Modernizing a "SPEKTR-1" Wah-Fuzz-Tremolo

[Paul Marossy]

My fascination with these very interesting/unusual Soviet-era guitar pedals continues. I got a good deal on a partially functioning "SPEKTR-1". I don't have it in hand yet, but while I'm waiting for it to arrive from halfway around the world I've been analyzing it and trying to understand some of its aspects. I'm not a fuzz guy... not in the least, but it has a really nasty sounding fuzz that I'd like to observe on the o'scope. I'm mostly interested in the wah function for now but I want to get it fully functional. Should be able to achieve that since there's no unobtainable parts in it. Converting the in/out jacks to 1/4" type won't be easy but it's do-able.

There's a few things in particular that I am curious about:

1 -- Can I combine the two power supplies into a single DC jack? It appears possible from some videos I've seen
      where it was converted from the 5-pin DIN jack to a more conventional modern DC jack. Looks to me like
      the PNP transistors are in a negative ground scheme. (?)

2 -- "V19" is allegedly a KS168A which appears to be a 6.8V zener diode. Does that sound right? Is that partly
      how they create the fuzz, by starving the transistors with a low voltage? Allegedly some examples may not
      have this diode but I have no way to verify that or why that may be the case.

3 -- The arrangement of V10-V13 reminds me of a full wave rectifier... what exactly is that doing? I've never
      seen anything like it. As far as I can tell from pictures I have found, they are 9V zeners. Why use
      zeners here?

Schematic can be found here: http://www.diyguitarist.net/Images/SPEKTR-1_Schematic_(Translated).jpg
Picture of PCB can be found here: http://www.diyguitarist.net/Images/SPEKTR-1_PCB.jpg

I'm currently trying to model the fuzz section in LTSpice but I'm not sure exactly what waveform I should be seeing on the output. What comes out is very sensitive to what exactly the power supply is.

[ElectricDruid]

What a fantastic old thing! It looks great, and those old soviet parts definitely have *bags* of mojo, lol.

1 -- Can I combine the two power supplies into a single DC jack? It appears possible from some videos I've seen
      where it was converted from the 5-pin DIN jack to a more conventional modern DC jack. Looks to me like
      the PNP transistors are in a negative ground scheme. (?)

It looks possible to me. Since one power supply is used purely for the LFO, I'd guess it's been done like that to prevent ticking getting into the audio.

2 -- "V19" is allegedly a KS168A which appears to be a 6.8V zener diode. Does that sound right? Is that partly
      how they create the fuzz, by starving the transistors with a low voltage? Allegedly some examples may not
      have this diode but I have no way to verify that or why that may be the case.

Sorry, no idea on this one.

3 -- The arrangement of V10-V13 reminds me of a full wave rectifier... what exactly is that doing? I've never
      seen anything like it. As far as I can tell from pictures I have found, they are 9V zeners. Why use
      zeners here?

Yeah, very fullwave-rectifier-ish. I would guess that's what it's doing. FWR is a perfectly good way to make a fuzz, and gives you a bit of that sixties octave sound.
What it actually reminds me of most is the envelope follower in the Dynacomp. That uses the same design of a phase-splitter transistor and two parallel paths. In that case, the signals are recombined by transistors, but here it looks like they're using diodes for the same job. I'd guess there's four of them because they need to be biased on to work - otherwise nothing would get through. But I'm only guessing.

This is the sort of thing for which I'd be quite cautious using LTspice. The circuit will depend on using those zeners in a biased-just-enough-on mode so that increasing signal turn them on for a half-cycle, and the other signal turns the other diodes on for the other half-cycle. I wouldn't trust LTspice's zener model to have that level of detail. I might be wrong.

Good luck, it looks like a great project.

[Paul Marossy]

3 -- The arrangement of V10-V13 reminds me of a full wave rectifier... what exactly is that doing? I've never
      seen anything like it. As far as I can tell from pictures I have found, they are 9V zeners. Why use
      zeners here?

Yeah, very fullwave-rectifier-ish. I would guess that's what it's doing. FWR is a perfectly good way to make a fuzz, and gives you a bit of that sixties octave sound.
What it actually reminds me of most is the envelope follower in the Dynacomp. That uses the same design of a phase-splitter transistor and two parallel paths. In that case, the signals are recombined by transistors, but here it looks like they're using diodes for the same job. I'd guess there's four of them because they need to be biased on to work - otherwise nothing would get through. But I'm only guessing.

This is the sort of thing for which I'd be quite cautious using LTspice. The circuit will depend on using those zeners in a biased-just-enough-on mode so that increasing signal turn them on for a half-cycle, and the other signal turns the other diodes on for the other half-cycle. I wouldn't trust LTspice's zener model to have that level of detail. I might be wrong.

Good luck, it looks like a great project.

Yeah, I don't count on LTSpice to nail stuff every time, but it does get pretty close a lot of the time. I do have a pretty good zener diode model.

Anyway, I have my doubts about V19 being a zener, or at least not a 6.8V. I don't see how the 9V zener arrangement would work with only 6.8V getting to them ("+U1" source). Seems to me like it's just a polarity protection diode. I'd have to see the original factory schematic to know for sure, and even then it could have a wrong part number shown. One thing I'll have to investigate further when I have it in hand.

[Paul Marossy]

So it only took nine days for it to arrive from half way around the world. I've been working on this thing the last few days. It was supposed to have wah functioning but no fuzz, "only silence". It was DOA on arrival.

The insides were a complete hack job - someone took apart the nice neat factory wiring harness so the wires were all loose, they disconnected things, connected wires in places they shouldn't have been, did various other things that are inexplicable and the wah pot arrangement had a strange hack job done to it as well, which was NOT how it came from the factory. Had to fix that too. Overall it's just very difficult to work on because the copper on the PCB is so delicate. 50% of the electrolytic caps were bad, so I replaced them all. In that process, about half of the 35-40 of the tiny wires inside broke off from where they were supposed to be so that added to the frustration as well. I re-did the wiring harness best I could. It didn't turn out too bad all things considered.

I have it now working except for the vibrato section. I am a bit baffled by it. Signal goes in but only about 18-20mV gets out, so when you switch it on the sound almost goes away completely. The LFO appears to be running at about 8 hz but you can't hear it... probably because the output is nearly non-existent. All of the transistors have some kind of voltage on them but I don't know what they are supposed to be. Overall, in the LFO circuit the voltages all seem to be low. Not sure what the deal is with that. Maybe someone can give me an idea of what I should be seeing for voltages? I did measure them all and write it down. Next I will check on the continuity between all of the components, maybe there is a broken track somewhere that visual inspection did not detect.

Verified that the schematic appears to be correct. Most of the circuit does run on on 6.8V like I thought, except for that "vibrato".

I made a webpage about this "SPEKTR-1" this morning and have PCB layout & wiring diagram linked there too. http://www.diyguitarist.net/DIYStompboxes/SPEKTR-1.htm

Now I just need to figure out why the "vibrato" section is not working. I'm wondering if it's a bad transistor or it if it's something loading down the power supply to where there are not functioning as intended. Not sure how that could happen though.

EDIT: OK I found two things were connected together that should not have been. Now I can hear tremolo. Still low output, which was much lower before I discovered a 820 ohm resistor had drifted up to over 52K! I've seen that happen on old carbon comps but first time I've seen it in an old Soviet guitar pedal. Now I just need to figure out why output level is still low (could that other circuit trim pots are adjusted to have too much output in comparison). Main issue now is that it is THUMPY about 8 times a second, and louder than my test signal set up. I guess I should test with a guitar & amp now and see what that does.