Not quite YATS...

earthtonesaudio · Post by **earthtonesaudio** » 20 Jan 2009, 13:21

Inspired by the info on synchronous rectifiers. If I'm imagining this correctly, the "gain" ought to go from less than one, up through "screamerland," and beyond into comparator/square wave fuzz territory... On second thought maybe that voltage divider ought to go higher than Vref. Some experimenting is needed.

Basically you have voltage control over the antiparallel MOSFET channel resistance.

RnFR · Post by **RnFR** » 20 Jan 2009, 22:47

is it possible to do something like this in a hard clipping format?

floris · Post by **floris** » 20 Jan 2009, 23:07

Are you sure the Mosfet is used as a variable diode?
I think it is now a variable resistor which means that it only controls the gain factor, not the clipping.

earthtonesaudio · Post by **earthtonesaudio** » 21 Jan 2009, 01:54

RnFR wrote:is it possible to do something like this in a hard clipping format?

Depends on what you mean by "hard clipping." If you mean diodes to ground, no. Not easily anyway.
If you mean diodes in the feedback loop of an inverting amplifier, like the Big Muff, yes. Just put the (+) input to Vref and put the (-) input to signal instead of ground.

floris wrote:Are you sure the Mosfet is used as a variable diode?
I think it is now a variable resistor which means that it only controls the gain factor, not the clipping.

You caught me. They are indeed wired as variable resistors. But there's also the body diode in each FET that gives you the clipping. So the FETs take the place of the diodes and the gain resistor. When they're turned on fully, it's basically a unity gain amplifier, and the body diodes are effectively shorted. When they fully off, it's effectively at diode-limited open-loop gain.

I just thought it would be cooler to call it variable diodes.

floris · Post by **floris** » 21 Jan 2009, 13:14

They are indeed wired as variable resistors. But there's also the body diode in each FET that gives you the clipping. So the FETs take the place of the diodes and the gain resistor. When they're turned on fully, it's basically a unity gain amplifier, and the body diodes are effectively shorted. When they fully off, it's effectively at diode-limited open-loop gain.

I thought that the gate had to be connected to the drain or to the source of the FET to be able to use it as a diode.
Connecting the gate to the source lets you use the FET's "intrinsic body diode" which acts like a regular silicon diode (so I'm told).
Did you try it?

earthtonesaudio · Post by **earthtonesaudio** » 21 Jan 2009, 13:35

In this configuration, I think the body diodes are active anytime the signal going through them is higher than 0.7V, assuming the channel resistance is high enough.

The gate-source clipping can't be used here unless you rearrange things or add blocking diodes.

I still haven't tried it. Shame on me!

Anyway here's an alternate version I think would work similarly: