Ludwig Phase II Clone Debugging and Miscellany

[R.G.]

I got a few minutes to look at this. The "echo" transistors might be a problem. This can be cured by putting a 17V zener from the junction of R64/R39/R40 to ground. That prevents the voltage from ever going over the design voltage of about half of the +35V supply, so it would correct Keppy's supposition about the size of the collector swings on Q7/Q8. I still need to think about why the original didn't work right, but this hack would fix the problem.

I'll go look at the triggering from the PUT. It would be unusual that the AC coupled output spike of the PUT would not trigger the flipflop through the cap that couples it, but hey, I've made a career out of being wrong and then correcting it.   

[R.G.]

OK, 17V zener cures the funny problem with the animation switch. I did not think completely through how Q7 and Q8 work.

On the triggering from a PUT:

The flipflop triggers when some signal source puts a negative-going spike on the combined collector supplies at the junction of the two collector resistors and the resistor to +35V. What is critical here is that the trigger signal have enough oompf to make the transistors flip state by dragging both collectors down.

PUTs and UJTs differ a little in exactly how they work, but for almost all applications, they're equivalent. I think that the values of R7 and the 0.01uF cap that couples the signal into the joined collectors may also need tinkering.

Keppy, if you'll decrease R7 from 220R to something smaller, maybe 100R, and maybe increase the cap from 0.01 to 0.022, things may start working. If you're almost there, the oscillator may half-time, only flipping every two oscillator strokes.

Oh, yeah - a couple of sharp eyes noticed that I transposed the two inputs in the wiring diagram. I'll go update the docos. Ideally, if someone can verify the fixes here, I'll roll all that into one update.

[Govmnt_Lacky]

Got a busy day today, but I'll have time to check it out tonight.

Sit tight bro... I'll run through your requests, and report back.

Dino is on the case! 

I'm gonna sit back and see what you find out. Gotta work today so I wouldn't be able to get to it until tomorrow evening anyways. 

[Keppy]

Thanks for the help everyone! It means a lot that everyone is still so quick to help with this.

The "echo" transistors might be a problem. This can be cured by putting a 17V zener from the junction of R64/R39/R40 to ground.

That's a good idea, simpler than what I came up with. Since Q3/Q4 only output about 13.5v, should I use a 12V zener instead, or maybe two 6.8V in series?

Keppy, if you'll decrease R7 from 220R to something smaller, maybe 100R, and maybe increase the cap from 0.01 to 0.022, things may start working. If you're almost there, the oscillator may half-time, only flipping every two oscillator strokes.

I'll try it out. I'm reluctant to do too much desoldering for fear of damaging the pads (etched boards), but I'll poke a resistor and cap in parallel with what's there and see if it helps.

Keppy, your builds are doing exactly what mine was (as I recall).  While this sounds frustrating to you, I'm really excited that you may have uncovered why mine isn't working.  I put this thing aside after spending close to a month debugging it almost exactly two years ago...I would love to get my clone running!

We'll get it running.

[R.G.]

The "echo" transistors might be a problem. This can be cured by putting a 17V zener from the junction of R64/R39/R40 to ground.

That's a good idea, simpler than what I came up with. Since Q3/Q4 only output about 13.5v, should I use a 12V zener instead, or maybe two 6.8V in series?

Whatever works. I was just guessing about the voltage and based the "half of 35V" on the fact that there's one common 4.7K and one 4.7K collector resistor on each flipflop transistor. If the signal's 13.5, stick in a 13.5V (kind of...  ) zener. I suspect the actual value isn't all that critical.

As a side note, the base-emitter of most silicon NPNs is about 6-7V, so two of them in series might be about right if you didn't have a 13-15V zener laying around. Season to taste. 

Keppy, if you'll decrease R7 from 220R to something smaller, maybe 100R, and maybe increase the cap from 0.01 to 0.022, things may start working. If you're almost there, the oscillator may half-time, only flipping every two oscillator strokes.

