OCD - 3 different revisions with 3 different Mosfet Clipping Setups...

[Bastelliese]

I'm sorry, but I'm still not really satisfied. Of course the Ge adds asymmetry to the clipped signal. That's undoubted. Add another Ge to the upper Mosfet and it's symmetrical again. What bothers me, is the way the Ge is connected. You are assuming that V3+V4 in the above picture will produce the same clipping, but KHStudio claims hearing a difference. Sure, there could be some differences in parts values or anywhere else in the circuit not mentioned yet. But I assume it's the manner of connecting the diode.
We also agree that there is no Mosfet-clipping in the OCD circuits, all being body diode clipping.
The body diode of a N-channel Mosfet is located between drain and source, drain being cathode. What about the gate? Why is it normally connected to source, couldn't it be left open and what was the differnce?

[GGBB]

I may very well be wrong, but you can't just claim I am wrong without backing up that claim with evidence. The scientific method is what we use to test a hypothesis, which in this case is "hearing a difference." Use the scientific method to validate the claim: There are a limited number of ways that current can flow through the two clipping arrangements according to the terminals connected to signal. Make a list of them all and then apply physics and math along with the known properties of the diodes and MOSFETs to test whether or not they actually will pass current from an amplified guitar signal with no DC bias present. If you have access to a scope, use that too. I applied known math to known properties to make my conclusion. I invite anyone to prove me wrong or show me that there are other properties I didn't consider - I already stated that a few days ago.

EDIT: You actually can claim anything you want . But your claim has no merit unless you back it up with evidence.

[Bastelliese]

Excuse me, I did not want to be overbearing nor question your knowledge. I didn't mean to proof you're wrong or even offend you I'm sorry for that. I just wanted to keep the topic open to get futher beyond and my impression was that you aren't that sure either. Please keep in mind that I´m writing in a foreign language which I'm not used to. So there might be some 'between the lines' content or connotation that's not intended. I'm sorry for that!

[duck_arse]

please, somebody, draw the circuit out. let's see what it is we are working with, if only the clipping section. as far as I understand this arrangement, having read an explanation of the correct wiring for mosfet clipping, the external series diode opposes the internal mosfet diode (the body diode you are all referring to?), so it can't conduct. I can't remember the correct gate connections, tho, I was hoping someone else would point to the discussion in question.

[Bastelliese]

Well, I might not be the perfect guy to explain the circuit and mosfet clipping, but I'll try my best to keep the discussion going.
First of all here's a schematic, which seems to have most common concensus:


For me the most helpful resources were the thread 'Understanding MOSFET clipping' at freestompboxes.org  and the website of amzfx http://www.muzique.com/lab/zenmos.htm which is a real good starting point to get into this stuff.

But I'll try to explain in my own words.
In circuits like OCD and similar, e.g. Joyo Ultimate Drive, where two Mosfets in anti-parallel wiring form the clipping stage, it doesn't matter whether the drains+gates or the sources+gates are connected and therefore the mounting direction isn't of interest. This is because of the forward voltage of the body diodes being far lower than the gate/source threshold of the mosfets. Therefore the body diode will conduct and clip the signal. The body or  inverse diode is intrinsically formed between source and the substrate and between drain/substrate.  Actually there are two body diodes per Mosfet, but usually the source/substrate diode is internally shorted out. When we talk about the body diode we refer to the drain/substrate diode and since substrate and source is shorted out it's referring to the diode between source and drain. With drain being the cathode of the diode assuming we have n-Channel Mosfets. The body diode behaves like a simple Silicium diode with a forward voltage of about 0,7V.
For OCD and similar circuits this means, there is no magical Mosfet clipping at all.
After all I think most of the unique sound of UD/OCD comes from the overdriven TL082.
Nevertheless I have to admit that I recently saw a video on youtube, where a guy demonstrated his modded Boss DS1. He replaced the original clipping diodes D4/5 with Mosfets, IF520 I suppose. The only other alteration he did was to change the tone capacitor to 22nF, I think. And this mod altered the sound drastically. Assuming the above written to be true, this would say the sound difference only comes from altering the tone cap value. This seems also unlikely.

