Considering A 20-40 Watt Stereo SET Amp

I believe that Canary Audio makes both zero NFB SET and push-pull 300b amplifiers. I don’t know if anyone has done a direct comparison between these two versions.

@charles1dad I have to imagine Canary has... PP amps, just like SETs, vary quite a lot depending on design concerns! For example, single-ended circuits exhibit in mathematical terms a quadratic non-linearity, which describes how the harmonic distortion arises. When measured, it is found that the math accurately predicts the harmonic content.

OTOH, a fully balanced differential circuit exhibits a cubic non-linearity. In this case the even orders are cancelled; ideally only odd orders remain, but they fall off at a much faster rate as the order of the harmonic is increased (with the 3rd harmonic masking them). This again is accurately predicted by the math.

So this is where I'm going with this: when you combine both single-ended and PP circuits in the same amp (such as seen in almost any traditional PP amp, like a Dyna ST70), you get both non-linearities. This tends to exacerbate the 5th harmonic (Norman Crowhurst also wrote about this problem).

So if the comparison is between an SET and and a traditional PP amp, you are getting more 5th than you would if the PP amp was fully balanced. I've no doubt this is a lot of why SETs can be preferred.

These are all readily solved engineering issues though- its not hard to build a PP amp that does not have that problem! This is part of what I meant when I used the phrase 'cast a wider net'.

I have owned two terrific push pull tube amplifiers, 100 watt KT 88/6550 and 40 watt el34. I have 300b SET mono blocks 8 watts. All three work very well with my 94 db sensitivity/14 ohms speakers. The SET is the most nuanced, open, transparent and in addition the most emotionally involving, 3-dimensional and tactile of the three amplifiers. Simply more real and believable presentation. That’s my listening experience. For other listeners if can certainly be different.

@charles1dad 

I don't doubt your experience.

Those larger amps you mentioned use feedback. A long time ago I read a passage by Norman Crowhurst (an early tube guru), discussing what happens when feedback is applied in an amplifier. He described the spot that occurs as the 'feedback node'; the cathode of the input tube. The thing is, that tube, no matter how good, isn't linear. So when the feedback is mixed with the incoming signal, its a distorted by the tube, so causes additional distortion as a result. Crowhurst described that as 'higher ordered harmonics, inharmonic distortions and intermodulations'. All of these are unpleasant to the ear and readily audible (usually has harshness and brightness at some level).  Those things did not go away simply because the amplifier is a bit newer...

A simple remedy is to not mix the feedback inside a tube. It can be done using a resistor divider network entirely outside the amplifier. It helps a lot if the amplifier is also reasonably linear without feedback! This practice is exceedingly rare in high end audio.

Most pentode and tetrode power amps I've seen really need that feedback to work. But if the amp was already musical without it?? That's pretty rare... I'm not contesting your experience. What I'm saying is to find out how this works, a wider net has to be cast.

I started studying SETs in the early 1990s (with early editions of Sound Practices magazine as the influence). I've heard some SETs that seemed quite nice, but so far none surpassed some (but certainly not all) of the PP amps I've played. 

IIRC, your SET uses a 300b. You might find it interesting to hear a class A PP zero feedback amp using the 300b. In that way variables are eliminated, all of which affect the distortion and therefore the 'sound'.

@stereo5 What you are up against is that its really hard and really expensive to make power with SETs. By really hard, one of the aspects of that is bandwidth goes down the more power you need (due to output transformer limitations). That is why the smallest SETs have the reputation for sounding 'the best'. By the time you get over about 7-8 Watts (300b territory) you have trouble getting the amp to have what might be considered 'hifi' frequency response.

When people compare SETs to PP amps, its usually an apples and oranges thing! That is because the SET has no-where near the power of the PP amp, and the PP amp doesn't have Directly Heated Triodes, like nearly all SETs do, and likely runs feedback and pentodes as well.

To put that PP amp on a level playing field with an SET, they should have the same or similar output power at clipping. People don't do that- because there are almost no PP amps that only make 5 to 7 Watts! And no-one even considers putting a 30 Watt SET against a 30 Watt PP amp. It would get creamed. That's mostly because at that power level, the bandwidth limitation imposed by the output transformer is readily audible. And measurable as well... and the PP amp would have a much greater amount of usable power...

RE usable power: SETs rarely use feedback so their usable power is up to about 20% of their rated power. Above that power level, they take on a 'dynamic' quality as higher ordered harmonics show up on musical transients (where the power is needed). The ear uses the higher ordered harmonics to sense how loud a sound is, so when they are only on the transients, you get a 'dynamic' quality.

Dynamics should come from the signal, not the amp.

If you are not willing to replace your speakers, which are not efficient enough to really work right with most SETs, then look at a PP amplifier. There are triode PP amps that run class A with zero feedback, FWIW. IMO/IME they sound better than SETs too: more transparent (due to lower distortion), more authority (due to better bass power response) and overall more true to the signal- the music.

SET guys won't like this of course, but I challenge any one of them to a simple set of questions: When did you compare to a PP amp of the same power? Barring that, did you compare to a PP amp using the same power tubes and construction technique (and parts)? I've done both. I can go into in depth...