Stereo Fuzz Circuit

BobTheDog · Posted: Tue Oct 14, 2008 2:20 pm Post subject:

And I think they are called Optical Compressors, I always wondered how they worked, you learn something new everyday!

Inventor · Posted: Tue Oct 14, 2008 3:25 pm Post subject:

Neato!

Uncle Krunkus · Posted: Tue Oct 14, 2008 3:51 pm Post subject:

Inventor · Posted: Tue Oct 14, 2008 5:34 pm Post subject:

You could make the two different channels clip at different levels by using different color LEDs.

Uncle Krunkus · Posted: Tue Oct 14, 2008 5:54 pm Post subject:

Inventor · Posted: Tue Oct 14, 2008 6:10 pm Post subject:

Yes, in fact I was shopping for LEDs on the DigiKey search function (which is really cool BTW) and one of the selection parameters was the voltage of a red LED. it varied from like 1.6V to 1.9V or more, if I recall correctly.

Uncle Krunkus · Posted: Wed Oct 15, 2008 2:05 pm Post subject:

Well, I found a nice candidate fuzz design, and modified it to breadboard up one channel. It's just two op-amps at this stage (1/2 an 074) this could mean fitting it all on one chip! Shocked

It's got nicely rounded knees, which I only just managed to square up by using 1N4148s. Ge's are much nicer. The wave just snakes along between two unspoken boundaries, bouncing and rolling like a slinky. Smile

LEDs seem similar to the Ge's. (BTW difference in drop is very obvious for LEDs)
This is the kind of fuzz I wanted to start with, as it's not too harsh, but easy to roughen up if need be.
I'll get the second channel up and running today, then I'll get out the coils! Wink

Uncle Krunkus · Posted: Fri Oct 17, 2008 6:44 am Post subject:

Well,
Interesting progress.
Scratch the transformer, I've found a better way.
I hooked up the X input on my CRO to the second channel so I could see just how much the two channels differed from each other.
Putting the transformer across the diodes didn't do much at all really, and putting it in series with one of the diodes on each side, had varying success, but didn't really create an effect very different from simple fuzz. In fact I found a few ways of locking the whole thing up into high frequency oscillations.
But,......
In the process of connecting and disconnecting diodes etc. I struck upon the most pure, obvious, simple way of creating stereo fuzz.
Clip the positive going signal off one channel, and clip the negative going signal off the other channel. This not only preserves the classic fuzz sound, it also creates this phase shift effect between the channels. It sounds absolutely beautiful!
I also found a number of optional configurations which include not just substituting different diodes, but bypassing them with various caps (to cut top end) and simply bypassing with a large(ish) resistor. (This just dulls the overdrive effect)
The next stage I'd like to try would be to subtract the original signal from the AC coupled fuzz channels. This should widen the stereo effect even more by reducing the common signal.

BobTheDog · Posted: Fri Oct 17, 2008 1:02 pm Post subject:

Well I wish I could understand a word of it but it sounds interesting.

Would this be a simple circuit for other people to build?

Andy

Uncle Krunkus · Posted: Fri Oct 17, 2008 2:43 pm Post subject:

Yes, it's a simple circuit. Very Happy

The options would require some extra switches and pots, and the associated wiring, but it would be fairly straightforward.
I think the best thing will be to keep it simple. How much feature creep could there be on a fuzz box? Laughing

Only 2 chips as far as I can see.

BobTheDog · Posted: Fri Oct 17, 2008 3:36 pm Post subject:

Could be our first "strings and things" project.

Inventor · Posted: Fri Oct 17, 2008 3:57 pm Post subject:

I understand what he's saying and it sounds kewl! Audio samples, schematics, etc!

Uncle Krunkus · Posted: Sat Oct 18, 2008 12:19 am Post subject:

I'll try and knock up a schem tonight or tomorrow. I need to try the difference amps out and see if that works. If it does, I reckon it'll sound brilliant (literally!) Laughing

Inventor · Posted: Sat Oct 18, 2008 12:29 am Post subject:

What's it called? Anti-proton fuzz? fuzzy navel? spin fuzz?

Post a crude schematic of it and I'll do an Eagle CAD version for ya.

This circuit sounds promising in its simplistic elegance.

Oskar · Joined: Jul 29, 2004 Posts: 1751 Location: Norway

Uncle Krunkus · Posted: Sat Oct 18, 2008 4:30 pm Post subject:

Here's a basic sample.
I'm no guitarist, but you should get the idea.
If you have a look under Soundforge etc. you'll see the spikyness (?) on the upper parts of one side, lower parts of the other.
This is with quite conservative options, ie; I'm using the smooth Ge diodes, part of the signal which could be clipped is going around a bypass resistor, and part is going around a bypass cap. These will be adjustable in the final version. As would switchable diode types.
Add to that I haven't added the difference amplifiers to subtract the original wave, this will "widen" the sound considerably, as the common part of the signal will be attenuated.

