Zener diode's as clipping diodes?

[tungngruv]

Can Zener diodes be used as clipping diodes? I have used LED's as clipping diodes and I have seen LED's used in place of Zener diodes to protect mosfets (I think). I have a couple extra and thought I'd ask before getting the solder braid out. Thanks for any ideas about this.

[Processaurus]

Yes!  but check this out, to get it to work, put a regular Si diode in series with each zener, but backwards, cathode to cathode for the first zener/Si pair, and anode to anode for the second pair, then tie the free ends together and use it like the usual clipping diode arrangement.  the "}" is the zener cathode.



---l>l---{<l---
---l<l---l>}---

They clip very softly in this configuration.  It has a higher clipping threshold too, though, so to use it in a regular circuit you'd need to increase the gain to get the same amount of distortion.  I seriously doubt I'm the first person ever to figure this out, but I'd be curious if anyone else here has "invented" this implementation.

))<>((

[johngreene]

Actually you only need 2 zeners and instead of putting them 'anti-parallel' you put them in series cathode to cathode. Each zener preventing the other from conducting in the 'wrong' direction.

And yes, it's been done many times before.

--john

(edit) the order doesn't matter, cathode to cathode or anode to anode will both work the same.

[Processaurus]

How about that.  Being able to set the threshold by the value of the zener could be good for shaping LFOs waveforms too, to get all hypertriangular and sinusoidal on it.

[Gus]

also good grid to grid on some power tubes in PP amps to limit the voltage drive.  Steve A had/has something at blue guitar I believe.

[tungngruv]

Thanks for all of the answers guys, this place is awesome.

[R.G.]

Guitar Effects FAQ, GEO, http://www.geofex.com

[Transmogrifox]

A while ago somebody posted a link to the circuit used by Crate for their "tube emulation" patent.  It used a zener in the clipping stage.  You won't be able to make it work on a 9V battery operated circuit because you'll hit the op amp rails before the Zener opens.  Even the 2.4 volt Zener will leave you dangerously close to the rails since the battery would have to be super fresh and right on 9V to work correctly unless you use a rail-to-rail opamp.

I would recommend a circuit with 2 batts: +/- 9V supply and a pair of 5.1 volt zeners.  That way you have about 2V to spare before you run into a normal op amp's rail.  The down side is that you need an op amp with a fast slew rate to avoid slew rate distortion (unless you want that).

EDIT:  Upon further thinking it occured to me that the gain is going to determine the slew rate requirements, not the clipped peak amplitude.  If you use the same gain as a TS, then slew rate is not a problem.  However, you have 8x the headroom before clipping than the TS, so you need 8x the gain to get the same amount of distortion.  This means you need 8x the slew rate.  On a 100mV amplitude input signal, this means you need about 11V/us for a 20kHz signal for a gain of 100.  Cascading stages giving little bits of gain per stage won't help, either.  The op amps in the final stages will still have to track the same steep slope no matter where or how it was made steep.   On a 13V/us opamp (TL071 for example) you could probably have a gain of around 400-500.  The reason being, I don't think it would be terribly ugly if you were slewing frequencies above 5kHz, so that improves your requirements by a factor of 4.  This makes me wonder if this is a significant why many super high-gain pedals get to sounding wooly and loose, instead of staying chunky and crunchy.  It doesn't take much more gain until the 13V/us op amp can't even keep up with 500Hz signals.
Take it easy

[WGTP]

R.G. has mentioned the 1N5819 as having a low threshold.   

[syndromet]

I modded my Samick solid state amp, and added a switch for clipping diodes. One of the diodes was a 6.2 V Zener, and it sounds great. Nice little breakup like when pushing the clean channel of a tube amp a little to hard. I Just put it in like a regular diode, no back to back or parallel or anything.

[Sir H C]

Zener diodes around 5v operation are usually pretty leaky and don't work too well.

[MartyMart]

R.G. has mentioned the 1N5819 as having a low threshold.   

I've tried using 1N5819's a few times as clippers, they sounded horrible !!
Technically, they should clip something like a 1N34A but I found them to be
un-usable  (  2c  )

MM.

[Processaurus]

I would recommend a circuit with 2 batts: +/- 9V supply and a pair of 5.1 volt zeners.  That way you have about 2V to spare before you run into a normal op amp's rail.  The down side is that you need an op amp with a fast slew rate to avoid slew rate distortion (unless you want that).

The opamp circuit I was getting good soft clipping with some 3.9v zeners was running off a +/- 9v supply.