adding a feedback control to a Shin Ei Fuzz Companion

[moid]

Hello everyone

Yesterday I built a modified Shin Ei Fuzz Companion (see below)


And it works gloriously. I was even smart enough to test the pedal before I boxed it (see, I'm learning!) but I have added two extra controls that are not on the above layout - one is a simple starve control that works perfectly (10K lin from the power jack to the power in of the board if anyone wants to try it - really adds sputtery nastiness to this pedal). The other control was what I thought (in my confused / befuddled mind) would be some sort of feedback loop - basically take a cable from the audio out of the pedal to lug 3 of a potentiometer (I used another 10K Lin) and then run a cable from lug 2 of the potentiometer to the audio in of the circuit which in my mind would allow me to add more of the circuit back into itself as I turned up the potentiometer (like a passive feedback loop pedal does)... but it didn't work... the first time I thought it worked when I hadn't boxed it but it turned out that I hadn't connected all the cables and I was so impressed that I built a pedal that just worked first time I didn't realise one of the knobs was doing nothing. So I drilled the enclosure to have 5 knobs and put it all together, discovered the unsoldered connection and being slightly smart decided to try using insulated crocodile clips to re connect everything before i soldered anything and then all I got in the end was a lo fi bass cut and volume drop! Not that I have anything against lo fi, but it's not what I wanted, and even with the potentiometer set to off, the effect is still present (plus random hiss noise).

Presumably my brilliant idea won't work because of some arcane electronic law... so I could drop the idea of that knob entirely (the rest of the pedal is fine after all) but I've now got an unsightly hole in the enclosure that needs filling. Does anyone have a suggestion for something to stick there that wouldn't be large (this whole mess is squeezed into a very cramped space) and might modify the sound but have the option to go from not modifying it at all to adding modification? I have tested a SPDT switch and attached it instead to create a lo-fi on/off function instead of a knob, but the lo fi effect isn't that endearing (it isn't wildy lo fi and just cuts bass and makes everything sound flat) so I'm not too keen to go in that direction... I don't need tone knob because the pedal already has one (labelled Scoop) and I'm out of ideas. Any suggestions would be gratefully received. Thanks!

[Ben79]

There's a mod by Mark Hammer to alter the 'body' of the tone.

Another thing I've done successfully is incorporate a "body" control. You see the 100k47k pair in series, between V+ and the collector of Q2? There is a .047uf cap in parallel with the 100k resistor. By varying how much of the entire 147k resistance that cap straddles, you can vary the bottom/body of the resulting sound.

One possible way to do this would be to connect an 1k resistor between V+ and one outside lug of a 100k pot, and a 33k resistor between the other outside lug and the collector of Q2. You now have 151k of resistance between V+ and the collector. Given that pots are usually a little lower than their nominal value, that probably puits you spot on 147k. Now, connect the .047uf cap between the wiper of that pot and V+. The total V+-to-collector resistance will remain fixed, but the resistance in parallel with the cap will be varied from 18k to 118k; from thinner-sounding than stock, to something almost suitable for a bass. Try it and see if you like it.

[crochambeau]

Can you draw out the feedback loop you employed?

I see three BJTs there, for an ADDITIVE effect you probably want to work in even numbers (unless you've an emitter follower in there, sorry, my mind hates trying to wrap around strip layouts, so...), I would say plop your feedback leg into (or from) the second transistor, use a DC blocking capacitor to keep the sections from beating on each other's optimal DC values, 10K might be a touch low, but that's easy to remedy.

[moid]

There's a mod by Mark Hammer to alter the 'body' of the tone.

Another thing I've done successfully is incorporate a "body" control. You see the 100k47k pair in series, between V+ and the collector of Q2? There is a .047uf cap in parallel with the 100k resistor. By varying how much of the entire 147k resistance that cap straddles, you can vary the bottom/body of the resulting sound.