Fuzz Field 19-10-08.mp3

Description:

Download

Filename:

Fuzz Field 19-10-08.mp3

Filesize:

962.25 KB

Downloaded:

1204 Time(s)

Inventor · Posted: Sat Oct 18, 2008 9:09 pm Post subject:

I listened to it with my new fancy Sennheiser headphones, and it was amazing! The stereo effect makes the sound seem to travel left and right inside my head... or was that just an echo in my empty skull? No, it was the funky diode action.

The creative, exploratory, experimental approach that you seem to take on your projects appears to have really paid off this time. I wonder if it is possible to enhance the stereo travel effect.

Anyway, well done!

Uncle Krunkus · Posted: Sun Oct 19, 2008 1:23 am Post subject:

Inventor · Posted: Sun Oct 19, 2008 3:09 am Post subject:

Uncle Krunkus · Posted: Sun Oct 19, 2008 3:42 am Post subject:

The reason my difference amp idea didn't have great success, was that I'd forgotten that the clipping op-amp has a gain of about 10 for the side which is not clipped (unadjusted), so I had to put the original through a similar X10 inverter for the difference to make a difference. Laughing

It's not a great clip, it's actually hard to see, but if you look carefully, you can see just how different the two channels are.
I've put one through the X input, and one through the Y. If they were the same (or had a linear relationship (ie; gain change only)) there would be a straight line at an angle on the CRO, a bit like a "/", but as you can (almost) see, we have a very much more complex set of twisting interactions going on.
Schem coming up soon.

Fuzz Field Movie 007.mov

Description:

Download

Filename:

Fuzz Field Movie 007.mov

Filesize:

1.5 MB

Downloaded:

1229 Time(s)

Uncle Krunkus · Posted: Sun Oct 19, 2008 5:09 am Post subject:

Here's the schem,
Updated 20-10-08
Can those of you who have more theory background than me check it over for any obvious problems or questionable practices?
Thanks heaps. Smile

For example, I don't think I really need C2/C5. I've tried bypassing them, and can see no real difference in the response. Would they change the way the clippers work? or act merely to de-couple DC?

By comparison, I know I want C4/C7 as I wanted to remove some of the DC offset from the clipped version before feeding it to the difference amps. I think this makes for more interesting crossover points. Any thoughts? I think maybe they need to be bigger though, as the clip has a bit of a ramp on it, (ie; they're discharging too quickly?)

I'm thinking of replacing R9/R15 with a dual gang pot to control how much of the original is subtracted from the clipped versions. I'm wondering if I should just replace R9/10 and R15/16 with dual pot based dividers (500K?) with the inverting inputs on the wipers? Any thoughts?

Fuzz-Field-20-10-08.gif

Description:

Filesize:

47.91 KB

Viewed:

250 Time(s)

This image has been reduced to fit the page. Click on it to enlarge.

Inventor · Posted: Sun Oct 19, 2008 8:22 pm Post subject:

Yes, thoughts. In the movie I did not detect the stereo shifting effect, I presume the movie is mono?

Those 100nF caps are 15 kOhm at 100 Hz, which is the region of your bass signal fundamentals, and lower obviously at the harmonic region. So they are comparable in impedance to the 10 kOhm resistors they are in series with, and they will shape the waveform. If I recall correctly, a cap on the input side of an inverting opamp has a high-pass effect. So you should hear some difference when you bypass them.

Similarly, the DC blocking caps of 100nF are too small, if they are for DC blocking only. I'd make them bigger.

Instead of using a dual ganged pot on the diffamp inputs, I'd just use a single pot on the prior inverting amp feedback path. That would be cheaper, easier, etc. - always avoid dual ganged pots when designing for cost, as I'm sure you know. Also, putting the pot there just "feels right", as in "I've seen that a lot" and in the other positions it doesn't "feel right".

Just my 2 cents, nice circuit. I may have more comments later once I have a little time to think about it, but I wanted to address your questions right away. Fuzz On!

Uncle Krunkus · Posted: Mon Oct 20, 2008 4:00 am Post subject:

Uncle Krunkus · Posted: Mon Oct 20, 2008 5:51 am Post subject:

I just updated that schematic a couple of posts back with some enhancements,
Got rid of the caps on the inputs to the clippers,
Added pots to vary clipped levels and invert levels,
Tweaked gains for each,
Changed the de-coupling caps to 1uF BP

And I must say it's really starting to sing now. It's got great control and variation.

Inventor · Posted: Mon Oct 20, 2008 9:24 pm Post subject:

What about C1? Shouldn't that also be 1uF since it looks like a DC blocking cap?

I must say, Uncle, it is a really elegant little circuit. Do you mind if I add it to the EChucK collection? Do you know about EChucK?