Tube DIY Asylum Do It Yourself (DIY) paradise for tube and SET project builders.

---

[ Follow Ups ] Thread: [ Display  All  Email ] [ Tube DIY Asylum ]

---

[ Alert Moderator ]

---

MOSFETs reality sets in

Posted by BBeck (A) on February 3, 2007 at 06:01:30

In Reply to: Re: MOSFETs? posted by Eli Duttman on February 2, 2007 at 21:36:03:

Eli,

I would agree that the only sensible way to operate a MOSFET follower (or a MOSFET gain stage for that matter) is with plenty of Vds (>>20 V in the case of your part at worst case signal swing), where the magnitude of Cdg drops, as does it variation. It’s hard to say from the data sheet how much capacitance variation remains because the axis scales don’t allow much visibility into Crss (=Cdg) when Vds is above 20 volts. Would 20% variation of Cdg across the signal swing matter, 10%, 5%? I just don’t know. The driving impedance matters in determining the phase shift caused by this capacitance, of course. The question then becomes: have these effects diminished enough so that the MOSFET follower is now audibly indistinguishable from a comparable triode follower? (“Comparable” is a loaded word, of course.) If it’s not audibly exactly the same, then there has to be some explanation for the difference.

On a related note, much has been made of the far superior gm of a power MOSFET compared to most hot rod tubes, and therefore of its superior driving ability in a follower configuration. This is mostly true, but maybe less so than it might seem. Take the well-worn IRF820 for example (similar arguments will apply to the IRFBC20). The IRF820 is specified as having a minimum gm (=gfs) of 1.5S at 1.5 amps of Id. That’s 1,500,000uS (=1500mS) folks. Fantastic, but only at 1.5 A! But we aren’t going to bias our driver at 1.5 A, are we? So what happens to MOSFET gm values at lower, more reasonable currents? In another forum, someone measured an IRF820 follower output Z when it was biased at 10mA to 25mA, a reasonable current range for either a tube CF or a MOSFET follower. The output Z value was then about 15 to 20 ohms. This corresponds to gm values of around 50,000uS (=50mS) to 67,000uS (=67mS). That range of gm values is quite high of course, but no where near the spec value that lured us to begin with. And, importantly, it is not orders of magnitude better than what a few high-gm tubes can muster at comparable currents (never mind paralleling driver tubes). My point is that the enormous gm possibilities in a MOSFET are only available when there is significant amperage flowing through the drain. So if you decide to bias your MOSFET driver at high values of Vds to minimize Cdg modulation (per the above comments), AND also crank a lot of current through it (say a few hundred milliamps of Id) to pump up the gm, you could be dissipating more power in your driver MOSFET than in your power tube! I’m not dismissing MOSFETs driver out of hand, only urging caution. I also have to wonder why people would struggle with these devices (under these design constraints) for relatively little advantage when there are hot-rod tube follower possibilities (including, but not limited to, White Followers) that we can consider.

This post is made possible by the generous support of people like you and our sponsors:

---

[ Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2025, Audio Asylum, all rights reserved ]

---

[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]

---