Clipping diodes: Schottkys?

[isher1992]

I still have a vague understanding of clipping diodes...forward voltages and all that.  I know in most of my overdrives and distortions, I prefer 1N4001 as opposed to 1N4148... But I've been looking at modding a DOD Grunge I picked up, and got to wondering, what would happen if I put a Schottky in series with a 1n4001, and have asymmetrical clipping?  A Schottky with a 1n4001 in one direction, with a single 1n4001 in the other.  I'm curious how this would impact the tone... would it clip harder, or a touch softer?  That's the part I am vague on.

[vigilante397]

what would happen if I put a Schottky in series with a 1n4001, and have asymmetrical clipping?  A Schottky with a 1n4001 in one direction, with a single 1n4001 in the other.  I'm curious how this would impact the tone... would it clip harder, or a touch softer?

Breadboard breadboard breadboard! 

You can do the math for what the output voltage will be, but the numbers don't always reflect tone. You never really know how something is going to sound in a circuit until you try it out

[isher1992]

I was just trying to get a ballpark idea.  I don't really want it clip noticeably harder or softer, I just want to subtley make it asymmetric.  I may just have to try it and see.  Who knows, it may be so marginal I don't notice.

I am curious how the Grunge would sound with 1n4001's though.  Also should change the opamps while I am at it.

[MrStab]

IME, a reeeally subtle asymmetrical clip (e.g a 1n4148 in parallel with a 1n4001) is still effective to my ear, and sounds a bit more natural than a bigger difference such as a lone Schottky paralleled with a 1n4148. not a scientific response at all, but i just wanna emphasise that you don't necessarily need a huge difference. what you're proposing would mean only about ~200mV difference in thresholds though, whereas some folk use much larger than that, so it's all down to taste.

[Keppy]

Try it. My favorite combination involves MOSFETs and Schottkies. They're not more or less valid than other clipping diodes, so you just have to see if you like it.

[aron]

Yes, something like this you would just have to try it and see how it affects the feel and the sound. I like a little mismatch at times.

[amptramp]

To answer the OP's question, it would clip softer because there is more headroom before clipping with a Schottky in series with a 1N4001 on one side and a 1N4001 on the other side, but only on one side of the waveform.  Symmetrical clipping tends to increase odd-order harmonics whereas asymmetrical clipping tends to produce more even order harmonics.

[Mark Hammer]

The degree of symmetry/asymmetry will always depend on the signal level, relative to the diode complement.  In theory, one could have eight 1N4001s in series for one half cycle, and a single BAT46 Schottky for the other half cycle, and not have asymmetrical clipping, if you hit those diodes with a high-enough signal.  Yup, that's right.  If the signal were 15V P-P, it would be squared off for both half cycles, with one a much smaller amplitude than the other.  The asymmetry arises when one half-cycle is more likely to clip, and add those harmonics, than the other half cycle.  So one chooses the diode complement, in part, to suit the anticipated signal level.

Works the other way, too.  Three silicon diodes in one direction and 4 in the other will not yield anything asymmetrical if the maximum anticipated signal level is not up to the forward voltage of the 3-diode path.

None of this is to dismiss the desire to aim for asymmetry.  It's just important to keep in mind that diodes have a fixed forward voltage, and your input signal level varies quite a bit.  So actually achieving asymmetrical clipping might not happen, even though that's what one is trying to do.  Matching the clipping elements to the circumstance is important in getting the desired result.

[anotherjim]

I've noticed that the effect of asymmetrical diodes on hard clipping stages -  the production of 2nd & other even numbered harmonics, doesn't always leap out at me.
If the diodes are driven from an amp that is clipping symmetrically, and the resulting wave is pretty much square (equal duty cycle), then odd diodes don't produce asymmetry - it just shifts the zero crossing up or down & it looks asymmetric on the scope versus vref, but it will self centre again the moment it passes through a coupling capacitor. No second harmonic seen & none heard!
Recently I did the "Keeley" mods to a Boss DS-1 (Highly successful!). I like the sound with the red LED versus the ordinary signal diode, but it doesn't sound an octave up at all, despite it MUST be asymmetric!

