GGG's volume circuit on Scrambler seems to distort

[mark2]

Ampeg scrambler - octave fuzz.



From this document, via GGG's "Updated Circuit With Volume Control"

I get what I believe to be the intended result (as determined from a video) on the output of the balance pot (R17).  The balance does a good job of blending between the textured signal and the clean'ish signal.

The clean side is a bit distorted though.  Not sure if that's intended or not, but it's fine as far as I'm concerned for the purposes of this pedal.

Immediately on the other side of C8 (0.1uF) the sound loses a lot of bass response and is far more distorted (or if the distortion is pre-existing, it becomes much more prevalent to my ears). I tried 3 different caps, 2 different types and the effect was the same. The result is that the "Balance" knob no longer seems to have any effect on distortion and the signal is always buzzy.

Is this expected for a 0.1uF cap? Eg high pass filter. If so, does the volume circuit (everything after the balance pot) supposed to be doing something to compensate?

I'm getting no boost in volume, only the ability to reduce it - which is completely unnecessary since the circuit already has an expected volume drop.  Because of this I'm tempted to just cut out the entire volume circuit and call it "done," but this problem is really bothering me and making me curious.

[Mark Hammer]

The stock Scrambler is not really capable of providing a clean sound at any setting, which may have something to do with use of a 5306 Darlington as its input "buffer".

I've been sticking a modified Distortion+ circuit ahead of the Scrambler, to make it a 4-knob box, for a number of years.  Part of the logic is that I could turn the output of the Dist+ portion down to deliver a "gentler" input to the Scrambler section, and also modify the output level.  Cranking the gain of the Dist+ provides for different qualities of Scrambler octaving, as well as different durations.  The pseudo-compression of the Dist+, when cranked, remains at the level the octaving "prefers" for a longer period of any held note.  Turning the gain down, and using modest volume levels gets you subtler octaving that occurs at the start of the note but gets out of the way quickly.  Lots more tonal variety.

That said, I wonder if sticking a more "normal" bipolar in the Q1 position might yield better cleans and subtler Scrambling.  I may just try that later today.

[Mark Hammer]

Well, it's a little later in the day.  Found a couple of Scrambler boards I had etched and stuffed, but never installed.  I installed a 2N3904 in place of the 5306.  It would clean up nicely, but there was little in the way of octave or fuzz.  Not enough push from that front stage.  Just replaced the 2N3904 with an MPSA18.  So, a change from an hfe of around 250 to around 890.  That was the trick.  A decent clean tone when the balance pot was set to full clean, and just enough fuzz and octaving as the balance control was rotated  past 10:30 from clean to dirty.  haven't tried out the Dist+ front end yet, but I suspect it will sound decent.

So, the moral is "Pull the 5306 and stick in a normal higher-gain bipolar transistor, like a 2N5088/89 or MPSA18".  On my unit, at least, nit yields an improvement in clean/dirty balance and overall clean capability.  The caveat is that there is no stompswitch connected so I haven't tried out the bypass/effect comparison.

[mark2]

Thanks for the info.

A key thing to note, though, in case you missed it: The "clean'ish" sound I can get out of the balance knob is pretty clean (not perfect but good enough).

The problem is that C8 seems to ruin the sound completely. This is part of GGG's volume circuit.

As I mentioned, I was thinking of simply removing it, but then I won't be able to use a really awesome enclosure I made for it (which has the volume knob).  So in short to anyone who might know:

A.  Is the volume circuit junk? And should just be removed?  (vs a problem I might have introduced).
B. If so, what about a simpler passive volume, e.g. just a simple pot.  Not necessarily because it's useful, but to fill a "Volume" spot on the enclosure.

[PRR]

Why blame C8? Maybe Q5 isn't happy?

What are the voltages around Q5? That would be a first-check for troubles.

[mark2]

Why blame C8? Maybe Q5 isn't happy?

I can check voltages and post later, but as I mentioned: it sounds great off the balance pot (directly on one side of C8), but is distorted, loss of bass, and tinny on the other side of C8.

[mark2]

With an 8.9V battery, Q5 is showing 0.3, 0.93, and 0.93 V.

