Difference between SET and push-pull using 300B

Basicaly, in a SE configuration you have one output device (tube, MOSFET, transistor, ...) which is usually working in A-class. That means that it is always open and transmits the whole "sinusoide" - actually the whole signal. It usually get very hot, and the losses to the heat are considerable. This design is probably the oldest one (with WE 300B).

Push-Pull design has two output devices at the output. Each device transmitts the half of the sinusoide. PP devices can work though in different classes. They are usually more efficient and provide more power at the output than SE. That is why they are preffered solution with more critical loudspeakers.

Typically SE is considered as superior design in respect of sound reproduction. It has though some limitations - the most important one is the power that it can provide and accordingly, the matching with appropriate speakers. PP can be also extremly successful in sound reproduction. It depends actually on implementation; output devices that are used; quality of transformers; etc. So there are seriously sounding PP designs as well as poor sounding SEs ...

I wouldn't worry though to much about all this. Listen to this amp (if possible with your speakers) and see whether you are happy with the sound. If yes ... then ... no problem which internal design has been implemented.

The primary issue between SE and PP is the output transformer, although PP often has some additional circuit complexity. But its the transformer that makes the difference.

When running PP, the signal has to go from positive to negative and then the other way. Magnetics being what they are, they resist the change in polarity. This is called hysteresis loss. In a SE amp, all that is happening is that the current through the transformer is changing intensity, but never polarity: there is no hysteresis loss.

Due to hysteresis loss, a little bit of extra energy is required to change the polarity. This energy comes from the low level signal, essentially trading distortion for signal. That is why SETs have such great low level detail. Of course, PP has bandwidth that SETs cannot match.

Take away the output transformer and you take with it the issues of hysteresis loss- and any argument for SET. Then you can have the low level detail and the bandwidth at the same time. To my knowledge though, there was only one 300b true OTL ever built, and it was a prototype. OTLs need different tubes that are lower impedance, and they have their own requirements for speakers, although they can usually drive anything an SET or other low-powered PP amp can.

Nice thread. Is it critical to have a low impedance or low capacitance speaker cable connection to reduce the loss of dynamics or drive coming from an SET?

I am struggling with keeping my SET amp which has 6 small power supply banks so it never runs out of juice on peaks but I could not believe the authority a 7 watt PP Integrated with mickey mouse caps had over my speakers.

@La45

Theoretically it could be an issue and probably speaker cables with less capacitance should work better. But I had once on test at home SET 300B (Audiogon Silver Night) connected to my speakers with Kimber 8TC - which are known for relatively high capacitance - and it worked very well. No problems ...

I believe that the careful matching with the speakers is much more of an issue, when we are speaking about SET 300B, although (and that has to be said as well) they are capable of driving speakers that would never say that it is possible by just looking on speaker figures/numbers ...

High capacitance is not that strange in a speaker cable, as the capacitance contributes to lowering a quality known as characteristic impedance. When the characteristic impedance of a cable matches that of the speaker, you get a little more efficient power transmission from the amp due to less reflections.