One possible way to do this would be to connect an 1k resistor between V+ and one outside lug of a 100k pot, and a 33k resistor between the other outside lug and the collector of Q2. You now have 151k of resistance between V+ and the collector. Given that pots are usually a little lower than their nominal value, that probably puits you spot on 147k. Now, connect the .047uf cap between the wiper of that pot and V+. The total V+-to-collector resistance will remain fixed, but the resistance in parallel with the cap will be varied from 18k to 118k; from thinner-sounding than stock, to something almost suitable for a bass. Try it and see if you like it.

Thanks Ben I didn't know that. This modified version of the Shin Ei fuzz has much more bass than the original and the Scoop knob controls whether you dial in the original higher sound of the original (I built an original shin ei first, then decided it needed volume and bass) or a much heavier bass effect (which sounds excellent). Work has of course gone crazy so I can't now look at this until the weekend

[moid]

Can you draw out the feedback loop you employed?

I see three BJTs there, for an ADDITIVE effect you probably want to work in even numbers (unless you've an emitter follower in there, sorry, my mind hates trying to wrap around strip layouts, so...), I would say plop your feedback leg into (or from) the second transistor, use a DC blocking capacitor to keep the sections from beating on each other's optimal DC values, 10K might be a touch low, but that's easy to remedy.

Can you draw out the feedback loop you employed?

I see three BJTs there, for an ADDITIVE effect you probably want to work in even numbers (unless you've an emitter follower in there, sorry, my mind hates trying to wrap around strip layouts, so...), I would say plop your feedback leg into (or from) the second transistor, use a DC blocking capacitor to keep the sections from beating on each other's optimal DC values, 10K might be a touch low, but that's easy to remedy.

Thanks for replying! I will draw a diagram soon (might have to be the weekend though sorry, have a huge deadline for Friday) to show my actual layout. From what you say about additive is this related to a signal being in phase or out of phase? I found a thread on DIYstompboxes I think where someone had a similar issue and the consensus of opinion was that adding an out of phase signal to an in phase signal would cancel out frequencies and volume in the way mine is.

Hmmm on re testing with a pair of insulated crocodile clips attached to a cable I am just getting a selection of high pitch beeps / low pitch oscillation (quite techno actually! - it makes for a interesting tone generator noise box, so I'll keep this idea for a future pedal perhaps - or maybe if I can't make a feedback loop work I'll install an SPDT switch in the empty hole and use it to switch from effects pedal to noise generator if required) when I connect the output to the input - I think the bass cut / lo fi effect happened when tried sending the audio out to other pots / resistors and capacitors in the circuit. I should really do this when I'm awake and not last thing at night, but that's usually the only time I have Sorry for any confusion I caused.

One thing I just remembered that may have caused the phase issue (if that's what it is) - I switched Q1 in this vero design to a 2n4401 - I put 2n3904s in Q2 and Q3 because the shin ei I originally built had those in it's vero diagram.

If I connect the output audio to lugs 1 or 3 of the Fuzz knob I get the lo fi sounds - the same goes for many of the capacitors and resistors on the board.

Edit: wow after reading what you said about the size of the pot and the specific transistor I discovered that a combination of 100K lin pot with lugs 3 connected to the audio in jack and lug two connected to the Collector of Q3 in the vero diagram results in either normal pedal sound at one end of the range, damaged weirdness half way through the sweep and odd oscillation/ techno pulses or muffled tones at the other end. In short, musical bliss! Yay, thanks once again for solving my musical issues! It's not what I intended, but it sounds great so it's staying!

[Ben79]

Sounds interesting. Would love to hear this ….

[crochambeau]

hahaha, I was going to recommend putting a little resistance into the mix after I read about your beeps and bloops, but it sounds like you're way ahead of me.

[moid]

Sounds interesting. Would love to hear this ….

OK I'll try to record something this weekend, I'll ask my son to play something while I adjust the settings - that'll be better than you listening to me mangle a guitar!... although I seem to have lost the oscillating bleeps somewhere which is a little depressing... the circuit is kind of held together by crocodile clips and loose wire at present, so I'll try to see if I can get them to come back - the more distorted and muffled and crumbly sounds are still present, plus some constant pitches that are a bit irritating on some settings... I think this might be a work in progress...

