OD distortion box question

I found this website which happens to have most of the famous OD/dist box circuits on one page.
http://users.chariot.net.au/~gmarts/ampovdrv.htm

The question is, presumably the change from germanium diodes to silicon-only availability, caused the postioning of the clipping diodes to be moved from, after the opamp gain stage, to in the feedback loop of the op amp. How does that position change affect the type of clipping. Presumably it's some phase/delay interaction with the opamp?

An additional question on that line is, why use 4 diodes? The 3 diodes for assymetrical clipping is easier to understand.

Now I've looked at the circuits I'm tempted to make a box with both hard and soft clipping circuits but mixed not switched. Also 2,3 & 4 diodes, but switched for those.

There's no problem obtaining germanium diodes, I've just bought some.

I've not looked at the circuit, but a silicon diode needs about 0.6V across it before it starts to conduct, so 4 diodes would require 2.4V. Mebbe that has something to do with it??

I don't know if this is any help, or if you already know this, but a Tube Screamer uses a diode in the feedback loop of an op amp to get it's clipping. I assume it's make the clipping more tube screamer-esque?

Hope that's helpful

Daniel C.

That's right. Diodes in the feedback loop deliver softer 'tubier' clipping.

clipping.

Any idea why the position change changes the clipping type? I haven't come across an explanation yet.

I'm not sure I fully understand the question, but if you're curious as to the difference between diodes after the op-amp as opposed to in the feedback loop, I can explain that.

Diodes post op-amp: in this circuit the op-amp is simply a gain stage. The diodes clamp the voltage at the output of the op-amp, meaning that the output can no longer swing rail to rail - it clamps at e.g. +/-1.2V (with two pairs of silicon diodes back to back). There is a little bit of softness, in that the diodes have a 'knee' - i.e. they will start to conduct from about 0.5V (each) and will be fully in conduction from about 0.6V. But basically it's just clamping the output voltage.

Diodes in the feedback loop: entirely different. An op-amp does whatever is necessary at the output to maintain its two inputs (inverting and non-inverting) at the same potential. At low voltages, the diodes do nothing - normally you'd have a resistor in parallel with them to define the gain. At higher voltages, the diodes start to conduct, and the resistor no longer defines the gain - the op-amp becomes a follower with unity gain, or if there is resistance in series with the diodes, it's this resistance (in parallel with the overall feedback resistor) that now defines the gain. To summarise, the gain changes; from high gain at low voltage to lower gain at higher voltages. By manipulating these two gains, one can control the 'clipping' softness characteristics much more closely.

I hope this helps.

That's pretty much it. In the feedback loop you get kind of compression effect, where quiter note is amplified and louder one is not (or at least not as much) and also gets clipped. This sounds closer to the original meaning of 'overdrive', i.e. overdriven amplifier. But obviously solid state is still solid state and you won't get that smooth transition from clean to overdriven tone, yet it is much smoothier than diodes in post op-amp stage.

That's what I was after. Thanks Steve. So how does the 2/3/4 diode variation change things in that area?

I haven't seen the configuration, but if it's just a question of 2/3/4 diodes in series, then it's simply defining the voltage at which the diodes start to conduct, and will have the same effect as varying the gain of that stage.

You can do all sorts of weird non-linear stuff by using sets of diodes and resistors in parallel.

So you might have a germanium (or shottky) with one value of series R, in parallel with a silicon with another value of R, in parallel with a si+ge and another R, then si+si and another R.

Indeed. I've seen this done.

But: I've always been more interested in *controlling* the non-linearity of an amp, rather than hard-wiring it as above. I've related the tale of a compressor that I modified to shape not the envelope, bue the waveform - worked remarkably well. I've also toyed with the idea of using analogue synthesis tricks - e.g. driving a VCA with an envelope follower, but wired with no control delays, i..e. to shape the waveform. Or use transconductance amplifiers... One day I'll actually get around to trying this.