Phase Splitter

[Stretto]

I've read somewhere that the basic 1 transistor phase splitter is generally a poor choice compared to the LTP/diff amp but I can't remember why.

I am attempting to replace a LTP in a tube amp with a simple mosfet based phase splitter. LTP uses negative feedback but I'm not quite sure how to integrate it into the single mosfet. The obvious way is simply add it in at the gate as there isn't any other place to do so.

Not sure if I should add some followers/buffers on the outputs or if this will make things worse(since these are discrete mosfets that may not be matched well) or if it will help at all.

Phase Splitter Tutorial & Circuits - Junction Transistors - Electronic Hobby Projects

with R3 = R4 = 100k or maybe a bit more. R1 and R2 = 1M. It will be driving EL34's or 6L6's(2 or 4) with a grid resistance of 2k and a pulldown of 220k.

Anyways, I'm looking for reasons not to do this besides the obvious "tube sounds better" or to use a better topology. I'm under the impression that the basic phase inverter should work perfectly for generating the two signals needed to drive the power tubes and it shouldn't matter how it's done. I don't plan on driving the the PI into clipping(I hope) and so the SS version should give me a more ideal PI without the cost of a tube.

Anyone know any draw backs to the basic ss PI circuit?

IIRC there some issue with balanced drive with the basic circuit which is why I'll probably end up adding two followers.

[CBS240]

The collector is not the emitter and thus has different capacitances, and that affects the phase relation of the two balanced signals. Phase shift other than inversion creates distortion. Two op-amps using inverted and non-inverted circuits is another approach but once again the two amplifiers are different leading to phase shift error that increases vs frequency. A signal transformer might be your easiest choice. It is possible though, in my balanced SS power amp I use active biasing of a complementary bridge circuit and common mode EC to create a balanced output from SE input. It mantains phase alignment with freq >1MHz but it is fairly complex.

[Stretto]

http://www.aikenamps.com/cathodyne.pdf

seems to contradict what you have said about high frequency.

In any case the two followers should take care of any mismatch as long as they don't create their own problems. I'm using it for audio anyways so the differences in phase should be negligible. I guess the only think to do is test it out and see.