Adding blend control to Big Muff

[PBE6]

Inspired by the DIY distortion pedal thread, I decided to crack open my  Green Russian Big Muff Pi (http://www.generalguitargadgets.com/pdf/ggg_bmp_rusg_sc.pdf) and add a blend control so when I use it live on bass my sound doesn't get subsumed.

Here was my bright idea:


Unfortunately it's not working quite right. With the blend knob all the way to the dirty side it sounds right, but with the knob all the way to the clean side there's a severe high end roll-off and the fuzz controls are still affecting the sound somewhat. I checked the connections and bias voltages and they seem fine, and the clean signal looks fine on the scope until it gets past the opamp output capacitor at which point the controls start affecting it. I'm a little confused.

One thought I had was that maybe the blend pot is too small to block out the dirty sound, but that still wouldn't explain the tone sucking. Does the circuit look reasonable? Or is there a severe impedance mismatch I'm missing?

Any help would be appreciated.

[edvard]

Interesting...  I built an op-amp Big Muff for a friend years ago, actually two of them.  He absolutely insisted on having a blend control, and oddly enough it wouldn't work with anything but a 20k linear pot, exactly as you show, so I don't think your troubles are a too-small pot.  I didn't use buffering or anything, just a couple low-value resistors to smooth the mixing (do a search for passive mixer), which is where I think your troubles are from.  Put resistors (680Ω - 10k, adjust to taste) between the output caps of your buffers and the outside legs of the blend pot. 
Another thing I'd check is if the clean side is in- or out-of-phase with the dirty side.  I do remember one box I modded for him where the two signals were out of phase, and as the blend swept through the middle, he got strange nasal sounds because the fundamentals were cancelling, but clipping harmonics came through fine.  He loved it.  Try using an inverting op-amp circuit coming from one of the signal taps to flip things around.

That same guy wanted blend on every box he wanted me to build, and every box I tried was a completely different scenario, which made designing blend controls a massive PITA.  I wish I had notated the original schematic, maybe I'd be a little more helpful, but I definitely remember it was based on passive mixer topologies I found in electronic reference books in the library, and a 20k pot. 

[MaxPower]

How about something like this: http://www.ecircuitcenter.com/Circuits/opsum/opsum.htm

V1 and V2 would be your clean and dirty signals. R1, R2, and R3 could be replaced with pots as level/gain controls.

EDIT:
Bloody hell. I tend to use two 9V batteries to power my op amp circuits. I assume this circuit will work okay with a 9V power supply. With proper biasing obviously.

[aion]

Might be a little overkill for what you need, but the Deluxe Big Muff Pi from around 1978 is a BMP paralleled with a Soul Preacher compressor and it lets you blend between the two. It can get some pretty interesting sounds. I repaired one of these maybe 3 years ago so I got a chance to play around with it - it was a lot of fun. I think they're pretty rare.

Here's the schematic:
http://s193.photobucket.com/user/therotagilla/media/DBM-78.jpg.html

It's the op-amp version of the Big Muff. There was a later version that used a 3080 chip for the Soul Preacher, but the one above (EH-3053) is considered to be a lot better by people who have played both.

[R.G.]

Here was my bright idea:
...
Unfortunately it's not working quite right. With the blend knob all the way to the dirty side it sounds right, but with the knob all the way to the clean side there's a severe high end roll-off and the fuzz controls are still affecting the sound somewhat.

There are two problems.
(1) The clean sound is not clean. The pickup driving the big muff input doesn't much care about the added opamp, but the 39K input resistance on the big muff is a big load, and the loading actually helps when it wipes some treble off the signal before distortion. However, if you take off the clean signal there, you still have the loading of the muff input. A buffer before the muff would prevent the clean sound from being loaded down, but would also affect the tone of the dirty sound.
(2) "Blends" done this way are variable mixers with poor performance. Their virtue is that they are quick and cheap. Remember the old saw - fast, cheap, good; you may pick any two. A lot of mixer front end design is intended to make the channels not interact. This is an application where your ears are telling that fast and cheap are not good enough for you.

[PBE6]

Thanks for the replies.

R.G., I have invoked that old saw many times myself (my favourite configuration is cheap and good, but I never seem to have the time!)

