Clipping Diode Shootout

[dmp]

A few I had saved

 

[user 132374]

Which germanium diodes did you try?

 

[skipwave]

The Cornish buffer is a bootstrapped EF.
I like adjusting the bootstrap cap value smaller sometimes with bootstrapped input circuits.
You can find bootstapped circuits in books

Another adjustment for the TS I have done is add a .1uf cap and a switch for .1uf or 1.1uf after the input EF to the gain stage to move the HPF up.

I did this on my builds of Gus’s 3-Transistor Fuzz design. It is such a handy option to have and makes it easy to adapt to pick ups with more or less low end.

I found that the higher corner frequency setting sounded better on bass with a clean blend adding back in the lows.

Thanks for an excellent pedal design and all the thoughtful mod suggestions @Gus !

 

[Gus]

another good link

https://www.thetonegeek.com/single-post/ayan-enterprise-smooth-slim-style-eq-device

 

[CJ]

Thanks Gus! That would be a quick build. Maybe put a position switch on the Cornish buffer bootstrap cap instead of the passive circuit?

 

[fazer]

A few I had saved

Thanks DMP. The Rat. Looks like a cocked wha wha at one frequency. I’ve never cared for the sound of a rat. The rc combos on the minus input dictates that. On the other hand if I were playing in a band all the time rather than recording alone, I would probably see the need to cut through the band I suppose. I know the 1200 hz point is similar to a marshall tone response to some degree and might work ok with a black face fender

 

[CJ]

Received somemore diodes fromsmall bear electronics,

1N69
1N965
Amperex OF-129
Germaninium Diode Hi-V Fwd (small bear name)
Russian D9J
Mullard OA10
Raytheon 1N128

what amazing results did i find? nothing. i ain't no diode guru, just doing the plots.

we also threw in the Bat41 and the BAS33 which is supposed to be a good sub in the analogman king of tone build.

results found were that the Amperex OF 129 has a very low fwd v,
the 1N965 is a close match to the ITT 1N100 and ITT 1N192
the 1N128 is a close match to the russian D9E
the russian D9E has a lower fwd v than the russian D9J
the 1N69 is a close match to the Sylvania 1N119
the small bear Ge Hi-V fwd is a close match to the Japan 1S1588 and 1N914

the Bat 41 which i never liked seems to take a while to get off the ground,
the BAS33 for the king of tone has the highest fwd v and no wonder it gives a pretty aggressive clip in that circuit,

the new plots are the colored pen graphs>

 

[CJ]

here is kind of a summary for the whole mess, forward voltages at 3 ma dc current ,

most clipping circuits probably work at much lower currents and diferent ac signal voltage levels making this experiment inconclusive at best, the final determining factor being the good ol ear test.

and we got a new song, key of A with a rock beat:

E I got stomp boxes coming right out of my azz
A And they all come equipped with buffer bypass
E I'm going broke buying parts from digi key
A I'm gonaa end up livin in a tree

sound like a hit?

 

[dmp]

clipping circuits probably work at much lower currents and diferent ac signal voltage levels making this experiment inconclusive at best, the final determining factor being the good ol ear test.

Maybe the slope at 3mA of current?

The Vf doesn't seem like it would be that important since the level coming into the clipping stage is just based on the gain driving it.
But the slope of the clipping would dictate how rounded the clipping of the waveform is?
I have no idea for sure, just thinking

 

[mr coffee]

Really been enjoying this thread. Thanks to all the great contributors!

All this got me thinking about the diode I-V curves and their impact in an overdrive pedal, and I'm wondering if we are not measuring at low enough current levels to get an objective view of what the effect of various diodes may be in the KOT circuit.

Looking at the diodes in the feedback path of the op amp, I figure the current through the diodes should be under 0.5 ma. if I am understanding this correctly.

The output voltage from the first stage passes through the 10K input resistor to the inverting input of the second stage op amp, The 220K feedback resistor has back-to-back diodes in parallel with it, for a maximum gain of 22x when the diodes aren't conducting and the first stage gain is maxed out. See the attached schematic if you aren't familiar with the circuit.