As far as I know there are three ways of getting Mosfet clipping (drain+gate connected):
-Using just 1 Mosfet. The body diode and the "Mosfet-diode" are anti-parallel causing an asymmetrical clipping.
-Using 2 Mosfets in series, opposite direction. One half-wave will pass the Mosfet and the body diode of the second one, the first body-diode and second Mosfet will block current. For the second half-wave vice versa. This is symmetrical clipping.
-Using 2 Mosfets in anti-parallel plus 1 diode in series to each Mosfet. The additional diodes must be oriented against the body diodes to block them and therefore with cathode to drain or anode to source. This will give symmetrical clipping with slightly higher g/s-threshold. Of course it can be made asymmetrical by adding another diode for one half-wave.

Now let's return to the original post by KHStudio and his drawing.
As mentioned earlier, all 3 versions show normal diode clipping. I own two Joyo Ultimate Drives and in the schematic gates and drains are tied together, whereas on the real PCB the gates and sources are connected. But as we learned earlier, this doesn't make any difference. I suppose this being the same for the OCDs.
The schematic on the right shows a symmetrical diode clipping arrangement, long known from circuits like Tubescreamer.
The middle one (V4) has a Germanium-diode in series to one of the body diodes, thus creating an asymmetrical clipping stage, known for example from Boss SD1.
Here the gates and sources are providently tied together, to prevent the Mosfets from unwanted gate/source current flow and stable conditions.
As GGBB considers, and I basically agree with him, the left V3 seems to be very similar to V4 and therefore should produce asymmetrical clipping.
But the question is still the same. Why is the diode connected in this uncommon way?

Everybody who's interested in Joyo Ultimate Drive, should read the thread 'Joyo - JF-02 Ultimate Drive "2010" MISTAKE' on freestompboxes.org.

Good night
Sascha

[GGBB]

Excuse me, I did not want to be overbearing nor question your knowledge. I didn't mean to proof you're wrong or even offend you I'm sorry for that. I just wanted to keep the topic open to get futher beyond and my impression was that you aren't that sure either. Please keep in mind that I´m writing in a foreign language which I'm not used to. So there might be some 'between the lines' content or connotation that's not intended. I'm sorry for that!

Not offended - sorry if I cam across that way. The point I'm trying to make is - let's not imagine something magical is occurring, let's use principles of math, physics, and electronics to determine what is happening. Electrons obey the laws of physics - always.

[GGBB]

Why is the diode connected in this uncommon way?

That's a great question.

What we don't know:

• the source of the diagram
• the source of the OP's pedal that implements "V3" clipping
• whether the "V3" pedal is an original OCD or a clone
• if a clone, whether it was a production pedal or pcb or a hobbyist pcb/vero/perf

What we do know:

• the original OCD V3 has been reasonably well documented and does not match the OP's "V3" diagram
• apparently not one single other report of an OCD (original or clone) having this style of diode connection
• this could not be an assembly error on a production pcb - you can't connect the diode like that by misplacing or misorienting components as happened in some of the Joyos - the actual traces of the pcb would have to change

Based on that, I'd argue there is no evidence to believe OCDs or production clones were ever made like this. We have one person's claim and absolutely nothing more to support it, versus solid evidence that V3 isn't connected this way. I will however accept that it could be someone's own creation.

I am not saying the OP is lying - I am sure he/she believes they are correct.

I am not saying the OP is stupid - I don't know him/her but I assume they are human, and, like every human I know, makes mistakes on occasion.

I am not saying that the pedal doesn't sound different/better to the OP - I'm suggesting that it's probably not the diode that is the reason. Lets suppose for a moment that the pedal actually does have the diode connected that way. Then, if my analysis of the physics is correct, it would clip the same as "V4" anyway, so there must be another explanation for the difference that the OP hears - it's not the diode.

So in answer to the quoted question - what I am suggesting is - the diode isn't normally ever connected that way. If it does exist, it does so either by a layout mistake in someone's DIY project, or by intentionally trying to be clever or misleading, or by a misunderstanding of how it is normally connected.

My vote goes to mistake - either on the part of the OP, or on the part of the pedal builder. If the OP or anyone else can clear up the unknowns I listed, I might be swayed to believe otherwise.