YADT - Clipping Diodes and arrangements

[KR Sound]

Yes, another diode thread. Here's a generic set up for some clipping options you find in a lot of circuits. You have your LED, Si, Ge, Sch options, right.

When you serialize the diodes like in 'C', the Vf is a sum of the two, so does it make sense to just use a diode that is equal to that sum?

Two BAT46s are showing 0.68 and so is a 914 or other similar silicon diode. Is there a discernable tonal difference between the two, or is it done to just have another option while using the same inventory?

I'd try them out right now, but my daughter is home sick from school and I'm listening in on a department meeting

 

[benny_profane]

Diodes are nonlinear semiconductors. You’re just seeing a snapshot of the forward voltage at a given voltage from your test equipment. You’d want to compare the Fv graphs to see if the curves are the same. (Note: the slope of the curve is what people are talking about when they refer to the ‘knee.’) So, although at one point the arrangements may be equal, that doesn’t tell you about their nonlinear performance similarity or behavior at different voltages.

 

[KR Sound]

Diodes are nonlinear semiconductors. You’re just seeing a snapshot of the forward voltage at given the current from your test equipment. You’d want to compare the Fv graphs to see if the curves are the same. (Note: the slope of the curve is what people are talking about when they refer to the ‘knee.’) So, although at one point, the arrangements may be equal, that doesn’t tell you about their nonlinear performance similarity or behavior at different currents.

Gotcha. That makes sense then. I’ll do some comparisons later this evening and see how they differ with the same Vf.

 

[benny_profane]

Gotcha. That makes sense then. I’ll do some comparisons later this evening and see how they differ with the same Vf.

Note: I edited the initial post to correct a current/voltage mix-up.

This is a good reference:
https://forum.pedalpcb.com/threads/measured-odd-and-common-diodes-iv-curves.19777/

 

[Cybercow]

Here's a comparative graph of some common diodes, created with a Peak Atlas DCA75 and MS Windows. You can see the differences & similarities of the Vf curves between them.
IMO, they don't deliver a "tonal' difference as much as 'distortion attack transient' differences where the curves differ more drastically.

 

[KR Sound]

Here's a comparative graph of some common diodes, created with a Peak Atlas DCA75 and MS Windows. You can see the differences & similarities of the Vf curves between them.
IMO, they don't deliver a "tonal' difference as much as 'distortion attack transient' differences where the curves differ more drastically.

Cool- here are my results with the 2xBAT46 vs a 914 and 4148




They cross at 5mA with the exact same Vf, but like you said the knee is different. How that affects the sound, I'll have to test that out!

 

[Aentons]

Cool- here are my results with the 2xBAT46 vs a 914 and 4148

View attachment 65005


They cross at 5mA with the exact same Vf, but like you said the knee is different. How that affects the sound, I'll have to test that out!

View attachment 65006

Cool I don't think Ive seen any knee comparisons of a series pair vs a single diode. It's looks like the pair caused the knee to soften.

 

[spi]

These curves would be more useful if I knew what mA the signal in a pedal is dealing with.
I've seen mention of current draw for a pedal, but that doesn't tell me what the current of audio where clipping is happening is.

Does anyone know what kind of current is going through the clipping diodes of a RAT or Dist+ for example?

 

[benny_profane]

These curves would be more useful if I knew what mA the signal in a pedal is dealing with.
I've seen mention of current draw for a pedal, but that doesn't tell me what the current of audio where clipping is happening is.

Does anyone know what kind of current is going through the clipping diodes of a RAT or Dist+ for example?

The current is a function of the voltage of the signal.

 

[HamishR]

At the risk of tasking this high-brow thread in a somewhat less scientific direction, I recently discovered that I like the sound of a pair of BAT41s used as clipping diodes in the typical TS-style feedback style. I really don't like the amount of compression you get with a pair of 1N4148s, 1N914s, etc. I have always used a quad of those diodes to get the sound/feel that I want because I like a livelier more dynamic OD.

Then I discovered that a single pair of BAT41s will deliver a similar kind of response as a quad of 1N4148s. It is slightly different but it's similarly dynamic and less compressed and it works for me. I like it so much I haven't even bothered to try a quad or trio of BAT41s yet.

 

[KR Sound]

@HamishR Here's some more quick tests of what I have here in front of me. Your Bat41s knee is totally different than the rest.





I wish I could order the Keys or customize the layout more, but here's a closeup of the names of all those on the right.

 

[HamishR]

That would be more interesting to me if I could understand it! It kinda matches with how I hear things I guess. I really like the way a pair of BAT41s react to my admittedly ham-fisted playing. At least in the Timmy-based pedals I use them in. It makes me wonder why more designers aren't using BAT41s. They sure feel more like the way my amps break up then a pair of 1N4148s do.

 

[KR Sound]

I’m going to try some in this Guv’nor-turned-Crunchbox thing with a few clipping options I’m working on. I’ve settled on just the red LEDs to start with, then BAT46s for the high distortion and compression and the 41s should fit nicely in the middle of them.

 

[jwin615]

Re current going through the diodes. I don't know the amount but can explain the function, at least in terms of hard clipping. Note this is an over simplification and ignores several other parts of the circuit and reactance in general.
I'm not sure anyone can tell you a set amount for I as it will vary depending on the previous level of amplification.
Circle back to Ohm's law
V=I x R or in this case, I= V / R
Your resistance is going to remain fairly steady across most of the spectrum, overlooking filters and caps here.
So if resistance is constant, the current should vary with voltage. Higher signal= different part of the "knee"
Again, all this will be different with different previous levels of amplification.
There are many ways to tune this as well. There's more than a couple amz articles on it.
Note, this is ignoring the inherent resistance of the diode itself, again, in oversimplification, and the fact that the resistance is integral to the "knee" as V increases, internal diode R decreases allowing I to increase.

 

[CR0SSBL4DE]

Cool- here are my results with the 2xBAT46 vs a 914 and 4148

View attachment 65005


They cross at 5mA with the exact same Vf, but like you said the knee is different. How that affects the sound, I'll have to test that out!

View attachment 65006

Similar threshold but different shape/slope. You will have a similar result using a red LED vs 3 silicon diodes.
Or using 2 or 3 germanium diodes vs a silicon diode. Or even a germanium diode vs a schottky diode.