Maybe the main issue is that the "clean" sound is not clean, as all I'm really looking for is a bit more definition in the blend. If that's the case, I may give the buffer a try. I use EMG active pickups live on bass, but I used a passive bass to test this circuit. I'll try it with both and see if I can find a sound I like.

Also, I thought I *was* being conscientious by buffering the signals on the front end and back end of this blend device! But you're saying in general this type of mixer is a poor performer? Why is that? I based mine on Runoffgroove's splitter-blender, did I miss something there? Would you recommend alternatives like using an inverting opamp mixer as MaxPower suggested, or a passive mixer as implemented by edvard?

[PBE6]

Might be a little overkill for what you need, but the Deluxe Big Muff Pi from around 1978 is a BMP paralleled with a Soul Preacher compressor and it lets you blend between the two. It can get some pretty interesting sounds. I repaired one of these maybe 3 years ago so I got a chance to play around with it - it was a lot of fun. I think they're pretty rare.

Here's the schematic:
http://s193.photobucket.com/user/therotagilla/media/DBM-78.jpg.html

It's the op-amp version of the Big Muff. There was a later version that used a 3080 chip for the Soul Preacher, but the one above (EH-3053) is considered to be a lot better by people who have played both.

Sounds great, but I can't seem to open the JPEG. Found this instead:

http://eu11.stripper.jp/pulcino/blog/archives/images/DeluxeBigmuff.GIF

[R.G.]

Maybe the main issue is that the "clean" sound is not clean, as all I'm really looking for is a bit more definition in the blend. If that's the case, I may give the buffer a try. I use EMG active pickups live on bass, but I used a passive bass to test this circuit. I'll try it with both and see if I can find a sound I like.

I would expect the active pickups to lose less from the big muff circuit loading. It's funny - the muff only loads the signal when it's switched in, but doing a clean blend then uses this treble-loaded signal as "clean". The issue is with the muff loading, but that's not audible until you buffer the clean and listen to that too.

Also, I thought I *was* being conscientious by buffering the signals on the front end and back end of this blend device! But you're saying in general this type of mixer is a poor performer? Why is that? I based mine on Runoffgroove's splitter-blender, did I miss something there? Would you recommend alternatives like using an inverting opamp mixer as MaxPower suggested, or a passive mixer as implemented by edvard?

Buffering before the pot is definitely a step in the right direction, but the side-to-side pot blender is just not all that great a performer. Fixing it requires adding enough parts to do a different version well in most cases. You're already two opamps and a follower into this. If you used a third opamp instead of the follower, you could do an inverting panner circuit and re-vert it to non-inverting in the mixer/output opamp, and the nominal performance as a blender/mixer should be pretty good. I'm partial to this circuit over the other similar mixer setups for the special case of alternating between two signals.

If I were doing this, I'd use a high(ish) impedance inverter for buffering the clean signal, an inverter for taking off the dirty signal, and the circuit in the upper left hand box in http://www.geofex.com/Article_Folders/panner.pdf. By pre-inverting the signals, the final panner/mixer stage re-inverts them to normal. The panning/mixing is done with currents, and the two sides do not interact at all. It's a special-case circuit, but it's really handy when you want to move/blend/mix two inputs only.

[PBE6]

Thanks R.G. I will check this out!!

[aion]

Might be a little overkill for what you need, but the Deluxe Big Muff Pi from around 1978 is a BMP paralleled with a Soul Preacher compressor and it lets you blend between the two. It can get some pretty interesting sounds. I repaired one of these maybe 3 years ago so I got a chance to play around with it - it was a lot of fun. I think they're pretty rare.

Here's the schematic:
http://s193.photobucket.com/user/therotagilla/media/DBM-78.jpg.html

It's the op-amp version of the Big Muff. There was a later version that used a 3080 chip for the Soul Preacher, but the one above (EH-3053) is considered to be a lot better by people who have played both.

Sounds great, but I can't seem to open the JPEG. Found this instead:

http://eu11.stripper.jp/pulcino/blog/archives/images/DeluxeBigmuff.GIF

That one is the 3080 version, the one people say is the inferior of the two. Here's a direct embed of the one I posted - maybe this will work!

