is the ocd mosfet clipping using s+g d connected mosfets? or g+d s ?

[Quackzed]

i'm looking at different schematics and different layouts and there doesn't seem to be a standard answer.
is it the 'amplified diode' setup for soft clipping? this seems like an obvious reason to use mosfets to begin with, but many of the schematics and threads show them set up as source-gate connected?!? that would be 'normal' diode connected. additionally theres an extra diode in series with one of the mosfets and i've seen it drawn up anode connected AND cathode connected to either the lone source or the source/gate connected mosfet setup. i'm bewildered  so will the real ocd connected mosfets please stand up ?

[dbp512]

I just built one shorting the drain and gate together for one leg, and the source as the other leg, as per the tagboard layout. It sounds good, so I haven't tried anything else.

Edited for clarity.

[Quackzed]

from this pic it looks like s-g connected. at least ONE of em anyway...

[midwayfair]

It's G+D -> S with a diode in series if you want something "special."

Discussed in this pedal's build doc, with links to further reading:

https://docs.google.com/document/d/1U9ZENwZcjz2GvbG60vABIp15d1c4UQJ7asshD11qTyM/edit

other link for the impatient or lazy:
http://www.diystompboxes.com/smfforum/index.php?topic=90474.0

[Quackzed]

but the picture above IS an ocd, and clearly shows the s+g d connection. sorry, not disagreeing, but i keep finding definite answers for both possibilities. also, doesnt g+d s have the higher threshold 'soft' diode clipping characteristic? ok, maybee the bit thats confusing me is this: WITHOUT the opposiong polarity diode for the 'something special' action ; is the d+g s connected mosfet just the normal body diode action due to the antiparallel mosfets lowest threshold body diodes? or more simply is one mosfet preventing the 'soft clipping' of the other with its lower threshold body diode? and vice versa?

[TejfolvonDanone]

Quick clarification:
First of all in every MOSFET there is a diode called body diode. This is from the S to D.

If you connect G+S together you use only the body diode as the gate-source voltage is 0 meaning that you turn the MOSFET off. It will have a similar sound as a regular silicon diode.

If you connect the G+D together you use the MOSFET not the body diode. This means that it will have a softer characteristics compared to a regular silicon diode because you use the MOSFET's characteristics (both the Vgs vs Id and the Vds vs Id) to clip.
Also if you do this the body diode isn't left out of the circuit: putting a negative voltage to the D and G will open the body diode.

[Quackzed]

so i guess it doesn't matter, because either way with back to back mosfets the body diode of one will conduct before the 'soft clipped' higher threshold 'magical' mosfet clipping can happen with the other. each mosfet has a body diode in one direction (source to drain  ->l-  ) and a higher threshold 'mosfet turn on' soft clipping action in the other ( when gate and drain  ie. g+d s  connected). one signal polarity hits the body diode of one diode first and prevents the other mosfet from doing its groovy soft thing and the other signal polarity hits the other mosfets body diode b4 the first one can do ITS groovy thing. vise versa.

so it SEEMS like 2 antiparallel source gate connected mosfets as clippers is basically the same thing as 2 antiparallel gate drain connected mosfets as clippers...

cool. I think now i'll go downstairs and kick the neighbors dog.

[Quackzed]

i posted my comment above b4 i saw this...

If you connect the G+D together you use the MOSFET not the body diode. This means that it will have a softer characteristics compared to a regular silicon diode because you use the MOSFET's characteristics (both the Vgs vs Id and the Vds vs Id) to clip.
Also if you do this the body diode isn't left out of the circuit: putting a negative voltage to the D and G will open the body diode.

this is what i was trying to figure out, weather the ocd was using the mosfet characteristic clipping or not. i had it in my head that it wasn't but alot of schems seemed to have the d+g together setup which would use the 'mosfet character' clipping. what i THINK is happening is that even if the mosfet character set up is used, if you do it to both back to back mosfets, it doesn't work. you just get normal body diode clipping in each direction, in one direction one body diode clips before the others mosfet action can happen and vice versa the other polarity... so the body diodes preempt any mosfet clipping action.
  like if you use just one mosfet in d+g connected setup you get normal diode action one way and soft mosfet clipping the other way. with 2 back to back mosfets set up this way the body diode of one happens before the other 'mosfet action' happens in the other, and the body diode of the other happens before the 'mosfet action of the first... no 'mosfet action'...

[TejfolvonDanone]

so i guess it doesn't matter, because either way with back to back mosfets the body diode of one will conduct before the 'soft clipped' higher threshold 'magical' mosfet clipping can happen with the other.

Yes that's how it works.

[midwayfair]

Zactly. The method the OCD uses isn't the 'special' way so if doesn't matter which way it's facing. I was simply trying to provide context since someone had already provided the OCD evidence.

