Blend between clipping diodes

[pokus]

Hey there. I'm about to modify a distortion circuit and I want to add something like a blend pot to mix the hard clipping sound of two different diodes together. It should be symmetrical clipping so there are two of each kind in parallel and reverse to each other. I've tried to just put a simple pot from the coupling capacitor after the opamp-stage to each pair of the diodes. Capacitor and output of this stage to lug2, pair 1 to lug 1 and pair 2 to lug 3. So when the pot was max. or min. it worked just fine, but in between the diodes didn't clip and gave me a louder but less distorted sound. The smaller the value of the pot the better the result. So how do I manage to blend between clipping diodes? Is it possible to build two different opamp-stages, each with it's own pair of diodes and then put a pot between them(like op1 to lug1, op2 lug3, out lug2)? Would that solve the problem? And can I split the Input then when putting a buffer in front of the splitting point? Thanks for help!

[antonis]

IMHO, blending 2 pairs of clipping diodes shouldn't result in any audibly interesting result..

You better try to wire a SPDT switch between the two pairs or a SPST switch for the pair of lowest forward voltage drop (leaving the other pair permanently wired..)

[pokus]

Ok, but the pot shouldn't stand alone. It should be a dual pot, one blends the diodes and one controls a notch filter. So you have a more compressed sound (with BAT or sth else) when the mids are scooped and the clipping sound of an LED when mids are boosted. I guess that can be quite interesting, but I'm not sure because I couldn't figure out the blend thing 

[GGBB]

As you've discovered, added resistance causes less clipping. This is because voltage drop occurs across the resistor in addition to the diode, therefore the effective clipping threshold is higher. Also, since we know that two parallel clipping diodes (same direction) with no added resistance will conduct/clip across the diode with the lowest forward voltage only, we can assume the same for diodes with identical added resistance.

I suspect the only way for both diodes to conduct simultaneously and equally is to have the voltage drops across diode plus resistor be identical for both paths. The diode with the lower Vf will require more compensation resistance. It is not possible for one path to drop some voltage and the other path to drop more voltage - once the lower threshold is reached, voltage cannot go higher in order to turn the other diode on. In theory at least. I don't know what role diode turn on speeds or junction capacitances might have, although I'd guess they're not something you would audibly notice, even if you could manage to set just the right amount of resistance to bump up the drop of a lower Vf diode to match identically the higher Vf of the other. So it is not possible to blend variably, but it is theoretically possible to have both on at the same time. How to do that practically and know that you have both conducting I don't know. Even then, would it sound different than just the one diode?

On the other hand, blending two separately clipped signals via the op-amp setup you described should work fine.

[marcelomd]

There are a few ways to blend diode clippers at AMZ blog:

http://www.muzique.com/lab/warp.htm
http://www.muzique.com/lab/sat2.htm
http://www.muzique.com/lab/tclip.htm

[GGBB]

There are a few ways to blend diode clippers at AMZ blog:

http://www.muzique.com/lab/warp.htm
http://www.muzique.com/lab/sat2.htm
http://www.muzique.com/lab/tclip.htm

Interesting. I stand corrected. Thanks for pointing that out. Note that the last two links aren't exactly blending the same signal clipped by two different diodes, they are about blending different frequency portions of the same signal clipped separately and blending different ways of clipping the signal (not the same kind of clipping using different diodes). And the first link is mostly about blending asymmetry not clipping networks but that is also mentioned.

I tested the warp control idea on the breadboard (which is what I should have done before my previous post). Sure enough, I could get two different LEDs in parallel to pass DC current at the same time by varying the resistance in front of the LED with the lower Vf. Basically this but both diodes facing the same direction:



Interestingly, the higher Vf LED passed current even at a lower voltage drop than it's Vf. The two LEDs had Vf numbers of 1.85 and 1.7. I used a 1k trimmer in front of the 1.7 LED, and fed both from a 9V battery through a 1k CLR. As I increased resistance in front of the 1.7 diode, voltage drop rose from 1.7 up to 1.85 - the range covered by the LEDs.

What I think is happening is that the two paths to ground form a current divider, and since current goes up with less resistance, as the trimmer was turned down, less current flowed through the other LED. And when current across a diode drops, so does the voltage drop. This is the turn on thing I mentioned before.

I'm not sure if I just got lucky with the 1k trimmer and LED choices, but that seemed to be a pretty good blend - both LEDs got brighter and dimmer in opposite directions across the whole range. Overall the lower Vf LED seemed to dominate slightly so maybe a bigger trimmer might have worked a little better. I would expect that trimmer size and Vf variance need to be chosen carefully through experimentation. I have a feeling that the greater the difference between the diode Vfs, the less well this would work, but I'm not sure.