[PBE6]

Ok, help?!?

I finally got around to building a new blend control for my Big Muff as per R.G.'s suggestion. The blend part works great! Somehow I managed to screw up the gain stage instead

Here's the circuit:


A probe on the inverting input shows that the blend is working very nicely (there is some very minor kinking on the fully clean signal, but I don't think it'll be a big problem as this pedal is meant to be dirty). The max voltage is around 350 mV. However, once it goes through the gain stage it's getting completely squared off. A probe on the output shows a square wave with a positive swing of 4 V and a negative swing of -3 V. Shorting the inverting input and the output appears to give a half-rectified version of the blend. Connecting a 4k7 resistor in parallel with the 47k resistor doesn't seem to have much of an effect.

I am seriously stumped. Anyone see something simple I'm missing?

[electrosonic]

At first glance, the last op amp is referenced to ground instead of Vref.

[Keppy]

At first glance, the last op amp is referenced to ground instead of Vref.

Same goes for the blend pot.

A probe on the inverting input shows that the blend is working very nicely.

Keep in mind that the opamp is attempting to keep its inverting input identical to the non-inverting one, which in this case means having no signal at all. Any signal you hear on the inverting input means the opamp can't do this properly, and is likely slewing between the rails attempting to do so.

[PBE6]

Yaargh!! Yes, looks like not using Vref as the virtual ground was (at least partly) the issue. Funny, I actually had it wired up that way originally but it wasn't working so I switched to regular ground because I could see the waveform blending. And that was dead wrong! Stompbox, thou art a cruel mistresses..

Of course, I managed to screw something else up now, too. I checked the 47k feedback resistor again and thought it was too small (max Vout was about 350 mV again for some reason??) so I bumped it up to 120k. It's definitely loud now (max Vout about 1.5 V), but crazy noisy with no input - not surprising. I'll have to scale it back tomorrow.

Even stranger, the blend goes mostly (but not totally!!) clean around the 3/4 mark, but then gets louder again and strangely dirty past that. On the scope it looks like rats have been chewing on the sine wave! Very strange, they're supposed to be relegated to another distortion pedal   kidding aside, what the heck would cause that to happen?!?

Time for bed, hopefully some rest will clarify things.

[PBE6]

Cracked this open again to see what the signals sound like and there is definitely interaction between the gain/sustain knob and the supposedly "clean" signal. That can't be normal =\

[Keppy]

I wonder if you're Vref isn't filtered enough. You could try putting the blend pot wiper to ground again, but cap couple the inverting input to the 15k resistors feeding it.

Also, by my math (done in my head late at night, admittedly) the 47k resistor should provide just under unity gain. Did you check the values of all four 15k resistors, plus the 10k pot?

That's all I got. If it doesn't help, time for voltages and a schematic with complete values, unless someone else wants to chime in.

[akkar]

If I were doing this, I'd use a high(ish) impedance inverter for buffering the clean signal, an inverter for taking off the dirty signal, and the circuit in the upper left hand box in http://www.geofex.com/Article_Folders/panner.pdf. By pre-inverting the signals, the final panner/mixer stage re-inverts them to normal. The panning/mixing is done with currents, and the two sides do not interact at all. It's a special-case circuit, but it's really handy when you want to move/blend/mix two inputs only.

Sorry if I resume this old thread... If I use for eg a TL074 for the inverters (creating a double power supply from the 9volts) should I connect the tap of the pot in the link to the 4.5V, instead of ground?

[antonis]

For a dual supply (+/- 9V), implement RG's first page configuration..

For a double supply (+18V), use +9V as V_bias..
(with adequate DC decoupling, of course..)

[akkar]

For a dual supply (+/- 9V), implement RG's first page configuration..

For a double supply (+18V), use +9V as V_bias..
(with adequate DC decoupling, of course..)

Thanks for the answer, however I don't understand what you mean by "first page configuration"

[antonis]

'Cause you quoted RG's http://www.geofex.com/Article_Folders/panner.pdf, where 1^st page op-amps are powered from dual (symmetrical) supply..
(not clearly shown but guessed from non-inverting inputs bias..)