------------------------------------
> (VTV bumpersticker)
Clearly, it is not only novelists who have
MEMORIES OF THE FORD ADMINISTRATION
----------------------------------------
(title of a novel by John Updike)
peter kalinkov
Andy Moss <en...@ibm.net> wrote:
>likely introduce instabilities. They won't hurt the sound of the
>amplifier. And a signal generator and oscilloscope will demonstrate
>that to you after you've finished the amplilfier.
>Cheers!
>Andy
Now that's quite a statement for a tube/lo feedback lover. Give the SS
army more ammunition.
Peter
Ned Carlson
On Sat, 14 Mar 1998 23:20:50 GMT, pe...@iafrica.com (peter kalinkov)
wrote:
In PP EL84 amps instability is unlikely to be a problem,
many guitar amps using EL84 run no feedback with
no problems, I believe the Vox AC-15 and Vox AC-30
(which Mr Moss should be familiar with) are two examples.
Ned Carlson Triode Electronics,2225 W Roscoe Chicago, IL, 60618 USA
ph 773-871-7459 fax 773-871-7938
12:30 to 8 PM CT, (1830-0200 UTC) 12:30-5 Sat, Closed Wed & Sun
http://www.triodeel.com
Text file catalogs:Catalog 'Bot at cat...@triodeel.com
Sander deWaal
On Thu, 12 Mar 1998 11:32:57 -0700, Max or Paul <ho...@supernet.ab.ca>
wrote the following:
,snip>
>> ECC 81 is preferrable, it's the middle ground btw. the low-mu
>> but powerful 82, and the hi-mu, but incapable of delivering
>> enough juice 83.
>>
>
>Its also the least linear of the three -- just look how bent its curves
>are. Why does the 'power' of the ECC83 even matter if he's using it to
>drive such small output valves (EL84) in class AB1 pentode? Mullard's
>5-10 used the '83 quite sucessfully in this position, and may be a good
>starting point-- just add more output valves.
The power of the 81 is important in order to have lower
plate resistors, which will make the tube more linear, and
able to sink current better from the coupling caps (blocking).
The 83 needs fairly high plate resistors, and in a cathodyne
phase splitter, this will result in a cathode rsistor
that's also too high.
Miller effect takes its toll, also.
>> The 82 is a nice driver tube, the 83 has lots of gain,
>> and the 81 is the tube for you :-)
>
>Also consider the 6CG7 the 6SN7 and the 5687 all are very linear, and
>easily available.
I agree about the 6SN7 and 5687, but this guy wanted a simple
amp, and these tubes don't have enough gain to use GNFB.
Best regards,
SBench
>Grid stoppers. You can leave them off if there is no oscillation. I
>always leave them in. (I notice now on reading the next post that Steve
>describes them as "parasitic". I've always wonder what they feed on.
>Imagine them stuck like limpets on the signal and thereby escaping the
>amp and passing on the soundwaves through your ears into your brain
>where they do terrible things...)
Parasitic oscillation is unwanted oscillation created by the
interaction of "parasitic" components, which are defined as those
elements that happen due to geometrical considerations of
building the device, but don't show up "on the schematic",
such as cathode lead inductance, plate lead inductance,
interelectrode capacitance etc. These create an oscillator
if the circuit Q allows it to happen. The series resistance
effectively deQ's the circuit, preventing this from happening.
Nost often high gm (not hi mu) parts are subject to this. For
example 417A/5842 will oscillate at some ungodly high frequency
if you omit the 4 grid stopper resistors.
Steve