Wich MOSFET you like the most for clipping?

[mateusborges]

Heyas!

Guys, I'm trying to find "the best" MOSFET to use as clippers. I know, I know, such thing doesn't exist, however, I've tried 3 combos with the ones I had available here (IRF520, 2N7000 and 2SK962) but couldn't find much of a difference between those, is that correct? There's any other MOSFET out there I should try?

I'm using this kind of MOSFET connection:



Cheers!

[Steben]

You mean the clipping is comparable and/or is the treshold leadi g to ampltude/level the same?

[mateusborges]

You mean the clipping is comparable and/or is the treshold leadi g to ampltude/level the same?

To my ears there's not much difference in any way between those that could make me go over one than the other, and for such small difference (at least in my testing) I much prefer using 2N7000 due to availabilty and size.

P.S. about testing: ProCo Rat circuit mostly, some TS808 circuit.

Cheers.

[mateusborges]

BTW,

setting the MOSFETs like this...


Will get me the same result as the way I'm doing? This one...



2N7000 N channel MOSFET G+D soldered, 1N34 anode soldered to source, then G+D leg and 1N34 cathode leg used as the clipper.

Cheers!

[Vivek]

Maybe the title should have been:

Which MOSFET you like the most for clipping? and do you really claim you can heard the differences between MOSFETS and regular diodes in a blind test, when normalized for equal clipping and equal volume ?



They say that the DNA of a human and a shark differ by 4%

Human and monkey by 0.5%

Maybe all methods of clipping differ less than 0.05% when normalized for equal clipping and equal volume.

[idy]

What do you think Matteus? Vivek is poking...In your opinion do MOSFETs sound different than just a diode?

I played with that a lot. The two sound different to me (at least when not just used for body diodes I actually never tried that!), and look different on O-scope. Subjectively, "more complex," objetively, the little spikes on the wave were more complex

[bean]

I've done IRF510 before and liked them a lot. I think I tried it after seeing the Beavis Audio Rat project, although I used them for soft clipping in a different circuit, IIRC. I remember them sounding a little different than 2n7000 but it's been several years.

If you look at some of the VFE schematics on my website you'll see Peter used BS170 often for clipping. Here's a couple of examples:
(PS These are "factory" schematics but I have the okay to post them)

[mateusborges]

-double posted-

[mateusborges]

What do you think Matteus? Vivek is poking...In your opinion do MOSFETs sound different than just a diode?

I played with that a lot. The two sound different to me (at least when not just used for body diodes I actually never tried that!), and look different on O-scope. Subjectively, "more complex," objetively, the little spikes on the wave were more complex

Heyas!

Yes they do! However I discovered that it's kind of a "type" thing. Mosfets clip like mosfets, germanium like germanium, sil like sil, etc.

Again, I'm faaaar away from being expert and I don't even have proper testing tools, but for my practical uses, so far, I couldn't find enough difference between, let's say, sil A to sil B, or germ A to germ B, given they have close stats, and I tested quite a bunch of them, to the point of inverting phase in recorded samples, reamped from the same source, and check for differences, and at least for sil vs. sil, and germ vs. germ, they are tiny, too tiny for my needs and persuit, however, between these "types" yeah, they differ enough for me wanting to have them at hand.

Couldn't test many Mosfets so far, only have 3 at hand. But yeah, set like the D+G - S -> A|K of a low germanium diode it clips way different than germ or sil alone.

Cheers!

[mateusborges]

I've done IRF510 before and liked them a lot. I think I tried it after seeing the Beavis Audio Rat project, although I used them for soft clipping in a different circuit, IIRC. I remember them sounding a little different than 2n7000 but it's been several years.

If you look at some of the VFE schematics on my website you'll see Peter used BS170 often for clipping. Here's a couple of examples:
(PS These are "factory" schematics but I have the okay to post them)

Wow, thanks a lot Bean!!

Cheers!

[GGBB]

BTW,

setting the MOSFETs like this...


Will get me the same result as the way I'm doing? This one...

I think those are quite different. MOSFET D>S in series with body diode vs Ge diode in series with body diode. Very different voltage drop at least - probably sound different as well.

[idy]

I think GGBB caught something. My notes always show the diode "pointing away from" the source. The way you show is just diodes.

[Vivek]

What do you think Matteus? Vivek is poking...In your opinion do MOSFETs sound different than just a diode?

I played with that a lot. The two sound different to me (at least when not just used for body diodes I actually never tried that!), and look different on O-scope. Subjectively, "more complex," objetively, the little spikes on the wave were more complex

It would be nice to see your Oscilloscope screenshots


Please remember, one of the suggestions I had for comparing the various clipping methods was to try and normalise clipping levels and volumes

I mean

If we put a 6V sine wave into a 0.6v clipper and then add a gain stage to get some volume

Then we compare that with a 3V sine wave into a 0.3v clipper and add a gain stage to get same volume as above

Now we have somewhat normalised clipping level and output volume for a slightly more fair test

[Steben]

What do you think Matteus? Vivek is poking...In your opinion do MOSFETs sound different than just a diode?

I played with that a lot. The two sound different to me (at least when not just used for body diodes I actually never tried that!), and look different on O-scope. Subjectively, "more complex," objetively, the little spikes on the wave were more complex

It would be nice to see your Oscilloscope screenshots


Please remember, one of the suggestions I had for comparing the various clipping methods was to try and normalise clipping levels and volumes

I mean

If we put a 6V sine wave into a 0.6v clipper and then add a gain stage to get some volume

Then we compare that with a 3V sine wave into a 0.3v clipper and add a gain stage to get same volume as above

Now we have somewhat normalised clipping level and output volume for a slightly more fair test

Yes.... but ..... having 2 or 3 geranium diodes will give you average textbook 0.6v as well but more volume because of higher internal resistance. It behaves differently.

[Vivek]

Audibly different ?

Or only different in Sim and Scope ?

and drastically different than Si with series resistor ?

[Steben]

and drastically different than Si with series resistor ?

No. That's exactly what it is about.
Question is: is there a difference between a SI diode and a SI diode with added resistor. Answer: yes. the resistor. Softer clipping, but not by knee or curve softness. It never "hits rails".
Now if the series impedance is 1M and the resistor is 10 ohms you wont notice much. It's all relative.

[anotherjim]

Anybody tried connecting N-MOS gate to some fixed positive voltage (above 4.5v) so when the source swings below 4.5v toward 0v you have a preset clipping point? That gate voltage should probably be adjustable as the turn-on Vgs might be anywhere up to around 3v. This could be close to the opamp rail clipping point (and you have the blocking diode drop there too) so you'd either want a R-R output chip or select the MOSFET for low Vgs threshold.