Ludwig Phase II Clone Tech Notes

[Keppy]

So I've been pondering that preamp circuit, and I was wondering if it was traced correctly. A 47k resistor from base to collector caused it to work in my unit. There is a 47k pulldown resistor between the blocking cap and the bypass switch on the wiring diagram. Did we get that right? Or is that resistor maybe supposed to go from base to collector instead? I know I'm asking for a triple-check here, so sorry. It's just that my brain finds it offensive that a circuit that works perfectly well in the originals does not work without mods on my bench. 

[R.G.]

You and me both!

[Keppy]

So with Q2 back in socket, the emitter drops to 2V and stays there, right?

That's right.

I'll have to go do some digging, but with E at 2V and B1=B2 = 2.47, the base/emitter is reverse biased and conduction should turn off unless the current through B1 and B2 is high enough to hold it on.

That's not. B2 (middle pin) is only about .7v.

[Keppy]

R29, R31 (close to the middle), and R20 (about 2/3 clockwise), should get you close. Other than R4, which adjusts the rate range, I've found the other trimmers only to affect frequencies when the unit is in fuzz repeat mode.

R20 on the fall plate board is not a trimmer. Did you mean R20 on the console board or something else? I thought R29 was the one that needed to be 2/3 up because of the switch from 150R to 100R (which I guess would be the middle on yours). Is there another one to try at 2/3?

[R.G.]

2N4870 appears to be a direct replacement for 2N2646, but in a plastic TO-92 package.

[R.G.]

I got a model running of the 2N2646.

Now I'm faced with not trusting it. 

Anyway, the 2N2646 is sensitive to the maximum value of resistance feeding into the timing cap (1uF in this case.) Over about 500K and it won't run because it won't charge the cap up enough. So for testing your board, it might be wise to short pads 28 and 11, thereby taking the external animation speed pot (500K) out of it, and using only the R4 trimmer to set speeds. It has enough resistance to do fine on its own. Actually, setting R4 in the middle should make the UJT work.

Here's the thing - you probably can't see the output on B1 or B2 on other than an oscilloscope. B2 is a fast pulse down, B1 is a fast pulse up. What you can look for is to look at wire pads 25 and 26 on the console board. These should flip-flop-flip-flop and be controlled in timing by the setting of R4. If they're flipping, that part is working.

[digi2t]

Did you mean R20 on the console board or something else?

Yes, trimmer R20 on the console board. My apologies, I should have been more precise. I just figured that since we were talking trimmer, that it was a given  .

I thought R29 was the one that needed to be 2/3 up because of the switch from 150R to 100R

You're right, I'd forgotten about that mod. For sure, the setting will be different to compensate for the different value. The sweet spot should now be around the 2/3 point.

If you need me to test something Kep, don't be shy.

[Keppy]

If you need me to test something Kep, don't be shy.

Thanks Dino! Could you recheck that preamp wiring like I mentioned at the top of the page?

[Keppy]

I just thought of something. Doesn't the animation move the formant filters around? If so, shouldn't I get the formants working first? Otherwise, I don't know how I'll be able to tell if the animation is fixed, short of keeping my meter on Q2 at all times.

This is one of those questions that's gonna make me feel stupid regardless of the answer.

[digi2t]

OMG!!!!! YOU GUYS ARE GONNA KILL ME!!!!!

There is a 68K resistor between B and C on that terminal circuit. I only just spotted it because it soldered to the backside of the terminal strip, hiding the resistor between it and the casting of the fall board. MAN I FEEL LIKE A NUMBSKULL FOR NOT HAVING VERIFIED BEFORE!!! I had a sudden spurt of curiousity because the soldering just struck me as a tad bizarre, and decided to check in the corner, and there it was. DAMN!!!

R.G., update your wiring diagram to reflect this, the rest of the circuit is correct. My sincerest apologies Kep, I know you're working hard on this. Hopefully, this will cinch it, formants and all.

[R.G.]

I just thought of something. Doesn't the animation move the formant filters around? If so, shouldn't I get the formants working first? Otherwise, I don't know how I'll be able to tell if the animation is fixed, short of keeping my meter on Q2 at all times.

This is one of those questions that's gonna make me feel stupid regardless of the answer.

Not at all. You maybe did this intuitively, but you were intuitively right.

You have to verify the simple stuff that everything else depends on first. It's like making sure the power supplies are right before testing everything else. It's pretty certain that if the voltages from the rocker pedal and animation don't tell the formant filters to move, they won't move. Even if they are working perfectly.

I would start with measuring the voltage on pins 8 and 9 of the fall plate/filter board. If those voltages don't move, the filters aren't being told to move. Depending on the settings of the switches and controls, they should move with either animation or the rocker. Using the rocker is probably easier to get going than anything else, and it's output is all DC.

A very helpful contribution to debugging (digi2t... wink, wink, nudge, nudge) would be to know the range of DC voltages the pedal rocker pot makes those two points move in a working unit. Once we get that right, we can look at the filters even if animation is not working.