I'll try it out. I'm reluctant to do too much desoldering for fear of damaging the pads (etched boards), but I'll poke a resistor and cap in parallel with what's there and see if it helps.

Good point. Solder a 220 on top of that existing one, and/or try paralleling another cap on the coupling cap.

[digi2t]

Sorry for the all caps. Not yelling, just want to contrast the answers from the questions.

Auditory test:
Animation on, Intens. FFM and Intens. Fuzz Rpt. all the way up, rocker centered, Counter mode. Start with the Animation toggle set to Both.

1) Do you hear both the Fuzz Repeat and the automatic filter sweeps?
YES

2) When you move the toggle to Fuzz Rpt. does the filter sweep stop? Do the Fuzz Repeats just continue, or do they get more intense?
FUZZ REPEATS ARE MORE INTENSE, AND FILTER SWEEPS ARE MUCH LESS. FUZZ REPEAT INTENSITY IS ALSO SOMEWHAT DEPENDANT ON PEDAL POSITION, AND MANIPULATING THE FFM INTENSITY KNOB MAKES ONLY A TINY DIFFERENCE IN THE OVERALL SOUND.

3) When you move the toggle to FFM, do the Fuzz Repeats stop? Does the filter sweep change at all from the center toggle position?
FUZZ REPEATS ARE MUCH LESS, AND THE FILTER SWEEP IS MUCH GREATER. FILTER SWEEP INTENSITY IS SOMEWHAT DEPENDANT ON PEDAL POSITION, AND MANIPULATING THE FUZZ REPEATS KNOBS DOES NOTHING.

Meter test:
Animation on, rate all the way down, R4 (internal speed trimmer) all the way down.
I'VE USED THE MIN/MAX SETTING ON MY DMM, THE READINGS ARE AS SUCH. I'VE ROUNDED THE READINGS.

1) Measure the voltage on Q7 collector with the toggle switch in all three positions
FUZZ RPTS. = 21V/29V
BOTH = 0V/12V
FFM = 0V/23V

2) Repeat for Q8
FUZZ RPTS. = 0V/32V
BOTH = 0V/17V
FFM = 18V/28V

Hope this helps.

[R.G.]

I believe that verifies the supposition that Keppy brought up about Q7 and Q8, and that the simulation run this morning thought was right.

Do you have anything that could be a 12V to 18V zener? I think that would fix it back to the original content, and keep the cheap center-off switch.

[Govmnt_Lacky]

Fairchild 1N5247

17V Zener 0.5W

Would a 0.5W Zener be OK? Or... would the 1W be better?

EDIT: Looks like 1W is only available as 16V or 18V 

[R.G.]

Hmmm... lessee here.

35V minus 17V is 18V. That's across a 4.7K resistor, so the current is

I = 18/4700 = 0.0038A, 3.8ma. The power in the resistor is 18 x 0.0038 = 68.9mW, call it 70mW.

The power in the zener is 17V times at most that same current (if both Q7 and Q8 are off, which only happens for microseconds) so it's go nearly the same power, about 70mW.  I'd say half watt is fine, even overkill, except that smaller power zeners are probably more expensive.

Note that using a zener between 12V and 18V doesn't change this all that much.

[Keppy]

I have plenty of 12v and 6.8v zeners. I can give this a try later.

[Govmnt_Lacky]

This can be cured by putting a 17V zener from the junction of R64/R39/R40 to ground. That prevents the voltage from ever going over the design voltage of about half of the +35V supply, so it would correct Keppy's supposition about the size of the collector swings on Q7/Q8.

So... just need to find a convenient GND location and... if there is one nearby... tack in the zener and all should be OK.... RIGHT? 

[digi2t]

Just out of curiousity...

Would the zener be required only for the 6027, or for the 2646 as well? Do the readings I supplied show proper operation? I mean, to my ears, everything sounds OK.

