Anderton Comparator Fuzz

[Aph]

Thanks, it was a bad LT1054.

But now it has a little hiss noise. If i touch the solder side with my finger hiss goes away. Is there anything I can add to circuit to get rid of hiss?

You're possibly hearing the artifacts of the LT1054 clock... which means you have great hearing, as that clock runs around 33kHz!

Check out this revised schematic. Add R7 and C5 and possibly C9 to your circuit. C5 can be as high as 150pf if needed. Read the note under U2 on the schematic.
Make sure you insulate your circuit boards (especially the bottoms) with cardboard or something so they don't short out, and then screw the bottom on your pedal to shield the whole thing, etc.

[rankot]

Or you have bought relabelled 7660 instead of LTC1054?

[Aph]

Or you have bought relabelled 7660 instead of LTC1054?

Mmmm, interesting, since his first 1054 was defective. The number of people running into bad IC problems on this forum makes me wonder.
I can't remember the last time I had a bad IC.

[caspercody]

I actually only had a MAX1044 chio, if that makes a difference.

[Aph]

I actually only had a MAX1044 chio, if that makes a difference.

Connect pin 1 of the 1044 to +9v and add C5. Do not add R7. I couldn't get a 1044 to work with R7 in the circuit.
Let us know.

A 1054 is still better. I would really use one of those. If you use a 1054, just be sure to keep pin 1 unconnected.

[caspercody]

I connected pin 1 to pin 9 on the MAX1044 circuit, and noise gone.

Works with daisy chain power supply.

Thanks!!!!!!!!!!!!

Rob

[Aph]

Great, Rob! Persistence pays off!

[Ratimus]

Whats the point of R3, R2. Why not just connect pin 3 (to Ground (Ironically pin 3 (lt1054) too)?

Bleeder resistors for the power supply bypass caps so they don't stay charged after you power it off.

[Andon]

Sorry to be bringing up an old topic, but it's one of the most recent (and comprehensive) regarding comparator fuzzes.

I breadboarded the Optimum Fuzz Adapter that Jack Orman revised, and it sounds great with humbuckers, but really lags and gets super gated with single coils, suggesting input impedance issues. I tried just tacking a 10K resistor to the input and that helped a bit, but not enough. Eventually I just took the input and connected it straight to non-inverting input and added a 1M resistor to ground and now it responds perfectly, though it lacks some of the gated charm.



What I did is highlighted in red. The 500K pot still works as it should, but again with less gating.

Was there another way to do this that I overlooked? Obviously it works, but I'm just looking for a second opinion I guess.

After that I also tried a modified BMP tone stack to the end after C3 and before before the volume pot (put that after a JFET buffer) and that gave it a lot more depth (which is unrelated to my question, but also neat).

[antonis]

The 500K pot still works as it should

IMHO, it doesn't..

It works more like variable frequency HPF rather than input amplitude set..

[Andon]

Ha ha, true, fair. Is there anything else I could have done beforehand, though? Op amp gain buffer/stage?

[iainpunk]

Ha ha, true, fair. Is there anything else I could have done beforehand, though? Op amp gain buffer/stage?

what specific type of opamp did you use? some opamps really don't like being used like a comparator, and the input bias current can shoot off to one side when its pushed into the rail voltages.
some opamps can do the same function as this fuzz without the hassle of Vbias and getting that to work. the TL071 TL081 CA3130 CA3140 and CA3160 work fine when biased to GND instead of Vbias.

cheers

[Andon]

I'm actually using a TL071, though I've also tried it with a TL022 and an NE4558, etc. to varying success.

EDIT: At this point it may also very well be that my single coils just don't like this fuzz, even though my other guitars do.   

[anotherjim]

So, you have a device that has a huge gain wanting to drive the output high or low depending on which of the 2 inputs is a higher voltage than the other.
To cope with different guitar output levels, it can help to able to trim the bias voltage on one input so you can set the point at which signal causes the output to switch. This will let you control how sensitive or gatey it is.
You can hack what you have to try it out...

Remove your extra 1M.
Disconnect R5 wiper and input signal direct to C4.
Connect top lug of R5 to +9v.
Connect wiper of R5 to the point between C4 and R1
Set R5 midway (never turn it full cw or ccw while trying this out).

R5 now gives a variable bias and you can see if you can find a setting to suit your pickups.
This is an experimental hack, the setting of R5 will be very tricky/fine. A better approach is available with a lower value pot in the middle of a resistive divider. The setting needed will be millivolts away from the fixed Vbias.

Personally, I like CMOS opamps for this. As Iain said, the bias can be ground making it super simple (no Vbias network). Also, those opamps tend to already have input offset that already suit guitar levels and the outputs switch slower/smoother. However, I've never got it to work with a TL0xx chip. TLCxxx instead works great, but they are duals. CA3130 would be a good single amp type.

[rankot]

Could you please reupload this image? This one is lost.

[Andon]

That was me working through a design problem I had when I was brainstorming a circuit and trying to wrap my head around the comparator fuzz architecture. Here's what I ended up going with:



That then feeds into a series 10K resistor followed by a 10K resistor to ground (a voltage divider of sorts to limit the volume of the comparator fuzz - it's very, uh, loud) and that goes into a BMP-style tone stack with a mid boost at noon rather than a scoop, and then eventually an output buffer.

If you'd like to hear it, here's a demo video (it's the fuzz on the left):

[rankot]

Very nice pedal!

[caspercody]

What, if any, tone circuit could be added to this circuit? I tried a BMP tone circuit, but all the way to treble it sounds more bass and treble control seems to be more at noon