Greatest circuit ever (or weirdest)

[kenpeter]

Differential White Cathode Follower for DAC

Twin triodes simultaneously doing THREE jobs very well:

With resistors R3&R4, this cascode acts as a unity gain
inverting plate follower.

With resistors R1&R2, this cascode also acts as a White
Cathode Follower of the average voltage. (V1+(-V2))/2
The output voltage swing may be small, but impedance
is both extremely low, and very symmetrical.... You can
definitely drive difficult loads.

The combined error correction signal at the top plate is
so strongly controlled, it tends to overwhelm any power
supply noise. You won't see any PSRipple in the output.
So its sort of its own smoothing regulator too...

-----------also note-----------

V1 sees 47K to virtual ground (between R3&R4)

V2 sees 47K to signal ground, and a very high
impedance across R1&R2 to the plate follower.
Assuming both inputs are truly differential, this
plate holds a duplicate copy of a similar signal.

So both inputs seeing the same 47K impedance...

[kenpeter]

And the .ASC file for LTSpice.
Credit to Duncan for dmtriodep.inc


What weird circuits have you been dreaming lately?

[smoking-amp]

"What weird circuits have you been dreaming lately?"

White CF from top to bottom, anti-triode from bottom to top. Feedback gains coordinated. Note hi-Z inputs.

Don

[smoking-amp]

Improved model

V1 needed some triodizing. No point in anti-trioding the top V2 unless the bottom V1 acts like one too.

[Michael Koster]

Very nice, I always wondered what I would do with a current mode
DAC. What's the CMRR? Or does it matter? Common mode impedance
is low.

"What weird circuits have you been dreaming lately?"

Who, me? Why are we all playing with cathode followers?

Augmented cathode follower of some kind. Anti-signal current mirror
on the botton, cascode-ish follower on top to maintain constant plate
voltage. Zout ~= 1/(2*gm) or can be configured for Zout ~= 1/gfs of
the P-MOSFET if you want a flat load line.

Michael

[Michael Koster]

I forget... Did we do this already?

Fixed cathode bias regulator with current mirror for a 2X improvement
in damping factor, LF extension, output power...

[Wavebourn]

I have nothing to brag;
just an ordinary pentode on MOSFET steroids:

http://www.diyaudio.com/forums/attac...amp=1241463993

[Wavebourn]

Originally posted by Michael Koster
I forget... Did we do this already?

Fixed cathode bias regulator with current mirror for a 2X improvement
in damping factor, LF extension, output power...

This is very raw, just a concept.

To do:

Temp stability? Vgs = Vg1-k?

Edit: however, power tubes have higher Vg1-K than Vgs is needed, so my concern is not valid.

Variant: large emitter under MOSFET than under a tube.

Variant 2: different resistors in emitters.

PS: I've finished soldering a prototype, am afraid to switch it on and find hidden surprises, so am reading the forum for a while...

[ErikdeBest]

Hi Anatoly!

Go, switch on the barracuda!!

Just to see if I got it: current through the 12L6 will develop a voltage over R6. Substract the voltage drop of the mosfet (about 4V?) from the voltage over R6, and divide the obtained value through the amount of current desired through the mosfet, to determine R8.

R5 is used to provide the cathode-bias, R7 to measure standing current...

[Zen Mod]

Originally posted by Michael Koster
I forget... Did we do this already?

Fixed cathode bias regulator with current mirror for a 2X improvement
in damping factor, LF extension, output power...

Papa patented that