So if you want to blend two different pairs of LEDs, putting a pot in series with the pair that has lower Vf diodes should work.

[GGBB]

I was fiddling with the breadboard a bit more - trying different LED combinations. I found that as the difference between Vf gets larger (I tried a 1.7V and 3.4V combination), the greater the resistance you need to in front of the lower Vf in order to turn on the higher Vf diode. And the lower Vf diode never seems to turn off completely, although I did not try a very large resistance like 1M. If the resistance gets too big, you don't have much blend range for the control - the higher Vf diode goes from off to full on over a short rotation. So this is more of a "blend in" a 2nd diode control rather than a "blend between" 2 diodes control.

[pinkjimiphoton]

use the search function for my transmission overdrive. works great. blends between two si in4XXX's and some ge/led's if memory serves. center pin of pot to output, pin one to one clipper, pin 3 to the other. i will have to check my other computer for the layout unless i can find it on google

http://tagboardeffects.blogspot.de/2013/06/pink-jimi-photons-transmission-od.html

[antonis]

So this is more of a "blend in" a 2nd diode control rather than a "blend between" 2 diodes control.

Well said...!! 

@pokus: The above arrangement serves more than asymmetrical clipping adjustment rather than blending..
(left diode clips during positive cycle where right diode + trimmer clip during negative cycle - there isn't any kind of "blend"..)

[GGBB]

So this is more of a "blend in" a 2nd diode control rather than a "blend between" 2 diodes control.

Well said...!! 

@pokus: The above arrangement serves more than asymmetrical clipping adjustment rather than blending..
(left diode clips during positive cycle where right diode + trimmer clip during negative cycle - there isn't any kind of "blend"..)

Not quite - I guess you didn't read my statement above the pic (emphasis added):

Basically this but both diodes facing the same direction:

So we're not talking about that diagram specifically. Otherwise you would be correct. 

[Phoenix]

Could always have two parallel clipping paths, and then an active (or passive) mixer. Then you could have any mix level you want, and also do any level matching (say to get a germanium clipper to the same volume level as an LED clipper) before mixing, etc.
While it would be marginally more circuitry, it would also solve all your problems.

[antonis]

I guess you didn't read my statement above the pic (emphasis added):

I have to admit I didn't... 

Perhaps because "both diodes facing in the same direction" arrangement confused me due to higher forward voltage drop diode uselessness...

[pokus]

Some great idea's. I will try out GGBB's idea today and maybe add a second pair of diodes in parallel the other way around
so it will give me symmetrical clipping. The only disadvantage of this could be, that I guess I now need a triple pot for my initial idea 
So when I want to do the blend thing between two different clipping circuits, can I just split the input signal after a buffer and will it makes sense giving the buffer a gain of 2, which keeps the output level the same? Or doesn't this make any sense at all in a RAT based circuit where the gain of the opamp stage is about 2000?

[antonis]

So when I want to do the blend thing between two different clipping circuits, can I just split the input signal after a buffer and will it makes sense giving the buffer a gain of 2, which keeps the output level the same?

On a splitting node what splits is current - not voltage..
(Kirchhoff's Current Law..)

Only concern you might have about signal level could be in case of phase splitting..

[GGBB]

Some great idea's. I will try out GGBB's idea today and maybe add a second pair of diodes in parallel the other way around
so it will give me symmetrical clipping. The only disadvantage of this could be, that I guess I now need a triple pot for my initial idea

You just need one pot. Follow the pic but replace each diode with an anti-parallel (both directions) diode pair. So you have a lower Vf pair in series with the pot, in parallel with the higher Vf pair.

[pokus]

Oh my mistake, of course the voltage keeps the same sorry 
I've build it with one pot now, a pair of 1n914s and some red LEDs. It works but I only had a B10K pot and there is still only a small area in which the tone changes but I guess with a 1k or a small rev log one it could work quite fine.
Nevertheless the one with the different clipping circuits should be the better option also because of the possibility of adjusting the volume. Any thoughts on passive and active mixer? Which could be the better choice in this case?

[GGBB]

Any thoughts on passive and active mixer? Which could be the better choice in this case?

Better is always subjective. They will be different. Passive gives you one signal clipped varyingly. Active gives you two differently clipped signals blended together. Obviously these are not just two ways to achieve the same result. I would expect that the active setup is going to give you a greater variety of possibilities because of the variable blend control, but none of them will be the same as what the passive setup will produce. You'll just have to experiment to find out what you like.