Multiple Nested Diode Clippers In Opamp Feedback Loop

[lovekraft0]

This may be old news to a lot of you, but I just discovered this while playing about on my breadboard:

It's not quite ready for prime time - it'll probably need about 10-20dB of boost at the input, almost certainly a cap in parallel with R1 for fizz reduction, and possibly tone controls on the output, but it has such an "amp-ish" vibe (especially as the notes decay) that I felt compelled to share it. Please feel free to use, modify, scale and/or mangle as you see fit if you find it useful - I've seen similar ideas in the late Fred Nachbauer's designs, and one of those Fender SS amps used something vaguely similar, so this isn't anything new, but even without any tweaking, it shows potential. If you can use it, have at it!

[R.G.]

It is a good circuit. About the only thing wrong with it is that it needs a lot of signal, as you alluded to. It might even want a bigger power supply to really open up, maybe bipolar. I really like the sound of circuits like this that lead you into clipping very softly.

Good work!

[Lateksi]

I've thinking about making something like that, but it seemed, that one 9v battery wasn't really enough. Maybe germaium diodes would help to lower the needed voltage.

[Paul Perry (Frostwave)]

If you havn't got Ge, then Schottky diodes might be worth a try, if you want to get the voltages down.

[Mark Hammer]

There was a circuit in ETI decades ago called the "Hyperfuzz" that used something like this.  I have a copy at home.  I'll see if I can scan and post it.

The suggestions for lower Vf diodes, such as germanium and Schottky are good ones. You can also make a point of selecting diodes based on Vf measurement too.  Personally, I find a fair amount of variation within a given part #.  A "common" 1N34A can have anything from a 190-260mv forward voltage.  If clean gain limitations of a 9v-powered circuit are an issue, then you want to aim for the lowest forward voltage within a selected diode category.

There is also this approach to gradual clipping from an old Electronotes issue, although like any of these multi-diode-stage designs it too requires rather substantial input signals to do what it is intended to do.  You will note that Electronotes was/is geared to analog synthesis buffs, so the expected audio signals are line level or hotter.

[MartyMart]

Yeah the Hyperfuzz had a pair of BC549's up front to "knock the shit out of the opamp" ( that's a technical EE term )
It has a couple of diodes in the FB loop and several w/reistors to ground in a switch matrix also, sounds good IMO

MM.

[Doug_H]

Looks like a great idea. I've seen this kind of thing here & there before, can't remember where. In the amp emulator thread there's a link to musicman amp schems that use a transistor in the feedback loop to gradually turn on a diode clipping path.

One thing I wonder though- why not just use a non-inverting op amp circuit? This would give you higher Zin, requiring less current drive, no series resistor which can introduce noise, etc, etc...

[DDD]

OK, Doug, you're right.
Moreover, the non-inverting circuit clips softer by itself, so it's preferable in the present configuration.
At the same time, the higher the value of the resitors in the feedback loop the lower voltage is needed to "open" the diodes, so less voltage of the power supply is needed. Make the resistance 5 times higher and you'll need about 20 to 40 per cent (approx.) less voltage to supply the circuit.

[stm]

That kind of arrangement has been used in some Fender Amps, as you can see below.  IMO based on the many sims I've done it works best when the forward voltages of the diodes are reduced as you have more clipping--the Fender circuit follows this rule.  In this respect, I'd suggest a pair of red LEDs for D1,D2 then a pair of silicon diodes for D3,D4 and finally a pair of Shotky or Ge diodes for D5,D6.  In this way you are more likely to get all diodes working without needing too high input levels.

Resistance scaling seems fine with the values you mention (470k, 270k, 150k), however you can be a bit more aggressive and use 470k, 220k and 100k instead, or even more gentle and use a milder series such as 470k, 330k and 220k.  Many factors to experiment with.

Has anyone looked at what Fender is doing in a few of their amps?  I think this is what they call "Dyna-Touch" or something along these likes.  Here's a pic of a section of the Champion 30 solid state amp.



I think their creating a "piece-wise" approach with this one.

Alf

Even though using the opamp in its noninverting form seems fine, I'd say in this case you'd need higher input voltages to travel all the dynamic range, as oppossed to the inverting case where it is a matter of reducing the input resistor to attain more gain only.

[lovekraft0]

Thanks for all the suggestions - here's a somewhat more complete idea including some of them (especially stm's idea about decreasing Vf for the diodes - genius!), with some basic fizz reduction and some gain at the front end.

There's a good chance that hot humbuckers will clip the input opamp at max gain, so that gain structure might need modifying, and I didn't include decoupling caps for the power supply, either. Looking at it now, it reminds me of the HAO Rust Driver - nothing new under the sun, I guess. Again, feel free to play with it!