Would this affect the sound on C8 though? Before it even reaches Q5

[idy]

Those voltages look messed up. Yes, it super common for a bad (or mis-biased or badly soldered or reveresed or gounded) component to affect the sound connected to it..."before" it. If two components share a node, and the sound is bad on that node, it could be either one!

Lift the the end of the cap close to Q5 and test the sound coming out of it when it is not connected to the Q. It the sound is still bad, then it wold be the capacitor. Not so likely.

Why is the Q not getting a higher voltage on its C? Should be a nice positiive voltage up top and then B and E separated by .6v. I think...

[Mark Hammer]

Well, it's a little later in the day.  Found a couple of Scrambler boards I had etched and stuffed, but never installed.  I installed a 2N3904 in place of the 5306.  It would clean up nicely, but there was little in the way of octave or fuzz.  Not enough push from that front stage.  Just replaced the 2N3904 with an MPSA18.  So, a change from an hfe of around 250 to around 890.  That was the trick.  A decent clean tone when the balance pot was set to full clean, and just enough fuzz and octaving as the balance control was rotated  past 10:30 from clean to dirty.  haven't tried out the Dist+ front end yet, but I suspect it will sound decent.

So, the moral is "Pull the 5306 and stick in a normal higher-gain bipolar transistor, like a 2N5088/89 or MPSA18".  On my unit, at least, nit yields an improvement in clean/dirty balance and overall clean capability.  The caveat is that there is no stompswitch connected so I haven't tried out the bypass/effect comparison.

Turns out I had two boards etched and stuffed, so I tried a 2N5089 in the other one.  Good enough, but a different feel.  Couldn't tell you what the hfe was on it, because it was already installed, but I think it was in the 500-600 zone.  But a similarly wide range of available tones.

I think a simple overdrive as front end is like the jam to the peanut butter of the Scrambler; a match.

[PRR]

> Q5 is showing 0.3, 0.93, and 0.93 V.
> Would this affect the sound on C8 though? Before it even reaches Q5

Q5 is somehow shorted-out, non-functional.

The simple interpretation is that it has zero gain, and is acting-like about 390 Ohms into the base. With a 0.1uFd cap most of the audio band is severely attenuated and generally distorted.

Fix what is wrong with Q5 and its bits.

[mark2]

Thanks for the tips and info. You were both right and it was a problem with the voltage on Q5.  There was a tiny solder bridge.

It works now!

It is *way* too quiet to get even close to unity volume (even with the blend fully on or off), but all the voltages are in line with the instruction sheet.

I've seen other reports of this, but also some people saying they *can* get unity.. so I'll do a bit more investigation and look for problems. Failing that I"ll probably box it up and call it done.  It definitely works great aside from the volume.

Thanks again!

[PRR]

The clean side should have lots of gain, over 10.

The scramble side is not easy to grok because gain depends on signal level. Mark is always right and if he says typical guitar wants a boost to scramble properly then boost it. But the clean side is a "unity" buffer plus an LPB so should boost significantly.

[mark2]

Yeah, so something is clearly wrong if I can't get unity on the clean side. Thanks.

[idy]

The original did not have Q5, the clean side is just the output of the input buffer. The fuzz side has a resistor to ground to cut the output to be more or less unity, easy to blend into the clean sound.

The GGG Volume pot mod adds a LPB after the blend pot.

I've made a few of these based on the original circuit (no booster), and they give more or less unity out. The pedal was not designed to be a booster, just a "color adder."

If your pedal is still not unity even with a booster attached to it, you still have some poking around to do. Is the sound louder inside the three transistor circuit that makes the fuzz? Aren't you getting unity out of the buffer Q1? You need a signal probe and a signal source like an ipod you can use to give it some input while your hands are free.

[mark2]

Thank you both again.  I used an audio probe to find where the "clean" volume dropped, then inspected around and found another tricky, tiny solder bridge.

I have no doubt this pedal is now working 100%!

I spent some time playing it, though, and also have no doubt this pedal is 100% on its way out the door. I can't sell or trade it fast enough.   Hopefully i can find somebody else who'll appreciate its funkiness, and use the proceeds to fund another build.