Mosfet's as clippers

[Khas Evets]

I always thought that using the diode on the mosfet would react as if it were a silicon diode, but there seems to be more to it that just that. This has been discussed tangentially in several threads, but I haven't seen it focused on specifically.

http://www.diystompboxes.com/sboxforum/viewtopic.php?t=244
http://www.diystompboxes.com/sboxforum/viewtopic.php?t=24884

So what is actually happening? Is the clipping threshold different? Is the knee different? Is the capacitance in the mosfet affecting the sound? What is the reason for putting it in series with another silicon diode (other than raising the threshold)?

[aron]

Try searching the first stompbox forum for MOSFET.

Tons of threads about the MOSFET clippers.

BTW: Credit John Greene for the use of the MOSFET clippers. Although I did use them "backwards", it was his inital idea that prompted the use of the MOSFETs.

[Paul Marossy]

I built one of Joe Davisson's circuits called the "Amber Boost" that used MOSFET clippers and CMOS opamp. Cool sounding circuit. I think they sound a little more tube-ish than run of the mill cipping diodes.

[amz-fx]

Is the clipping threshold different?

Depends on whether you are using the gate-source junction or the drain-source reverse diode.   The first is much larger than the second, which is usually around 0.7v

Is the knee different?

Not really

 Is the capacitance in the mosfet affecting the sound?

Not appreciably

What is the reason for putting it in series with another silicon diode (other than raising the threshold)?

To steer the signal -  and therefore the clipping - through the drain-source diode

None of the methods shown online are exactly the same as the setup in the OCD...  and there are other ways possible as well....

More on diode clipping, including harmonic analysis, on my cd-rom at http://www.muzique.com/amz-cd.htm

-Jack

[aron]

So what is actually happening?

I don't know but despite all that has been said,

it _sounds and feels_ different.

Try it yourself and listen.

[lion]

I'm getting a little confused - but if I've got it correct, there's 3 ways of using a mosfet as a diode:
G+S - D = normal silicon diode
D+S - G = low leakage, high res diode
G+D - S = diode conneted mosfet = very soft knee = as used i the Shaka B.

Questions:
1) What exactly was John Greene's setup?
(I have a note from '98 stating that he used a VN0300(?) and RG said about his findings: You may have the  next quantumstep extension to the TS style distortion here)
2) Whichs is the resulting diodes anode/cathode for each of the above mosfetconnections?
3) Vf? Measuring the diode forward voltage - how?

Erik

[WGTP]

Check out R.G.'s response on pg. 2.  

http://www.diystompboxes.com/sboxforum/viewtopic.php?t=23474&start=15

[gez]

I came up with a circuit similar to the Amber Boost a few years ago, but I wired it up so that the thresholds of the MOSFETs were adjustable, this means you can get asymmetrical clipping and also the transition from distortion to clean happens a longer way down the line so it's more transparent (I have weak single coils and the high thresholds of the MOSFETs was a real problem).

It requires one n-channel and one p-channel though. The sources are tied to the output of the op-amp and the drains also tie together and connect up in series to a pair of back to back diodes in the negative feedback loop. If you replace the 10k divider for the op-amp with a pair of trimpots then the gates of both FETs connect to the wipers of each pot. The p-channel's gate connects to the wiper of the pot with one outer lug connected to ground, the n-channels gate to the wiper of the pot with one outer lug connected to 9V.

It's tricky to set up but worth the effort. Set both wipers so that they point to the junction of the two pots and also short out the diodes, then gradually move the wiper of the lower pot towards ground with the gain pot full on. Distortion will increase untill the channel is fully open and then it acts as a follower, there'll be a sudden drop in volume and no more distortion. Ease the pot back from this till you hear fuzz again. Repeat the process with the other pot, moving it gradually towards 9V. When you've done this bring the diodes back into the equation and adjust both wipers towards the centre of the divider till you get the type of distortion/asymmetry/threshold that you want.

From this thread:

requesthttp://www.diystompboxes.com/sboxforum/viewtopic.php?t=244

Actually, the circuit I outlined won't work.  The original I built used JFETs and did work but required testing the FETs so I came up with a simpler MOSFET version to post on the forum.  I don't think I ever tried the MOSFET version which is why I didn't pick up on the inherent flaw (substrate brought into conduction), my memory (hazy at the best of times) probably just remembered the JFET circuit when I made that post.

The idea I posted should work though by using just a single MOSFET.  In one direction the substrate will conduct and in the other it functions as a MOSFET connected diode with variable threshold (set by trimpot).

[lion]

gez - OK, and thanks for the drawing!

WGTP - I'm always very impressed by RG's in depth responses. Although I mostly don't understand all of it, I've learned a lot from him (and others - and I really appreciate that). But this one is totally over my head.

R.G. wrote: "But what happens if we flip this thing around the other way and make the source more positive than the drain/gate? It ** should** block any current flow if it operates like a normal diode, which is what we want.

