Interesting MOSFET question (hiwatt)

[jon]

I'm interested in trying the Hiwatt dc coupled long tail pair phase inverter in an amp. But I'm short one triode. Could I use an irf820 in place of the follower directly before the pi in this circuit? Having read the MOSFET follies article over and over I think it would work.

[Tillydog]

I believe that you could drop one straight in there and it would work OK.

What I have reservations about is what it would do to the frequency response - the input capacitance of the IRF820 is 360pF (from the Vishay datasheet) as compared to a few pF for the triode.

I haven't worked the numbers, but I would expect dropping another 350-odd pF after that 1M8 and 22n would have an audible effect (however it should be possible to adjust those values to compensate).

Or... a late thought: maybe because it's in a source follower configuration, the effect of the capacitance will be negligible.

I will have a think!..... (But it will "work").

[jon]

I might have to give it a try.

[jaysg]

Yes. I did it in my Alembic F-2B.

[Tillydog]

...maybe because it's in a source follower configuration, the effect of the capacitance will be negligible.

That^

I've just run it through LTSpice, and the MOSFET follower doesn't make a jot of difference compared to the triode (even hanging a big capacitor between gate and source or grid and cathode of the follower makes very little, if any difference).

As you were. Sorry for scaring the horses!

[mhuss]

Yeah, that will work fine. The triode is just an impedance buffer between the fixed voltage divider reference and the PI.

--mark

[Cameron]

The other style PI using a high-voltage transistor....

[Jerryz1963]

I read the article too. very interesting. He addressed that and, as I recall, what the author stated was that since it is being used as a buffer, the apparent capacitance is actually extremely small, regardless of what the actual gate capacitance is. I have read that many have used MOSFETs as buffers with no adverse tonal consequences. As I recall, the only major caution is a diode or zener to protect the gate.

[Mescalero]

Additional capacitance, including that of a Zener or TVS, will nominally increase the required drive current at high frequencies. There is more than sufficient drive current to feed 360pF.

[Jerryz1963]

I didn't bookmark them, but if you do a general search about mosfets or FETs in general versus tubes, get creative with your search terms, there are several articiles about subbing in MOSFETs or FETs in general in place of tubes. It was one of those articles that inspired me to change a tube tone control to JFET, Took a little futzing around, but the final result was better. The tubed version (as the vendor candidly admitted) had a hum problem reported by many of his customers. converting to JFETs solved that and the tone control seemed to work better. Not surprising since the Baxandall works best with a very high gain device.

Anyway, the upshot of it all was that, in many cases, you can directly sub in assuming your sand device can handle the voltage. Otherwise, you just have to make some simple adjustments