• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Global Feedback in Push Pull Amps

I'm on my 5th and 6th Push Pull Amp projects.

Popular wisdom is that ANY global feedback slurs the phase response and destroys the stereo image.

In every case, however, I have found that a very small amount of global feedback (3 to 5 dB) actually improves the stereo image and makes it more focussed without making it any smaller.

Does this agree with your findings?
Is your experience different from mine?

Cheers,
Ian
 
in SE also, I think

I run some experiment on a 6EM7SE and EL34SE, in my opinion a little overal feedback (actually 6dB) gives better balanced, equilibrated sound. Little impression, averaged over several weeks and with different music, but on the average I and a friend of mine prefer with feedback. Maybe because I used in both cases self winded transformer, surely of inferior quality to the commercial ones.
 

dhaen

diyAudio Moderator Emeritus
2002-06-10 12:01 pm
U.K.
www.keystrobe.co.uk
My experience with messing about with feeback on PP amps is limited to one:

It was a KT88 own design. I arranged a feedback point in the cathode of the input valve. The circuit was such that rotatong the pot caused the feedback to vary between 0 and about 20dB, without change in overall gain. The amp was stable over the range.

With full feedback the sound was tight and sharp. Not dissimilar to a good SS amp.
With feedback at 10dB, it started to get looser, without any improvement.*
As the feedback was wound from 10 to 0, something strange happend: the image became substantially more open and real.

*On further investigation the distortion spectrum at this value had lost most of the even harmonics, the odd harmonics were still present.
 
I have almost same experience:

The PP amplifiers need some global feedback. The bass is not well controlled without feedback, if the output tube is pentode or tetrode.
So in this case I used 15-20dB feedback to get good bass response.
With ultralinear of triode, some 12-14dB is enough. This will gives better controlled sound, without slowing down the amplifier.
One thing is common: You need properly designed phase response, if You want to use global feedback. If no, You can get some oscillation, even for short times, depended by the input signal.
But global feedback is not solution for everything. If Your amplifier is not well designed (high distortion, poor phase response) the global feedback will not helps to repair it!

sajti
 

analog_sa

Member
Paid Member
2002-08-14 1:47 pm
Cascais
With full feedback the sound was tight and sharp. Not dissimilar to a good SS amp.
With feedback at 10dB, it started to get looser, without any improvement.*
As the feedback was wound from 10 to 0, something strange happend: the image became substantially more open and real.


In full agreement. Pentode amps present a real dilemma - no feedback is obviously better for dynamics and images but the bass suffers and the distortion is often obvious.

Triode amps, otoh, consistently sound better without global NFB.
 
Quote:
The PP amplifiers need some global feedback. The bass is not well controlled without feedback, if the output tube is pentode or tetrode.
So in this case I used 15-20dB feedback to get good bass response.



This is where, a little lateral thinking can be applied, how about a cathode feedback winding from the secondary of the the output to the output tubes to keep the output impedance down. And still achive good damping factor.
And no global feedback around the amp at all, to keep those mids and highs un-corrupted and true to life.

http://www.diyaudio.com/forums/attachment.php?s=&postid=554384&stamp=1106011421


Cheers George
 
I went from global feedback to none in UL PP amps. I preferred the natural, easy sound without feedback, but even in UL I couldn't get rid of a rather etched sound. When I went over to triode outputs the sharpness disappeared, leaving just a smooth, natural and very detailed sound. I'd never go back to pentodes or global feedback. To get the watts I just parallel up the smaller triodes - 12b4, 6S4A, 6CK4, 6AH4 and eventually 1626 when I get round to it (it'll need a few!) . You can use the double triodes too - 6BX7, that kind of thing. With a bit of lateral thinking you can use a whole stack of real triodes, and the sound - for me - is just on a higher level. I'm happy.
 
georgehifi said:
Quote:
The PP amplifiers need some global feedback. The bass is not well controlled without feedback, if the output tube is pentode or tetrode.
So in this case I used 15-20dB feedback to get good bass response.



This is where, a little lateral thinking can be applied, how about a cathode feedback winding from the secondary of the the output to the output tubes to keep the output impedance down. And still achive good damping factor.
And no global feedback around the amp at all, to keep those mids and highs un-corrupted and true to life.

http://www.diyaudio.com/forums/attachment.php?s=&postid=554384&stamp=1106011421


Cheers George


The cathode feedback results high level of driving voltage. Large local feedback can results even 80-100Vrms, which can results distortion in the driver stage...

sajti
 
andyjevans said:
I went from global feedback to none in UL PP amps. I preferred the natural, easy sound without feedback, but even in UL I couldn't get rid of a rather etched sound. When I went over to triode outputs the sharpness disappeared, leaving just a smooth, natural and very detailed sound. I'd never go back to pentodes or global feedback. To get the watts I just parallel up the smaller triodes - 12b4, 6S4A, 6CK4, 6AH4 and eventually 1626 when I get round to it (it'll need a few!) . You can use the double triodes too - 6BX7, that kind of thing. With a bit of lateral thinking you can use a whole stack of real triodes, and the sound - for me - is just on a higher level. I'm happy.

Currently I use 12dB feedback in my PP triode connected 6550 amplifier. I found that the sound is more smooth, and the bass more powerful with it. I thinking to avoid the feedback, but I have redesign the driver stage, with lower gain.

sajti
 
sajti said:


Currently I use 12dB feedback in my PP triode connected 6550 amplifier. I found that the sound is more smooth, and the bass more powerful with it..

sajti


Hi there......on most p-p amps you will also need global nfb (out tranny to cath input tube) to get the signal to noise ratio down to acceptable levels.....a conventional 3 stage p-p amp with a pentode on the input and cathode coupled long tailed phasesplitter can see s/n down
to -85dB/W using 25-30dB global nfb; while a 3 stage design using a see-saw paraphase splitter will hover around -70dB/W with 15- 20dB global nfb.
The thd between the two examples at 1KHz is typ 0.05% vs 0.25% resp...

Ironically it's the latter that sounds the better...the bass is looser, still well damped and sounds generally less tight.

The bass from a p-p amp with no global nfb can honk/boom with a reflex LS and result in uncontrolled overshoot. Why ? The driver impedance will rise at the LF end and won't be damped by the output stage....Most well designed p-p tube amps with global feedback loops will provide 10-15dB feeback at the LF end to damp LS driver resonance. Increasing this nfb by decreasing the interstage coupling cap values, and also nfb component values one risks LF amp instability leading to very low frequency oscillation and problems with long leads at the HF end.......The o/p transformer parameters and interstage coupling caps are highly influential.
Unless you are confident in designing a p-p tube amp from scratch, My motto is stay with a design that has been tested with time.


richj
 
Here's a design that povati posted in another thread, that is push pull and uses cathode feedback which should give good damping factor still, even after the global feedback is removed. Does'nt look too complicated, the drive and input are paraleled 12ax7's.

http://www.audiodesignguide.com/ba/vtaf.jpg

Cheers George
 
Re: some good reference about the 3D subject

plovati said:
A fundamenta one abot distributed cathode feedback:

Another way to use cathode feedback (positive) to avoit cathode cap:

http://213.156.45.208/primo/materiale/20_010_050_001_Koch.pdf

bye

Hi there.........in the end ....the same results acheived in the Quad p-p amps with cathode feedback winding put amongst the UL windings ? ? The leakage inductance would be the same.

richj