But I have on a breadboard, a RAT type clipper that I'm planning to build to use up some odd germanium bits. An 0A97 diode and a faulty AC127 with a emitter-collector short circuit & using its working base-emitter as a diode. Got bags of second harmonic. Then I noticed I've a coupling cap before the clippers, so when the amp clips, it isn't a square wave hitting the clipper, but a high pass filtered "saw" wave due to the cap. The output from the diodes is definitely asymmetric since the faulty transistor and diode have different Vf.
The thing with the germanium bits isn't, I'm sure, essential so long as there's different Vf - although it does no harm to claim some unique 40year old f*****-up Mullard mojo!

[ashcat_lt]

I don't completely believe that you can get real consistent asymmetrical clipping in this situation.  I have a Rat clone that I built with several asym options, and they don't work for the same reason.  The reason was almost touched above: asymmetrical clipping looks pretty much like a DC offset, and the coupling caps around it won't support that.  They will tend to float up or down to center the signal between the diodes.  This means that when signal is initially applied, you might see some asymmetry, but if you watch a sustaining note, you'll see it float up and get symmetrical again.

I haven't found a good way to get around that, myself.  Pulldown resistors can help, but sometimes they need to be at values that affect other things in the circuit.  One would think that adding a DC bias on your own, and/or biasing the "other end" of the diodes differently could do it, but I'm not completely sure that's as easy and straight forward as it sounds either.

Note also that two diodes in series to ground clips exactly the same as one would if it was working on a signal half the size.  This also works out to where DC offset with symmetrical diodes is about the same thing as no DC offset with an asym pair. 

[Keppy]

This means that when signal is initially applied, you might see some asymmetry, but if you watch a sustaining note, you'll see it float up and get symmetrical again.

I'm curious as to whether that can be used to good effect, so that a note's harmonics change as the note sustains rather than remaining static. Have you noticed any positive effect?

This "float" can be prevented by not using a coupling cap between the opamp and diodes, and by connecting the diodes to Vref instead of ground, as in the OCD. Whether that has a noticeable effect, I don't know.

[PRR]

> a Schottky in series with a 1n4001

Series, it does not break-down until 50V one way, 30V-50V the other way. In 9V work it never clips.

As MrStab mention in passing, think "parallel". That's nominally 0.6V one way 0.2V the other way.

As mentioned, very heavy clipping and especially with capacitors in the chain will ultimately go essentially symmetric.

You come into the garage very drunk with 6 foot clearance one side, 2 foot the other side. ONE side of the car is beat-up much more than the other side.

You come into the garage VERY-VERY-VERY drunk 6 foot 2 foot blah-blah. BOTH sides of the car are beat-up beyond recognition.

Musically useful because notes fade. The pluck comes out AAAA!!, flat both sides. As the string decays, it transitions to an AaAa! half-bent asymmetry. Then to an aaaa non-bent clean sound. Working among these three zones gives you "timbre" control similar to a vocalist and other musical instruments.

[Keppy]

> a Schottky in series with a 1n4001

Series, it does not break-down until 50V one way, 30V-50V the other way. In 9V work it never clips.

Paul, I think you misunderstood the OP. Right after the sentence you quoted, he wrote:

A Schottky with a 1n4001 in one direction, with a single 1n4001 in the other.

So it seems the OP is looking at a Schottky in series with 4001 oriented in the same direction, with the combination paralleled with a 4001 with the opposite orientation.

Bad ASCII version:
--->->
<---

[Mark Hammer]

Personally, I think a great deal is misattributed to asymmetry per se that is really due to other things.  Raising the threshold for one half cycle gets you a little more headroom, but more importantly, more output.  And with greater output, the next thing in line gets pushed differently.  Is what you hear a reflection of the particular order of harmonics the pedal generated, or is what you hear a consequence of being able to have the added harmonic content disappear faster as the signal falls below the forward voltage for the half-cycle with more diodes, or is it simply because your output is hotter into the next stage or the next device?  I also hasten to remind people that one runs into headroom limitations whenever feeding a guitar signal into a 9v-powered circuit and applying even modest gain.

So yeah, you know you monkeyed with the diode complement.  But is what you're hearing directly because of that, or is it because of something else?  I think the jury is taking a long sequestered vacation on that one.