But it doesn't. There is an internal "substrate" diode that's an inevitably part of the making of the MOSFET on silicon that acts like a standard silicon diode with its cathode to the drain and its anode to the source. This does not affect normal drain-positive operation at all, but it lets current flow through unimpeded when the voltage is reversed.

If you use two diode-connected MOSFETs back to back, the substrate diodes are also back to back, and the 0.7V substrate diode drop always conducts before the higher-voltage paralleled MOSFET diode can. You **are** using a standard pair of silicon diodes, just doing it the hard way.

This gets fixed if you put another silicon diode in series with the MOSFET so the current can only flow the "correct" way, through the drain toward the source. The external silicon diode then blocks the reverse voltage, and the opposite-polarity MOSFET can do its MOSFET-diode thing in the reverse direction without being shorted out by the substrate diode of the "off" MOSFET.

So - each diode connected MOSFET is really a silicon diode connected in series with the MOSFET with its cathode connected to the MOSFET gate and drain, and the + input of this mess to the anode of the silicon diode, the output being the source of the MOSFET. Take two of those, connect them up back-to-back, and you get a MOSFET diode clipper."

Can anyone do a translation for dummies like me?

Erik

[gez]

This is a fixed version of what I posted a while ago.  Haven't tried it, but it should work...



The comments I posted in my original thread apply to the JFET version and should be ignored.

[gez]

Can anyone do a translation

Try imagining the MOSFET as having a 'hidden diode', as shown here:



When biased normally, it's reverse biased and doesn't conduct.  However, make the drain more negative than the source and the MOSFET's channel is shunted by this diode.

Read through RG's post again and hopefully it'll be a little clearer.

[WGTP]

I found it, but can't claim to understand it.  Gez thanks for the drawings.  I'll have to work on that further.    

[brad]

This is a fixed version of what I posted a while ago.  Haven't tried it, but it should work...

Hey, that looks freakin' interesting!!  

So you adjust the lower pot, then the other pot, and tweak them accordingly again when you put the diodes in?

I recall seeing a circuit by Joe Davisson (I think) called the "Dual MOS" once, and I think that was a Mosfet clipper.

[gez]

So you adjust the lower pot, then the other pot, and tweak them accordingly again when you put the diodes in?

Leave the diodes in place when you're setting this up, those comments referred to the JFET version only (I got muddled).

With both wipers pointing towards the middle of the trimpot divider (half supply) the n-channel FET won't conduct until the output falls to such a point that the gate-source threshold is met. With the p-channel it won't conduct till the output rises to the point where its threshold is met.

MOSFET's thresholds are often around 1-2V, but can be more (3V max according to data sheets of many devices, but I've yet to measure one that high so far!) so if you have weak pickups there's often a not too subtle transition from distortion to clean half way through a note's decay! By adjusting the trimpots away from the centre point of the divider you lower the thresholds. Not only does this get rid of the problem, it also means you can tune the thing for asymmetry.  There comes a point when both channel are fully conducting if you turn the trimpots too far, so the clipping diodes are the only things left in the equation - bear this in mind when tweaking this thing!(that's what my comments about removing the diodes were about, but that will only work with the JFET version)

The above is only a tube screamer type of affair, don't expect metal mayhem!

[gez]

Haven't tried this either, but if you took the diodes out completely you'd get substrate conduction in one direction (about .7V clipping threshold)and the threshold of the FET in the other, which can be adjusted by the trimpot, so you could get a really low (or high) clipping threshold in one direction.  



It would be a simpler way of dialing in some asymmetry.

With both circuits make sure you use an amp with plenty of headroom.  You'll also need a tone stack and vol pot on the output, plus some supply filtering (I just kept things simple when drawing this up).

[WGTP]

I'll have to try some of this.  REMINDER:  I have fried seveal 2N7000's using them as diodes, so be careful.  When it is warm, in the northern hemisphere, is the best time to play with MOSFET's.  Less static electricity.  

[lion]

gez - thanks for trying to help with the 'hidden diode' drawing. Actually I got that part of RG description - but not much else. I'll just have to try harder  :?

The FET-Screamer looks interesting. WGTP - if you try it, please keep us posted.

Could the n/p-channel mosfet setup be used in (implemented to) a standard TS circuit?

Erik

[gez]

Could the n/p-channel mosfet setup be used in (implemented to) a standard TS circuit?

Sure, but you'd end up with a higher clipping threshold, so if you have weak pickups you might want to have a little boost at the input (reconfigure the input transistor as a common emitter circuit with gain of two or three).  Same applies to the first circuit I posted (the second will allow a really low clipping threshold in one direction, hence more distortion, though you might want to add a little gain afterwards to compensate for the loss of 'volume')

[gez]

Sure, but you'd end up with a higher clipping threshold,

Er, which is why I incorporated trimpots, to solve this problem. :roll:

Would somebody please give me a prod with a big stick, I need waking up! (bangs head against monitor in repetitive Reichian manner)...

[Nico666]

I have been through all these threads but I can't find that answer.
What will be the threshold voltage; the same as Mosfet (normal wiring)+Ge or less?

Nicolas