What if O.H. Schade had MOSFETs?

[Michael Koster]

Maybe it's not the best bedtime reading (or maybe it was the spicy BBQ
late at night) but as I was reading "Beam Power Tubes" by O.H. Schade
and contemplating the Eimac 4-65A tetrodes lying in a pile of wood
shavings on the workbench, I sketched a tetrode with plate feedback
and a high impedance triode as the lower leg of the grid divider and
called it a night. What a silly way to use a triode, I thought. And I want
grid current...

Well, by morning the triode had morphed into a pentode, then into a
MOSFET, then the tetrode itself morphed into a MOSFET, with the
following connections.

If feedback can turn a pentode into a linear triode, why not start
with a MOSFET?

I think the basic relationships are similar to what Schade describes
but I make some simplifying assumptions below. I also left out some
protection diodes. Probably some errors also... But in the context of
plate feedback discussions elsewhere I thought it would be of interest.

Cheers,

Michael

[kenpeter]

You are only now discovering this?

[Michael Koster]

- You are only now discovering this?

I guess...

Thought it would be fun to play with and build a one-stage SE'T'
amplifier with only local feedback. The idea is to use it as a triode
in-circuit.

Do you have a reference for prior work? I couldn't find anything on
a quick search.

Michael

[consort_ee_um]

I wondered if this was the same Schade who did the classic work on rectifiers and it was. Schades curves avaliable here page 118 onwards Also an example of Power supply design for an audio amplifier page 126
http://www.ieeta.pt/~alex/docs/Appli...20Handbook.pdf

[smoking-amp]

Hi Michael,
I think that circuit is pretty much what Ken was calling a Fetron or Fetrode. I think he said he found it somewhere else too. There was an article in Electronics World some time back on making a SS sub for tubes, not sure if that would be the original.

OT:
ever get any info from Bob Carver on his mysterious 6AL5 DC restorer?

Don

[kenpeter]

If nothing else, Nelson Pass' "Zen" is the same basic circuit.
And his slightly overcomplicated "Aleph" current source an
Anti-Triode (or Anti-Fetron)...

I doubt Nelson designed either with any regard toward Triode
or Anti-Triode emulations. Its still a good circuit, no matter what
angle of sandy or hollow state thinking brings one around to it.

More than once I've posted P-Channel Fetrons for discussion.
Seeing no need to offend the glass world with a redundant N-
Channel competitor.

And the same circuit bent around a GTBT as a potential stand-
alone emulator for an Anti-Triode (not requiring a reference).

[kenpeter]

Copy and paste: one of many days I wasn't quite thinking straight.
Only first of these three circuits halfway seems to make any sense...
Often finding my own past ramblings as difficult to comprehend as
anyone else.

[kenpeter]

And another nutty idea gone entirely wrong...

I think my intention was P-Fetron on the left side,
and Brute bridged P-Antifetron on the right side?
Abuse a bifilar dual voice coil sub as a very low
impedance PP transformer to directly drive a full
range loudspeaker in tandem.

Or maybe the driver side was the P-AntiFetron?
And the passenger side then P-AntiAnti? I forget
now what bizarre point I was trying to prove by
this exercise???

------------------------------------------

Anyways, the point I'm trying to make here today:
That they all share a simplified version of Schade's
local feedback scheme. A resistive divider repairing
the "broken triode" hidden in Tetrodes, Pentodes,
and certain sandy devices.

[Michael Koster]

Hi Ken,

I just returned from a similar archaeology expedition, and found
indeed a number of your fetron references wrt. O.H. Schade, which
is apparently what I just recreated. Now I know what you were
talking about all this time!

I'm a little curious why emulating 3/2 power is important. What's
wrong with a device that has more constant mu and Rp over it's
op range?

I didn't call it a fetrode because that's a registered trademark
of a biosensor company.

So at least it looks like I've distilled it down to a few parts with a
simple explanation ;-) I think it can make a practical one-element
amplifier with a nice triode characteristic, as well as a teaching tool
about triodes and pentodes for people raised on solid state...

Also this simplified circuit can be directly mirrored to P-version,
with an n-buffer and p-gain element.

I'm expecting my plate curves to look a little different, though...
more like DHTs; very linear down to the bias current baseline.
I think I'll at least try to get some curves from an actual circuit here.

Cheers,

Michael

[Michael Koster]

Don
- ever get any info from Bob Carver on his mysterious 6AL5 DC restorer?

First, an interesting assertion that trying to negative peak detect
the common cathode current would fail because of common mode
current. (I actually did have a problem with a positive feedback of
some kind that appeared to be related to common mode current)
I wonder if he follows these groups...

He didn't elaborate further except to say that balance control
wasn't as important as bias stability was, that anything with a time
constant has a negative impact on music, that sampling schemes
aren't reliable and don't account for drift, and that his scheme
continuously corrects zero-signal current without a time constant.
Apparently *not* by negative peak detecting cathode current minima.

So what could a vacuum diode do uniquely to this end?

I should try again to follow up but have been working on other
pursuits. I do agree with avoiding sampling and time constants and
also think continuous correction is possible. Maybe as I get more
interested in class AB operation I'll look into it again.

Michael

PS FETRON may not be a current trademark, but was apparently
used by Teledyne (an ex-employer of mine) as a trade name for
a solid state FET based replacement for vacuum tubes. They were
used in telephone equipment and found some use in guitar amps
(Mesa Boogie). They were packaged in 9-pin and TO-3 like cans.
So FETRODE and FETRON are already well known names for other
devices (though the FETRON is a similar use, it was a high voltage
depletion mode FET, not a Schade application of local feedback).
So I'll stick with MOStrode for now. I can only find MOST RODE
motorcycles...