[Keppy]

So... just need to find a convenient GND location and... if there is one nearby... tack in the zener and all should be OK.... RIGHT? 

Yes. I just verified this with a 12v zener. I tacked it between the bottom of R64 and the top of R29.

Just out of curiousity...

Would the zener be required only for the 6027, or for the 2646 as well? Do the readings I supplied show proper operation? I mean, to my ears, everything sounds OK.

Both. This affects neither the oscillator or the flipflop, just the "echo" transistors Q7 & Q8, which transfer the triggers of the flipflop to the FFM and Fuzz Repeat circuits. Your readings show incorrect operation, electrically if not audibly. Here are the correct readings:

Q7:
FUZZ RPTS. = 13.5V/13.5V
BOTH = 0V/13.5V
FFM = 0V/13.5V

Q8:
FUZZ RPTS. = 0V/13.5V
BOTH = 0V/13.5V
FFM = 13.5V/13.5V

All those 13.5V readings are probably okay anywhere 12-17V. The reason for these "ideal" voltages is Q7 & Q8 are supposed to read the same as Q3 & Q4 (0V/13.5V at all times), but with the ability for the toggle to disable one or the other completely. If yours sounds fine, no worries, but it's just one tack-on diode, and I know you're the curious type....


I tried paralleling the cap and resistor with no success. I have a couple UJTs on order, so I'm confident I'll get these working even if we don't work out the PUT issue. It would still be nice to work it out, though. I verified that the flipflop can trigger the filters in my build, but the oscillator still won't trigger the flipflop.

[Ry]

I'll have to dig through my zener drawer and see what I have, I'm sure I have something that will work.  This is exciting, I'll have to carve out some time to dig into this!

[digi2t]

Guess I'm digging through the zeners tomorrow as well. 

[Keppy]

I tried paralleling the cap and resistor with no success.

I forgot to mention that with the resistor paralleled, the tick did get louder, but still did not suffice as a trigger.

[Govmnt_Lacky]

So I am assuming that the zener solved the switching issue with Q7/Q8 

If it did... does it matter what threshold zener is used? (i.e. 12V, 16V, 17V, etc.)

[Keppy]

So I am assuming that the zener solved the switching issue with Q7/Q8 

It would appear so.

If it did... does it matter what threshold zener is used? (i.e. 12V, 16V, 17V, etc.)

I'd say as close to 13.5V as you can get, because that's what the flipflop puts out. No more than 17V (1/2 supply voltage) or the collector on the grounded-out transistor will still swing a little.

[digi2t]

Dang... now it's got me thinking, "Were the original units like this too?". I can't remember. I just remember A/B'ing the clone with them, and getting the same sonic results. Or maybe it's been lost to the fog of time.

I wonder if Jimi's still got his?

[R.G.]

So I am assuming that the zener solved the switching issue with Q7/Q8 

It would appear so.

If it did... does it matter what threshold zener is used? (i.e. 12V, 16V, 17V, etc.)

I'd say as close to 13.5V as you can get, because that's what the flipflop puts out. No more than 17V (1/2 supply voltage) or the collector on the grounded-out transistor will still swing a little.

Correct, all counts.

Going back through this I remember some of the issues. Q7 and Q8 were not in the original, and the switch controlling them was a center-both switch. Those switches are rare, and I inserted Q7/Q8 so a center-off switch could be used. Works, but introduces the problem with the odd changes in voltage to the circuits it drives.

A quick look at the second-run boards shows that a good place to put the zener is with its cathode to the south end of R64  and its anode to the north end of R29. This is between the Fuzz Mix and Fuzz Repeats pots, next to the Repeats pot. Here's a quick look at where to put it on the top/component side:


It's also easy to hit those same pads from the bottom/solder side.

One thing that made this a little abstract is that the signals are AC coupled from the collectors of Q7 and Q8 and so the raw DC levels were removed, leaving only the AC portions, which were square waves, albeit of the wrong sizes.