clipping diodes

[Earthscum]

So far, I've been a fan of any assym clipping of more than 2/1... 3/2 is pretty sweet for overdrives. I haven't tried 5/3, not sure why... I've done 4/3. I went about measuring diode Vf's after purposely selecting combos by ear and found them all to be voltage-wise within the same range of 3/2 to 5/2.

[MartyMart]

I like LED's and 1N4001's in some stuff, but LED clipping can be a bit "Farty" as it decays
which is a bit unpleasant.
Sonically the ones connected to the VB or 4.5v line seem to sound better than the "to ground"
stuff.
I have just modded an OCD to have mosfet & Ge one direction and mosfet & 1N4001 the other and that
is quite a nice "amp" breakup type of tone.
I think a build with several options and the pot also as Mark described is in the pipeline !

MM

[Mark Hammer]

Personally, I think LEDs are fine for clipping purposes in a mains-powered amp that operates off +/-15v or so.  The amount of gain required to get them to do anything, tonally, when operating from a puny 9v battery, however, more than likely exceeds what the semiconductors in the circuit are up to.  In other words, a big chunk of what you may be hearing is not the LEDs, but the op-amps buckling under pressure.

Why?  Although people may talk about pickups producing a volt of output, or whatever, in reality, that higher-amplitude transient is VERY brief, and things soon settle down to a much lower-level amplitude.  So, unless you are applying gobs of gain, any clipping is likely to be short-lived, like a parade that goes down the street in sports car gunning it, rather than at a leisurely pace where everybody can wave and be seen.  If one is applying gain so as to assure that any clipping will be noticeable over a longer portion of the note's lifespan, you need to apply enough gain such that the low-amplitude portions of the note after that initial 200msec are amplified enough to reach the clipping threshold...or at least the bottom of the knee.

But what about that portion of the note while the signal is still high?  If I apply a gain of 70x to an 800mv transient, I get a hypothetical output amplitude of 56V.  Obviously, the 9v-powered op-amp is going to respond with a finger wag, head jiggle, and "Oh no you DIH-int", since it is unlikely to be able to swing beyond +/-3.5-4 volts.  So, again, some of what I'm hearing is the diode, but a big chunk is the op-amp running out of steam.  That may well be part of why people find different tones with different op-amps.  If the clipping threshold was so far below the maximum voltage-swing of the chip that only minimal gain was needed to produce clipping, my guess is chip type would not matter.  But if part of the sound is the manner in which the chip reacts when swinging as wide as it can, chip-to-chip differences can start to become meaningful.

The lowest point/amplitude where clipping could occur for one half wave, and the additional amplitude required to result in clipping for the other half wave, is to my mind, the critical element here.  And if I haven't clubbed the point to death yet,  like a baby seal with a stale baguette, diodes have fixed forward voltages, and individual differences, so you have to do the measurement, do the math, and figure out what sort of Vf to Vf gap you need between half-cycles to get the degree of asymmetry you want for the desired tone.  Having a gain of 500x with a Ge+schottky diode produces asymmetry, in theory, but all you really end up with is a squared half cycle on one side and a squared half-cycle at slightly different amplitude on the other.  That's asymmetry of level, but not clipping.  Conversely, use of 1+2 Green LEDs would also produce asymmetry in some alternate universe, but in the 9v battery one, you'd get some very brief clipping that adds a bit of grit to the pick attack on one half cycle, and precious little on the other, if any.

In effect, you need to think of the asymmetry as "reserve capacity".  In other words, clipping of the other half-cycle that doesn't kick in until you dig in.  You need to ask yourself how much harder you want to have to dig in before that reserve capacity is utilized.

Somewhere out there in web-land is the schematic for the old Elektor project from the late 70's, which had variable clipping point and hardness for each half-cycle.  Ah, here we go:  http://www.diystompboxes.com/pedals/Elektor_variable_fuzz.PDF

[earthtonesaudio]

I have to agree with earthtonesaudio. Diodes are often modeled as simple switches that turn on when forward biased voltage exceeds their threshold voltage, but in reality they have a bit of a knee around the threshold voltage. So, a second set of diodes with a slightly higher threshold could definitely still be clipping, but the first set will be doing most of the work.

After I posted that I tried to find examples of different diode types with their curves plotted on the same graph.  The best example I could find was this:

from here.


...Where it appears the curves for Green and Yellow intersect near the Volts axis, and then Green and Super Red intersect near the top. 

Personally I expected to find more evidence showing overlapping conduction curves, but there's very little out there.  Without researching it myself I assume this lack of data is because the effect is quite small.  In most cases (i.e. the "normal world" outside of guitar electronics) it's probably negligible.  One exception might be clipping for a distortion device, but even there I think it's hard to tell.  I have experimented with placing diodes in parallel in distortions and the difference between Si+Ge versus Ge alone is (to me at least) completely inaudible.

[merlinb]

The best example I could find was this:

From a very recent thread;
http://www.diystompboxes.com/smfforum/index.php?topic=87129.0