PI Snubber Question

[rp]

Does a pf cap across the plates of a paraphase have the same effect as across a LTP? Seems so, but can't hurt to ask.

[pdf64]

Similar, in that it will create low pass filters, but particular source impedance at the plates will vary according to the specifics of the circuits.
eg if a balanced paraphase, then the source impedance of the plate in the feedback circuit will be much lower than normal, so the corner frequency of the low pass filter may be higher than might be expected.
Note that even with regular CC stages, the plate source impedance is affected by whether the cathode is bypassed or not.
I suspect that even if the source impedances of the paraphase phase splitter plates is rather different, both will have the same filter corner frequency, as the source impedances of both plates are summed to form the R term of the frequency calculation.

[xtian]

Confused because I thought a paraphrase is a single triode, thus only one plate.

[didit]

Confused because I thought a paraphrase is a single triode, thus only one plate.

Cathodyne.

See http://www.valvewizard.co.uk/paraphase.html

Best .. Ian

[xtian]

Thx.

[max_lwedge]

The small cap across the PI plates in either a LTP or Cathodyne is there to stabilize the negative feedback loop to not oscillate, or “slug the pole”. Output transformers have a -3db point or pole (high frequency knee) at around 60 kHz, and can have a near 2nd order response and a phase shift of well over 90 degrees. Combine that with the PI coupling cap and you have an oscillator. The small cap across the plates should put the pole at the power tube input grid at least an octave below than the transformer’s pole to keep the amp stable.

A paraphase is a different beast. I don’t know of any guitar examples that use negative feedback. You cannot put a cap across the PI plates since there is an extra pole and zero (low frequency knee) cascaded from the inverting to the non-inverting side. It is tricky to stabilize a paraphase with negative feedback but it can be done, the Magnavox stereo is a great example. The schematic may look odd at first. The pole at the non-inverting PI plate is slugged by itself with a 100pf cap to ground. The non-inverting side is fed by the already slugged inverting PI output pole, and it uses cathode feedback to lower its Miller effect and thus keeps that pole higher than the slugged pole. That achieves high frequency stability.

You’ll notice that the PI to power tube grid coupling caps are a factor of 10 different in size. That is done for low frequency stability, aka preventing rumble or motor boating. The inverting side has a zero at about 66 Hz with the 4n7 cap and the 510k grid leak. The non-inverting side is fed by the already slugged inverting PI output zero, and its zero is about 15Hz, from the 47n and 220k grid leak. Transformer zeros are around 10-15Hz, so the 66Hz zero from the non-inverting side is well above the other 2.

[rp]

A paraphase is a different beast. I don’t know of any guitar examples that use negative feedback. You cannot put a cap across the PI plates since there is an extra pole and zero (low frequency knee) cascaded from the inverting to the non-inverting side.

Thanks for this. I removed the cap across the plates where it seemed to do nothing. On a LTP I can usually hear 100pf knocking out some sparkle. I tried the 100pf to ground as on the Magnavox 88 example and that was what I was looking for, brightness seems right, sweeter tone, but I only have a G&L to try, might be too much w/ humbuckers.

FYI here's the 5C8 schematic again in case you are curious over about a guitar amp w/ a paraphase and NFB. I had a previous post asking about how it worked.