Totally flumoxed - clipping diodes doing nothing?

[timmyo]

Gah - time to ask the experts - I've been staring at this thing for 2 days 

So - I have put together a basic using the following schematic / layout : http://www.beavisaudio.com/bboard/projects/bbp_BoutiqueTubeScreamer_Rev1_1.pdf

Cutting a long story short, the clipping diodes seem to be doing nothing. Well almost always...

I started with a pair of LEDs and played. Fine. Then changed to some different LEDs. Same sound. Then Assymetric LEDs. Same sound.

suspiscious.

So I removed all the LEDs. Same sound. Completely the same.

If I short out the clipping diode section I get a drop in volume and drive. If I leave the clipping section open cct I get the overdriven sound I was hearing in those tests. If I add back to back diodes in there I still get that same sound.

I scrabbled around for what other diodes I had : putting OA91 germaniums back to back gives a different (quieter, splatty decayed) sound, and some BAT 85s give a raspyish sound, but any LEDs may as well not be there.
I am confused.

Is there any "gotcha" I may be missing? (the LEDs are good - I checked them and they light with +9v and a resistor in series)

fwiw the OpAmp is a 5532 (and yes tried 2 of them - same results) and there is 4.5v showing at the input to the clipping section.

Anyone offer any insight?

thanks,
Tim

[Mark Hammer]

LEDs will require a signal that is +/-1500mv in order to go into clipping.    GE diodes will require a signal in the vicinity of +/-230mv in order to clip.

A standard TS will provide a maximum gain of 127x.  Assuming that a guitar signal will start out with a peak of 80-100mv when picking a single note (250mv if you slam a chord), and then decline very shortly thereafter, and assuming that the TS shaves off all the bass on the input, dropping that overall amplitude even further, you may well only hear the clipping provided by the LEDs for a brief instant at the initial pick attack, and only if you crank the gain up.

The modus operandi of the TS is that it provide gentle clipping, and to do that with only modest gain, those diodes have to have a lower forward voltage.  Otherwise, if one wishes to use LEDs, the gain has to be increased, conceivably by using a 1M gain pot.

[BAARON]

LED clipping is really gentle, and with a large clipping threshold as Mark mentioned above.  Symmetrical LED clipping should be smoother than no clipping diodes at all, but the amount of clipping and the volume will still be similar.

And if you have asymmetric LED clipping (which I assume means one LED in one direction and two in the other) then you're going to end up with a clipping level (probably 1.8v : 3.6v) that's getting close to the power rail clipping that the op-amp would experience naturally (±4.5v)... it's pretty hard to notice a difference at that point at all.

Remember that to double the apparent volume level of something (as our ears hear it), you have to increase power by a factor of 10.  0.3v clipping diodes (Germanium) vs 3.6v clipping diodes (two LEDs in series) will have a Very noticable difference (12x), but 3.6v series LEDs vs 4.5v supply rail clipping is only a 1.25x increase... far from the 10x you'd need for an audible doubling in volume.  Hence why the difference is so much more subtle than Ge vs LED.

[timmyo]

Guys, thanks for that, that really helps me make sense of it.

It may be that I'm not quite hearing the subtle differences as I've been testing it late at night after the kids are in bed using headphones via PC modeller - perhaps once I get it up loud through the amp the subtle differences will be more apparent.

Cheers for the pointers,
Tim

[Mark Hammer]

One of the things that escapes a lot of people - not just yourself - is that the opportunity to hear something as distorted is a function of how long and what portion of the input signal remains above the clipping threshold; what I like to call the "proximity to clip".  To be able to hear something as distorted, the clipping needs to continue well beyond the initial 10millisecond transient peak.  One of the reasons why pedals that produce distortion have so much gain in them is to assure that the incoming signal remains above the clipping threshold for a long time.