Alternatives to diode clipping???

[tedmich]

One of the reasons some folks do not like diode clipping is because one pair or locus of diodes is doing all the heavy lifting. When distributed over multiple loci,the sound can become much more palatable.

here's a circuit from Bernie Hutchins that presentsd an interesting approach to diode clipping:
And another from the late Fred Nachbauer:

see also PAiA's Quadrafuzz:

PAiA: Quadrafuzz Design Analysis by Craig Anderton

LEDs in the pre tube cathode bypass are reported to sound very good, see Merlin's discussion on this.

[Mark Hammer]

The Quadrafuzz was a good start to an idea. Unfortunately, one of the things Craig Anderton didn't do with it was adjust the gain of the different bands to produce the same degree of clipping in each. And with LEDs for the clipping elements, that posed a problem.

But, yes, the basic concept is a useful one.

[classtone]

Boy Howdy, I have a Nady To 2 as well and I'm interested in the removal of the diode mod you did. Can you post pics or otherwise tell me how I can do this? It looks like they must be mounted to the sheet metal holding the tube and circuit board...

[jrfrond]

I am surprised that no one has yet mentioned opamp clipping (either "hitting the rails" with a low-voltage supply, or by using an opamp in open-loop mode), as well as MOSFET buffer clipping i.e. using multiple stages of MOSFET buffers (e.g. CD4049) biased in a linear fashion to achieve "tube-like" clipping. Neither one of these uses diodes.

FWIW- any circuit that clamps the signal to a pre-determined level i.e. has non-linear output, is technically a limiter.

[Mark Hammer]

I find that 4049-based circuits give a nice grunt. Of course, I tend to use bandwidth-limiting cap values, so I don't hear any fizzies that might be there. But, with that caveat, yeah, do consider something that uses a 4049 or 4069.

[jrfrond]

I find that 4049-based circuits give a nice grunt. Of course, I tend to use bandwidth-limiting cap values, so I don't hear any fizzies that might be there. But, with that caveat, yeah, do consider something that uses a 4049 or 4069.

That's what I also like about the 4049 circuits. It has been done in guitar amps, going back to the Sunn Beta Lead series, and is easy to implement in a pedal. In ANY distortion design, even tubes, you need bandwidth limiting, so there is nothing wrong with applying these techniques across the board. For me, the most important step is initial low-frequency cut BEFORE distortion stage, or the intermodulation distortion can get real nasty. That applies to amps as well. I generally apply high-cut AFTER distortion, as the odd-harmonic edge tones and the resulting IM distortion they create adds "snarl". THEN it's time to cut back on those "fizzies". However, there's no free lunch. Too much high-frequency gain, and you can create an oscillator.

[Mark Hammer]

Not that there is anything wrong in your approach, but I take a different one. I prefer to tame the high end before hitting a clipping stage, and then gradually "spool out" the top end. The basic strategy is to avoid production of harmonics of harmonics of harmonics. Keeping upper harmonics safely away from the clipping threshold also tends to emphasize/spotlight lower-order harmonics.

Case in point. I adapted Craig Anderton's initial "Tube Sound Fuzz" that appeared in Guitar Player ( http://gaussmarkov.net/layouts/49r/49r-project.pdf ). His later one that appeared in the EPFM book, and was further developed by Jeorge Tripps as the Way Huge Red Llama pedal, simply uses a 4049 on its own. In contrast to those, I used an op-amp input stage that allowed me to shape the tone of the signal hitting the*MOSFETs in the 4049. I recognize that distortioin is always a matter of taste as well as gear employed, but I find I prefer the arrangement shown over the bare 4049.

You will note that the bandwidth-limiting caps in the 4049 stages also tame the treble. This is pretty much my go-to pedal if I'm "in a Mike Campbell mood" that day. A meaty tone with some bite, but not overly strident, and certainly not fizzy. The resonant boost section of the circuit shown is not required; it's just there to make productive use of the second half of a dual op-amp.

Stellan Lehrberg in Sweden pursued this strategy evern further with his "Slowfinger" ( FortuneCity Web Hosting ). although from the soundclip, his appears to have more sizzle than mine.

[jrfrond]

Y'know Mark, it all depends on what the sound in your head is, and what you are trying to achieve. Everyone hears and wants something different. It's ALL cool! If everyone heard it the same, there wouldn't be so many choices in effects and amps. My approach is: if it sounds good (and nothing is smokin'), it IS good!

[Boy Howdy]

Hey classtone, don't do it!

Yeah, lately I'm kind of regretting removing the diodes from mine. Let me explain.

Lately, I've been using the TD-1 (80s model) a lot (it's very similar to the TO-2). Since I removed the 470 pF cap to get back the treble at that stage of the circuit I find that if I use a 7 band eq after the unit I can get some really great sounds.

The TO-2 is a great unit as is. It could use a bit of bottom maybe. I changed the second coupling cap to a .01mf on mine but I think that's a bit much, a .005 would probably be just about right. But it's fine as is.

As for removing the diodes, the only way the pedal is really useful that way is 1) as a volume booster to overdrive the amp's front end (It get's REAL loud that way. The amp must have a master volume to control it.) and 2) in my amp's pre amp out/in loop. In this mode the amp's preamp overdrives the TO-2's tube. It sounds great like that, but again, very loud.

If I were to do it again, I probably wouldn't. If you do this mod make sure that you can find your way back to stock again. That's where I am now, trying to figure out which ones of all those holes the diodes came out of. And that circuit is a bit delicate what with all those wires to come loose.

All that said, If you want to do this mod, instead of removing the diodes, remove just one leg (so you can easily reverse the mod) of the electrolytic cap just before the diodes. This will remove the diodes from the circuit and is less work, and is more easily reversed.

[mictester]

The Quadrafuzz was a good start to an idea. Unfortunately, one of the things Craig Anderton didn't do with it was adjust the gain of the different bands to produce the same degree of clipping in each. And with LEDs for the clipping elements, that posed a problem.

But, yes, the basic concept is a useful one.

It's trivially easy to do - replace the 3k3 into each of the clipping op-amps with (say) a 10k pot in series with a 1k resistor. You can then easily vary the "Drive" to each clipper.

[Mark Hammer]

Correct. Which is why I'm surprised it wasn't done originally. Not like it would have rendered the circuit dramatically more "complicated".