Phase Inverter comparisons

[raiken]

A quick PSpice sim of a 550V Vao cathodyne yields the following results:

Va (quiescent) = 413V
Vk (quiescent) = 138V
Ia (quiescent) = 1.366mA

This gives a Va of 275V, so it is within the 300V Va(max) rating. Static anode dissipation is 375mW, so that is okay, and cathode voltage of 138V is within the max 180V rating. It also looks like it can handle about a 200V input before the anode waveform starts compressing at the bottom.

[MWJB]

What about the preceding gain stage, given a traditional cathodyne inverter topology?

[raiken]

What about the preceding gain stage, given a traditional cathodyne inverter topology?

It can be whatever you want, since the cathodyne is typically AC-coupled in and self-biased. Anything that will give you a clean 100V swing will drive it to an undistorted output of 100V on each side (assuming you even needed that much - in many cases you wouldn't need anywhere near that much swing).

If you want to design a DC-coupled driver stage, you just need to make sure the quiescent DC voltage at the plate of the driver stage doesn't allow you to exceed the 180V H-K rating of the cathodyne stage.

[defaced]

What about output impedance and the ability to drive big tube? I'm working with an amp that uses 4x KT-88 which is currently setup with a cathodyne but the original design used post PI buffers and I'm wondering if I can forgo those if I switch to an LTP.

[Zoe_N_Iain]

I'm not going disagree anything there, that's perfectly reasonable. You don't even "need" a plate resistor that big (obviously what sounds best is best on that front)

Quite true. In actual fact, I have quite a big cathode resistor on the 340k stage, so I still use a bypassed voltage divider to get the cathodyne's grid voltage where I want it. Like the sound better than any other way I've tried, really. And the cap is handy for rolling a bit of bass off if it's necessary, which it sometimes is.

My only observation is that 450v supply at the PI node is not particularly high.

No it isn't, though it's about as high as 2x 6L6GCs @ 500v will allow with the 4uF polypropylene caps that I used for preamp/PI filtering on those amps. You need a decent amount of series resistance to ensure stability, and the preamp is pretty current hungry (not 12AX7) so there is naturally a reasonable voltage drop across the resistors.

I haven't at any point suggested that it's not a good idea...

You intimated it, though (not having a go at you, just pointing it out)>

"A cathodyne PI usually has both triodes run from the same node and plate voltage at the first might therefore be limited (<200v?)"

"I stated that cathodyne PI's gain stages were "rarely run much above 200vdc on the plate"

While agree with you that classical amps may not attempt to run cathodynes at high B+ voltages, that doesn't mean that you can't. Or at least moderately so, if you don't consider a 500v B+ high for a guitar amp (I grew up on valve TVs and later radar, so I consider anything under 10kV to be a low voltage, to be fair).

I think the whole discussion got unnecessarily heated, simply because Merlin was trying to point out there's more than one way to skin a cat and you got the hump with him over it.


You want some more heresy to go at?

I like Marshalls better with 6L6GCs in the PA and less NFB than usual to drag the breakup point down to about the same place.

I use transistors everywhere they are a better choice than valves. And with a bit of imagination and experimentation that's a lot of places. Purists and valve snobs be damned.

I like pentode preamps, but I set them up with sensible plate resistors, so they don't GET microphonic. A small signal pentode is good for only slightly more gain than a 12AX7 if you want it to work properly for the same 5+ years. And 6BR7s sound better, are cheaper NOS than EF86s are new, and can give more gain without microphonics anyway.

I eschew inline tone controls, at least some of the time, and instead do fun things like biasing triodes with inductors and caps. And Bridged-Ts for mid controls (parametric, too! Much more useful than your FMV 'not much good, but very lossy' affair).

I don't like voltage dividers much. They just add noise. Why not just use a triode with half the gain, if you're going to throw half away, anyway? What's the point. If you amplify a signal, why not use all of it to do something useful?

I bias valves with LEDs and diodes.

I don't use electrolytics, unless it's absolutely necessary (it often is, I will concede).

Anything else I can upset you with?