[Quackzed]

ok. thanks guys. at first i thought, 'wait a minuite! all these clones are set up for the mojo mosfet action. but it looks like the pics tell a different story and theres no mosfet action with the source gate connection set up. aha!!!' then i walked a little further down that road and realized its a cul de sac, and i found myself right where i started, both setups are basically doing the same thing. oh well. the thing sounds pretty good.
         it's most obvious 'nice' feature is how it responds well to playing dynamics. it does seem to clean up more as you play softer, and get gainier more as you play harder. has a big dynamic range between soft playing and hard playing. not too much volume difference which i like, it stays up in the mix, but a dramatic clean to distorted dynamic range... i like it!

[midwayfair]

Most of the OCD's clipping is actually the op amp. Removing the mosfets completely doesn't change the distortion much, some versions it doesn't really change at all.

[PRR]

> is it the 'amplified diode' setup for soft clipping?

I'm not seeing "soft clipping".

Idiot Assistant (SPICE) simulation. Because the handy MOSFET was a huge device with nFd of stray C, I used a low value series resistor; else the waveform is swamped in low-pass.

https://s25.postimg.org/xx9rdckcv/IRF150_clip.gif

SPICE may be an idiot, but its MOSFET models are highly developed because the first valuable work SPICE did was modeling MOSFET logic chips in a rapidly expanding logic-chip industry. The through-threshold action is critical to estimating logic delay and risetime so should be good.

The 2.9V up-swing is MOSFET action. The 0.8V down-swing is plain old diode action. The diode is a little softer, and considerable soft when scaled to the same peak signal voltage.

"MOSFET clipping" may be more marketing than musical?

[Quackzed]

i can see it happening in falstad... not sure, though i've read that a lower gain mosfet will be softer than high gain... (i'm not getting body diode action in falstad's tho.) ...
http://www.falstad.com/circuit/circuitjs.html   ...FILE... then ...import from text... -copy/paste below- 

$ 1 0.000005 5.023272298708818 57 5 43
R 480 304 352 304 0 1 40 5 0 0 0.5
g 624 416 624 432 0
r 560 304 480 304 0 1000
w 560 304 560 256 0
w 560 256 624 256 0
w 624 256 624 288 0
O 720 288 832 288 0
w 624 416 624 320 0
f 560 304 624 304 0 1.5 0.02
d 672 384 672 304 1 0.805904783
g 672 384 672 400 0
w 720 288 672 288 0
w 672 288 672 304 0
w 672 288 624 288 0
o 6 32 0 550 2.3974087538301907 0 0 -1

[Ripdivot]

so i guess it doesn't matter, because either way with back to back mosfets the body diode of one will conduct before the 'soft clipped' higher threshold 'magical' mosfet clipping can happen with the other.

Yes that's how it works.

So could you not just put one soft clipping mosfet one direction followed by a non inverting buffer and then place the other soft clipping mosfet in the other direction after the buffer?

[Quackzed]

yeah, you could do that. I didn't mean to imply that theres no way to get back to back softish mosfet clipping, just that if you use 2 mosfets together (without an extra diode), the body diodes in the mosfets will prevent each others 'active' mosfet softish clipping from working.
you can also place a diode in series with each mosfet oriented opposite each respective mosfets internal body diode, this will prevent the body diode from conducting and will allow each mosfet to 'mosfet soft clip' in one direction and not conduct in the body diode direction (due to the added opposing direction diode). so with the extra diode added to each back to back mosfet , each mosfet will only clip in its 'mosfet action' direction, and wont be able to conduct in its body diode direction and therefore wont be able to prevent the other mosfet from doing the same. the extra diode will add threshold to the already highish mosfet action threshold in each direction, so a low threshold ge diode may be a good idea, as well as a higher pedal voltage to avoid rail clipping...
   

[GGBB]

yeah, you could do that. I didn't mean to imply that theres no way to get back to back softish mosfet clipping, just that if you use 2 mosfets together (without an extra diode), the body diodes in the mosfets will prevent each others 'active' mosfet softish clipping from working.
you can also place a diode in series with each mosfet oriented opposite each respective mosfets internal body diode, this will prevent the body diode from conducting and will allow each mosfet to 'mosfet soft clip' in one direction and not conduct in the body diode direction (due to the added opposing direction diode). so with the extra diode added to each back to back mosfet , each mosfet will only clip in its 'mosfet action' direction, and wont be able to conduct in its body diode direction and therefore wont be able to prevent the other mosfet from doing the same. the extra diode will add threshold to the already highish mosfet action threshold in each direction, so a low threshold ge diode may be a good idea, as well as a higher pedal voltage to avoid rail clipping...

You can also place the mosfets in series with opposing 'polarity', so that in each direction the body diode of one mosfet enables the soft clipping in the other mosfet.

[Quackzed]

  so the body diode of one IS the opposing direction diode for other! NICE! thats pretty clever, i would not have thought of that... it isn't a visually obvious way to set up opposing clippers,in series like that but yeah, i get it-that works. 

[GGBB]

Another 'trick' for asymmetrical clipping is a single mosfet - body diode clips one direction, mosfet clips other direction.