Simulating the Bernie Hutchins distortion

Started by soggybag, September 14, 2024, 07:15:40 PM

Previous topic - Next topic

soggybag

For fun I mocked up this idea from Bernie Hutchins, post by Mark Hammer years ago.

https://tinyurl.com/22hkkh53

How can I get a better scope view? I'd like to set this up see the effects of the circuit. I'm guessing, besides adjusting the timing and the view, I need to get the input wave form to vary peak voltage to get the various stages to have an effect.

ElectricDruid

Yes, running different amplitude signals through it would be good. It looks like a sort of analogue piecewise-linear-interpolation-of-a-curve. Roughly!

What *is* that 10uF doing on the output of the op-amp? I thought that was supposed to be a no-no. Does it affect the response at all? It's the only thing there that could be frequency-dependent. Otherwise you could run the whole circuit at 2Hz instead and see much better what's going on.

Rob Strand

#2
Quote from: ElectricDruid on September 14, 2024, 08:09:51 PMYes, running different amplitude signals through it would be good. It looks like a sort of analogue piecewise-linear-interpolation-of-a-curve. Roughly!

What *is* that 10uF doing on the output of the op-amp? I thought that was supposed to be a no-no. Does it affect the response at all? It's the only thing there that could be frequency-dependent. Otherwise you could run the whole circuit at 2Hz instead and see much better what's going on.

We don't know what the opamp model is.  If the opamp is ideal then it will totally ignore the 10uF since it has zero output impedance, infinite drive capability and hence no effect on stability.  In practice it's a different matter.

As for fine details in the non-linearity:   We don't know how well the diode models represent reality.  That's especially true for zener models.   In a simplified simulation, it's highly likely forward biased zeners behave exactly like silicon diodes whereas in reality they follow a different curve (a different ideality factor).

Using LTspice instead of Falsad will give you more resolution on the plots.
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

soggybag

I put the cap at the output to have something to graph.

antonis

Quote from: soggybag on September 14, 2024, 10:45:39 PMI put the cap at the output to have something to graph.

Try to put the cap in series (outside the feedback loop) with a high value resistor to GND.. :icon_wink:
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

Rob Strand

#5
Quote from: soggybag on September 14, 2024, 10:45:39 PMI put the cap at the output to have something to graph
In general a 1MEG resistor is a much better choice.

The 100k feedback resistor has a bad connection (red dot?).   The two wires on the left of the resistor need to be deleted and redrawn.   At the moment it's like a single wire passing over the end of the resistor without connecting to it - well I think that's what is wrong.
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

amptramp

I don't know how much it will affect the simulation, but a small-value cap from the op amp output to the inverting input is a normal part of the design since capacitance to ground at the inverting input results in a rising gain vs. frequency response which intercepts the falling gain vs. frequency response of the op amp and causes oscillation.  Various fuzzes of this type use 47 pF to 100 pF for this.

Also, op amps have a rising output impedance with frequency so that a direct connection from the output to a load capacitance will add a lag to the feedback that causes instability because a feedback lead is what is necessary for stability.

m4268588

I don't know about this circuit, but is direction of zener correct with this?
https://www.falstad.com/circuit/circuitjs.html?ctz= ...

soggybag

That was really helpful.

There was a discussion about this, long ago, on this forum. I had been wondering for years what is really happening with this circuit. Seems like there is a little crossover distortion, and a lot of that tube screamer, rounded clipping, at the top, with heavier clipping at the bottom of the wave.

ElectricDruid

To be honest, it looks pretty complicated for what it achieves. There's a lot of weighted stages in there to get something which is pretty much just soft clipping on one side with rounded-but-harder clipping on the other side.
I suppose the devil is in the details though.

Mark Hammer

I think it is worth noting that Bernie & Co. work in the domain of synthesizers, rather than FX pedals.  So bear in mind that this circuit is likely anticipating line level signals, or greater, not puny mag pickup signals.

soggybag

I understand that this was meant for modular synth voltages. I was just trying to get an idea of how this would differ from your standard tube screamer.

From the simulation, it looks like there is a small about of crossover distortion, a flat line at the zero crossing, and a TS like flattening at the top of the wave, with a much more pronounced flattening at the bottom of the wave.

Here is a picture from the sim by m4268588 above.