• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Which high-voltage MOSFET for source follower?

I'm about to order some parts from Mouser for a project, and wanted to get a few MOSFETs for source follower use. I have the ZVN5045 part used by Eli D for the RCA-redux RIAA, and I have some IRF510 for low voltage situations.

I'll get some IRFBC20 for high voltage source followers, but is that the best part these days? I remember an FQP-something part with a really long number, which some had recommended for its very low gate-source capacitance. Is IRFBC20 still a very good choice?

I'm thinking of using a MOSFET as a split-load phase inverter (fetodyne?) with a paralleled 6SL7 or 12AX7 as the voltage amp. I'd want the 'fetodyne' to work with about 10mA, maybe more, to drive push-pull power triodes (2A3's, EL34-triode, KT88-triode, GU50-triode, PPP EL84-triode, etc.). But with a wimpy 12AX7 driving it, the input capacitance will need to very low.

Suitability for use as a source follower for driving power triode grids is desirable (enhancement mode).

Another use would be for a Starving Student Hybrid Headphone Amp type thing but with tube-friendly B+ of about 150 to 250V. Again, low input capacitance will be needed.

Any favorite MOSFETs for this kind of application?
--
 
Last edited:

exeric

Member
2008-03-24 6:23 am
Maybe you have read this recent thread I started:http://www.diyaudio.com/forums/tube...other-semiconductor-substitute-cathodyne.html

The ones in that thread are the only ones I've heard about. If you decide to go the "fetadyne" route make sure you bias the source at 1/4 the voltage applied above the upper load resistor. It makes a huge difference in reducing the upper harmonics, beyond giving you the maximum potential voltage excursion across each split load.

That's way more important than direct coupling the fetadyne. You can do both, of course, but one has priority over the other. ;)
 

jcx

Member
2003-02-17 7:38 pm
..
the problem is getting small die (for low parasitic C) in high enough power packaging

using 1 A or higher Idss die for 10 mA isn't going to get you low C, although 1 A may be the lowest many manufacturers offer so you may have to sort through data sheets for low C versions
 
Maybe you have read this recent thread I started:http://www.diyaudio.com/forums/tube...other-semiconductor-substitute-cathodyne.html

The ones in that thread are the only ones I've heard about. If you decide to go the "fetadyne" route make sure you bias the source at 1/4 the voltage applied above the upper load resistor. It makes a huge difference in reducing the upper harmonics, beyond giving you the maximum potential voltage excursion across each split load.

That's way more important than direct coupling the fetadyne. You can do both, of course, but one has priority over the other. ;)

Has this been confirmed in real life or based only on a simulation?
 

exeric

Member
2008-03-24 6:23 am
Has this been confirmed in real life or based only on a simulation?

It's based on two things. My own simulation and a comment (sorry I don't have the thread at my fingertips) by Trobbins of an article that said IM distortion was reduced if the bias was set that way. That seems to be consistant with what I was simming. The split load voltage excursions seem to modulate each other a bit if it is not biased in that way.
 
Last edited:
I've been using the FQPF2N60C lately. It is a good part, cheap, and has a plastic tab so there is little concern with shorts.

Thanks. I bought some. Does the plastic insulated tab mean it can be screwed to the heatsink without the use of a mica spacer kit? Just thermal grease and that's it? That would make things a little easier.
--
 

exeric

Member
2008-03-24 6:23 am
It's based on two things. My own simulation and a comment (sorry I don't have the thread at my fingertips) by Trobbins of an article that said IM distortion was reduced if the bias was set that way. That seems to be consistant with what I was simming. The split load voltage excursions seem to modulate each other a bit if it is not biased in that way.

Here's the article that I think Trobbins was referring to:http://dalmura.com.au/projects/williamson_verstaerker.pdf
 
Does the plastic insulated tab mean it can be screwed to the heatsink without the use of a mica spacer kit? Just thermal grease and that's it?
--

That's what I do.
 

Attachments

  • ssp2.jpg
    ssp2.jpg
    123.1 KB · Views: 554
Toshiba TK10E60W would be worth a look - although its gate capacitance (700pF) is higher than the IRFBC20 (400pF) this is largely bootstrapped in a source follower configuration. The other two capacitances are lower. Its also got half the thermal resistance (junction to case) of the IRF so will run cooler and hence potentially have higher gm.

<edit> On second thoughts, scratch what I wrote about the capacitances - the headline figures are given at different voltages (25V vs 300V) - this makes a really substantial difference in favour of the IRF on the output capacitance.
 
Last edited: