If you robustly provoke me then you may sometimes discover that I am imperfect. On such occasions the Mods usually put me straight, and I usually apologise.
I'm happy using feedback in my push-pull, solid state bass amp. Avoiding it for the wrong reasons may seem misguided. Maybe I enjoy the challenge of achieving linearity in a simplistic way.
I wonder that a push-pull, single ended supply can be modulated by the error signal and whether this deserves a large enough cap just the same. What of higher order distortion, ie what does push pull help, or make worse, or potentially produce?
A Class A push-pull amp will suffer from this effect just like SE. Class AB would be worse - but people building push-pull almost always use feedback to stabilise gain, so the total effect will be much smaller. Whether the supply is single rail is irrelevant.
I don't understand what you mean by error signal or difference signal in this context. The supply voltage follows the signal envelope. The amp gain follows the supply voltage (if feedback is not used). Hence the signal suffers from intermodulation, although it might not be heard as such. More likely to be heard as a matter of 'dynamics' as loud passages cause a temporary drop in gain; this could exaggerate the difference in volume between loud and quiet and so be misperceived as an improvement when in fact it is a distortion.
I don't understand what you mean by error signal or difference signal in this context. The supply voltage follows the signal envelope. The amp gain follows the supply voltage (if feedback is not used). Hence the signal suffers from intermodulation, although it might not be heard as such. More likely to be heard as a matter of 'dynamics' as loud passages cause a temporary drop in gain; this could exaggerate the difference in volume between loud and quiet and so be misperceived as an improvement when in fact it is a distortion.
Class A push-pull has potentially a constant draw between the two devices. Whether this is seen as a reason to use less capacitance in the supply, I guess it depends on who designs it. The rail voltage could be modulated by an imbalance, maybe this could be caused by biasing, or a kink in the device characteristic.
This is not entirely tru,e because second and any even harmonics generated inside the PP amplifier act in both tubes in parallel, draining extra power from the supply, and, obvious, are not 180° out of phase, they are in phase.
I don't recall a push pull amp that doesn't have an inverted and a non inverted signal input. Even though that was 28 years ago...
Then, you never has seen a push pull amplifier. What do you believe that the phase spliter do? Its outputs are 180° out of phase, and then are re-combined in the OPT.
yes but by the way you posted, I thought you said the two inputs are in phase at the push-pull amplifier's input
Output from both tubes in a push pull amplifier are in phase for even order harmonics generated inside the the PP amplifier itself, but it conserves the 180deg imbalance for input signals, whichever it be.
This is how frequency doublers used in RF circuits work: some kind of such, has its inputs in parallel and the output in push pull or vice versa.
This is how frequency doublers used in RF circuits work: some kind of such, has its inputs in parallel and the output in push pull or vice versa.
Attachments
Last edited:
With no distortion the difference between Class A SE and Class A PP is that SE puts signal current through the supply. This might mean that SE needs larger caps than PP.AllenB said:
However, with distortion both SE and PP put harmonics and IM through the supply. Harmonics are not too bad as if the supply can cope with signal then it can probably cope with harmonics too (as they will be at a higher frequency). IM is the issue, including the 'DC' from second-order which will vary like the music envelope. This needs bigger caps, or an amplifier which does not vary its gain with supply rail voltage.
So small PSU caps are fine provided you use enough feedback in your amp.
Hi,
People act as though feedback is acceptable, it isn't !! Imagine putting an after-the-fact, out-of-time signal back into a prior stage, that has already processed the musical event, and is processing NEW and fresh. different information.
Are we STILL in the dark ages, to accept out-of-time signal processing ?? Not for my ears, thank you very much.
Also the bigger the cap, the worse is the sound on a Loftin White SET 2A3 amp, proven 100 times over, the last 27 years. That is just the way it is. 🙂
Jeff
People act as though feedback is acceptable, it isn't !! Imagine putting an after-the-fact, out-of-time signal back into a prior stage, that has already processed the musical event, and is processing NEW and fresh. different information.
Are we STILL in the dark ages, to accept out-of-time signal processing ?? Not for my ears, thank you very much.
Also the bigger the cap, the worse is the sound on a Loftin White SET 2A3 amp, proven 100 times over, the last 27 years. That is just the way it is. 🙂
Jeff
It seems that feedback here is understood fundamentally wrong as something happening sequentially, as if the music signal propagates through the amp like some raw material through a processing plant, and the finished product is fed back to the input much later.
Windmills don't work that way.
Rundmaus
PS. If feedback worked like Jeff seems to imagine, it would be impossible to linearize an amplifiers transfer characteristic by feedback techniques. The widespread use of operational amplifiers in everyday life electronics and millions of cheap transistor amplifiers sold since the seventies prove your understanding of feedback as completely wrong. I would suggest some introductory reading about basic electronics instead of repeating something like "my ears prove x and y" again and again.
Last edited:
Unfortunately this is a subjective thing. Voltage sag and distortion are mensurable, and then not subjective at all.
Rundmaus, this idea of feedback as regurgitated information appears to be widespread; perhaps as an attempt to explain in a non-scientific way the (audible) effects of badly applied feedback.
However you cannot perfectly correct in the past (Input) an event of the present (Output) Time travel to the past violates Thermodynamics.
It would be a funny old world sans feedback. Not just for technology but for biological systems as well.
- Status
- Not open for further replies.