Wouldn't the input current be (round numbers) 5 volts/10,000 ohms or 0.5 ma. ? And wouldn't that mean the feedback current through the diodes in the feedback loop would also be no more than 0.5 ma. ? I'm guessing around 0.1 ma. for an undistorted guitar signal?

And with the 6.8K resistor in series with the the clipping diodes, it's hard to tell how great the effect the dynamic impedance of the diodes has. That's above my pay grade in understanding the circuit.

AND all this also got me thinking - what if we scale the current in the second stage up to the higher levels CJ and dmp measured ? If we replace the 10K input resistor with 1.5K, make the feedback resistor 30K, and make the resistor in series with the diodes 1K?

We would get higher feedback currents in the diodes (5volts/1500 ohms or 3.3 ma) without changing the gain of the second stage. Might make it sound smoother or at least slightly different. And the diodes chosen might make more of a difference in the sound you get. FWIW

 

[CJ]

I need a better meter to track those diodes at the 100 microamp and below regions.

And using a king of tone peddle for a test bed might be a good idea. Get some scope shots with different diodes installed, it has those switches for hard and soft clipping. At least on the Aion FX version.

Diode types do make a difference I was about to ditch the Aion FX g 2 Cornish big muff pedal because it did not sound that great with bat diodes. Stuck some 1n128 diodes in there and it smoothed out the tone quite a bit.

Tests could also include trying two different diodes together. Asymmetric clipping. This would require long hours of testing as the permutations would become endless.

 

[dmp]

Yeah, I expect the situation for the soft clipping (D1-D4) is very different than the hard clipping (D7-D8)
I've been building and experimenting with hard clipping lately
And agreed that different diodes make a big difference in tone

 

[CJ]

i guess i could use a 100K resistor as a current sensing resistor, 40 volt supply would give 400 uA max, and read 500 mv at 5 uA, what kind of comp program/I/O device are you using for your graphs?

 

[Tubetec]

Its a purely subjective quality were after here,
the numbers have little to do with it , having a base line measure of when the various diodes start to conduct is useful.

 

[CJ]

Oh a false clock tries to tick out my time
To disgrace, distract, and bother me
And the dirt of gossip blows into my face
And the dust of rumors covers me
But if the arrow is straight
And the point is slick
It can pierce through dust no matter how thick
So I’ll make my stand
And remain as I am
And bid farewell and not give a damn

 

[mr coffee]

CJ,
Sorry if I misread your earlier post

Stuck some Ge diodes in the king of tone pedal last night and it kicks the donkey's butt. The hard clipping circuit was too harsh with the stock diodes.

I thought the schematic I posted was your target circuit. I guess there are a lot of stomp boxes called "king of tone".

@CJ, dmp and Tubetec,

I think CJ's exploration of the I-V curves at signal level currents are a scientific approach to understanding what the significant parameters of "good sounding" clipping diodes are in various circuits, and disagree with dismissing the scientific approach in favor of golden ears subjectivism.

The higher currents CJ and dmp already posted are highly relevant in the "hard clipping", aka 1K-feeding-a pair-of-diodes-to-ground circuits.

@CJ,
I think the 40 volt, 100k resistor approach makes good scientific sense for diodes in the feedback loop. FWIW.

OT-That must be a famous poem, but I can't place it. Who's the author?

 

[CJ]

klon centaur diy, raytheon 1N128 which resembles the rusky d9e,

double gang vol pot, switch is for buffer bypasss or true bypass


sounds killer! but not $8,,000 kiiller, if i pay that much for a stomp box i want it to play the guitar and distort the signal, and drive the car home after the show.

 

[CJ]

got a box fabbed for the centaur,

i know what your sayin, "that don't look like no centaur to me"

i know, but i don't want to get in trouble with klon by doing exact artwork,

 

[dmp]

What, that looks awesome!
I'll have to check what diodes I have in my klon clone. Some kind of 1n34 I think

 

[musipol]

Bob Dylan Lyrics????