Diode Clipping - does the type (Ge vs Si) make a difference, or just voltage?

[amonte]

For example, lets say that a Ge diode clips at .2V, but a silicon diode clips at .6V.

Is there any audible difference between one silicon diode and three germanium diodes in series? 

Also, is there a chart somewhere that has all of the clipping voltages for Ge, Si, LED, MOSFET, FET, etc.?

[Tim Escobedo]

On paper, there's a difference. But depending on the application, such as simple clipping circuit, the effect is overwhelmingly a result of the voltage drop, so much so as to often mask any other difference.

[Mark Hammer]

Diodes are devices that conduct in selective ways, and conduction takes time.  The speed with which the diode begins to conduct has some impact on the nature of the clipping and many people report hearing such differences.  My own bias, however, is that what people hear when using this diode type or that one, has a great deal more to do with the forward voltage and voltage drop than anything else; partly because the speed differences are occurring at frequencies that are often well above where the guitar and speaker are situated in the spectrum, and because the impact of the level changes introduced is quite substantial and more likely to overshadow any real consequence of diode type and conduction speed.  I wouldn't say that there is NO audible consequence of conduction speed, but not many hear it, and everyone hears the consequences of changes in clipping threshold, so I lean in that direction.  EDIT: I guess Tim agrees.

The quality of clipping is always a function of how much of the signal over the lifespan of any given note is likely to be above or below clipping threshold.  Feed even the most serious head-banger shred-o-matic high gain device with a voice mic signal, and you'll get barely a glimmer of distortion.  Feed your guitar to an unattenuated/unpadded mic preamp and you'll clip the daylights out of it.  Take a pretty sensitive fuzz and feed it with a big body jazz guitar using flatwound strings and a wooden floating bridge, such that the strings have a big transient and a quick decay, and you'll get a different sound out of the fuzz than if you used a solid-body axe with roundwounds and a long natural sustain, because the jazz box's signal level drops from peak to sustaining fundamental quite quickly with great contrast, and only the initial part of the note is above threshold.  Feed ANY instrument into a booster before you plug into a distortion and the distortion will be more intense and last for more of any note's lifespan because it remains above threshold for a longer time.  Plug into a compressor THEN into a distortion and the distortion will have a more consistent tone because the same level (and its relation to clipping threshold) will apply across all notes.

Because the way in which clipping diodes are used also alters the output level of whatever stage they are situated at, this has implications for the degree to which clipping will take place elsewhere, whether within the same circuit, or in subsequent pedals, or in the amp.  In the case of double clipping circuits (e.g., the Big Muff), the ceiling that diodes in an earlier stage may set on the level determines what the consequences of the gain applied in the next stage might be for the dides used there.  Stick diodes with a higher clipping threshold in some pedals in place of lower-threshold units, and you may well be hearing less clipping from the pedal, and more clipping from the amp, combined together, not to mention perhaps clipping from the FETs in a phaser or whatnot.

Generally speaking, they proceed in this order: Schottky, germanium, silicon, 1N4xxx series, red LED other LED colours, going from lowest to highest clipping threshold (the forward voltage at which the diode conducts).  Within each category, you WILL find individual variation, and are recommended to use a DMM (or get one if you don't have one) to measure forward voltage of the specific component to adjust your build to taste.  For example, I've fond Si types can range fropm just under 500mv to somewhere around 650mv, and Ge types can go from 190mv to 270mv or so.  Sometimes you find advantages to matching diodes within pairs, and sometimes sonic advantage is provided by deliberate mismatching.

[mac]

I once compared 1n34, aa117, 1n60 and oa90, all within the same forward voltage, say 250mv ± 5%. Circuit was a std Dist+.
They sounded different. The 1n60 and oa90 closer. But I can't say why. Maybe the ±5% was making the difference, maybe they have different freq response, maybe both.
I tried to compare Si vs Ge but could not find a combination of same type Ge matching a pair of Si.

My Dist+ has selectable diodes, 1n4148s and 1n60s. Besides the volume drop when I switch to Ge, the sound gets darker.

I do not know about diodes construction techniques, but I guess that diode A & diode B have different dopping proportions, materials, etc., that can make them clip different.

Now this questions makes me think about different Si transitors having different Vbe. If FVD makes a difference with diodes it can make it with trasistors as well.

mac