:rock: hahaha, I was going to recommend putting a little resistance into the mix after I read about your beeps and bloops, but it sounds like you're way ahead of me.

Actually I thought that's what you were talking about when you said that 10K might not be big enough? I did try 100K lin, 250K lin, 500K lin and 1M lin pots... I'm sure one of them gave me techno bleeps but it doesn't seem to be the 100K, so maybe I'll swap pots around again to go chasing those beeps... One issue my son pointed out that I didn't notice before is that when I add more of this feedback into the audio the volume definitely drops, and by quite a bit... so I'm wondering if I stuck an LPB1 circuit with a trimmer on a separate veroboard and put that between the Connector of Q3 and Lug 3 of the Feedback pot with the trimmer set to add a boost to the volume would that possibly counteract the drop in volume overall? Or, if this is acting subtractively, would it make the result worse and the audio would get even quieter? Or would you have any better suggestion than my botch fix technique? Thanks

[crochambeau]

Actually I thought that's what you were talking about when you said that 10K might not be big enough?

Yeah, it was, but I'm used to having my hints lost to the nuances of the internet..

Yeah, plant a single transistor circuit in there and see what happens. IF you have the room and a spare transistor, it is wholly possible to build a section that will allow you to select the polarity of the signal. It's not really suited to the outside world, but tucked inside a circuit where you're in command of what sorts of loads and signals it sees... (by which I mean to say it's not an ideal circuit to stick by it's lonesome in an enclosure and expect the world..)



Those values may need to be ignored..

[moid]

Thanks for the circuit diagram. I'll need to draw it out on vero (I know you're not keen on vero but if I put a diagram up would you mind looking at it before I build it please?). If I can squeeze it into less than 10 rows of height I can hopefully get it into the box

Do I connect the left side of the circuit with Lug 3 of my feedback pot and the right side of the circuit to the audio in?

or

Do I connect the left side of the circuit with Lug 2 of my feedback pot and the right side of the circuit to the collector of Q3?


Is the switch an SPDT? It would switch between one polarity and the other?

I don't have any BC547 transistors. Could I use any of the following transistors? (I saw a page on DIYstompboxes suggesting 2n5088 instead of the BC547?)

I have the following transistors at the moment:
BS170
BC108
BC107
BC182L
MPSA18
2n2907A
2n2222A
2n5088
2n3904
2n4401
2n5457

Thanks again

[crochambeau]

It's labelled BC547 because that was the NPN transistor profile in my library, I use 2N3904 on it, any general purpose NPN should work.

Signal comes in at C40, a SPDT switch does work, center lug being output, C41 is inverted & C42 is not. Which way you orient it is up to you, so long as you keep the ins/outs connected so the feedback is a path it really shouldn't matter which comes first (if I understand your questions)

Happy to have a look at the vero, but I absolutely cannot guarantee my assessment will be accurate.

[moid]

Thanks I'll use 2N3904 then, I've got plenty of those. I assume that the capacitors are electrolytic if you say one is inverted. In which case does the solid red rectangle side of each capacitor refer to the strip on an electrolytic? I've started drawing this and wasn't sure about the orientation.

[crochambeau]

Any leg of the electrolytic aiming toward the transistor is generally positive. The only exception will be interstage (I'm guessing you're pulling signal from the collector somewhere - THAT voltage will probably be higher than the base, so the cap will prefer the positive leg connected in that direction.

Also, that is just a screenshot of a subsection in one of my circuits, so the values are far from critical. For a feedback loop you could probably get away with .01 or something, anything, it's a noise leg, go crazy.

[moid]

OK I learned another new thing about transistors and capacitors - thanks! Would you mind looking at the vero layout below please? I've tried to label everything - I know you said to experiment with values, but I've put yours in for starters just to check the circuit is OK. There is a cut under each of the capacitors that are connected to the Collector and Emitter of the transistor. Thanks!

[crochambeau]

You want the center leg of the switch connecting to the pot (I know my schematic has those pins in an odd order; that's for a clean drawing, NOT a layout), otherwise it looks right to me!