There is a 68K resistor between B and C on that terminal circuit. I only just spotted it because it soldered to the backside of the terminal strip, hiding the resistor between it and the casting of the fall board. MAN I FEEL LIKE A NUMBSKULL FOR NOT HAVING VERIFIED BEFORE!!! I had a sudden spurt of curiousity because the soldering just struck me as a tad bizarre, and decided to check in the corner, and there it was.

R.G., update your wiring diagram to reflect this, the rest of the circuit is correct. My sincerest apologies Kep, I know you're working hard on this. Hopefully, this will cinch it, formants and all.

Got it! I'll go update the info. It almost had to be something like that.

Don't be too upset - remember that we can expect this to be like peeling onions. 

This thing is on a par for complexity with the old Maestro Rhythm'N'Sound unit. It's a beast. Patient application will get us there.

[digi2t]

A very helpful contribution to debugging (digi2t... wink, wink, nudge, nudge) would be to know the range of DC voltages the pedal rocker pot makes those two points move in a working unit. Once we get that right, we can look at the filters even if animation is not working.

Voltage between points 8 and 9 while moving the pedal, or across the pot? Or a point to ground? Let me know which points to measure, you'll have it later tonight.

[Keppy]

No worries, Dino! So far, that resistor's been the easy part!

Now to other parts: I want to make sure I wired this thing correctly, so please confirm/correct the following:

- Animation footswitch wiper is grounded in bypass and lifted when on (or is it the other way around?).
- "Off" terminals of vowel switch have no connection.

So for testing your board, it might be wise to short pads 28 and 11, thereby taking the external animation speed pot (500K) out of it, and using only the R4 trimmer to set speeds. It has enough resistance to do fine on its own. Actually, setting R4 in the middle should make the UJT work.

No need to short, as turning the pot all the way up accomplishes this (verified with continuity test). Reading the schematic, the resistance between the emitters of Q1 & Q2 should be the total of R4, R5 and the speed pot. I have verified this with my meter, so I seem to be good resistance-wise.

Here's the thing - you probably can't see the output on B1 or B2 on other than an oscilloscope. B2 is a fast pulse down, B1 is a fast pulse up. What you can look for is to look at wire pads 25 and 26 on the console board. These should flip-flop-flip-flop and be controlled in timing by the setting of R4. If they're flipping, that part is working.

Not flipping. Got some weirdness, though. I'm reading a stable 15.7v difference between these, but only .06v to ground from either one. 



Other notes:
- I used an audio taper pot for the Bypass Balance. I'm glad I did.
- I used a 25k log pot for the Lo Z Balance, as it was the lowest audio taper I had left and any value large enough should work as a variable resistor (as opposed to a voltage divider). I've only tested it with a guitar, but it works fine. Note that it will never turn the Lo Z input all the way off.
- R65 indeed causes oscillation if turned up too far. Well, turned too far clockwise based on R.G.'s layout, anyway. So I got that going for me.

[Keppy]

I measured the voltage to ground at fall plate pins 8 & 9. At both points, I get about 2.5 volts at the heel, 4.5 at the toe. Maybe start with those, Dino.

Additionally, in parallel mode, the pedal controls the amount of noise I hear. Heel down, less noise, toe down, more noise. It sounds like it's affecting the treble as well, though it's hard to hear through the hiss at the moment. The noise is dependent on R65, which seems to be best set in this regard almost all the way counterclockwise, as stated by Dino & Jimi earlier.

I just thought of something. Doesn't the animation move the formant filters around? If so, shouldn't I get the formants working first? Otherwise, I don't know how I'll be able to tell if the animation is fixed, short of keeping my meter on Q2 at all times.

This is one of those questions that's gonna make me feel stupid regardless of the answer.

Not at all. You maybe did this intuitively, but you were intuitively right.

What I meant was, if I was wrong I misunderstood the circuit, but if I was right I should have thought of this earlier. I'd hate to get the animation working and not know it because I couldn't hear the filters!

[Keppy]

I've been thinking that some of the switching diagrams could be clarified a bit. For example, Dino tells me I got the fuzz toggle backwards. In most cases I was able to refer to the schematic, but not all. Perhaps some on/off notes would be helpful. Incidentally, the fuzz toggle wiper to fall plate point 19 activates "Voice Fuzz." A toe/heel marking for the treadle pot would be good too.

On a sort-of-related note, does the direction of the arrow on a variable resistor symbol (like the Animation Speed or Lo Z Balance pots) mean anything? I've been wondering about that.

[Keppy]

So for testing your board, it might be wise to short pads 28 and 11, thereby taking the external animation speed pot (500K) out of it, and using only the R4 trimmer to set speeds. It has enough resistance to do fine on its own. Actually, setting R4 in the middle should make the UJT work.

Just noticed that my earlier post failed to say that I did try this, and got no result.

[R.G.]

Voltage between points 8 and 9 while moving the pedal, or across the pot? Or a point to ground? Let me know which points to measure, you'll have it later tonight.

Voltage to ground on each separately. Looking for minimum and maximum voltages to ground as the rocker is moved to its extremes.

No need to short, as turning the pot all the way up accomplishes this (verified with continuity test). Reading the schematic, the resistance between the emitters of Q1 & Q2 should be the total of R4, R5 and the speed pot. I have verified this with my meter, so I seem to be good resistance-wise.

Oh, sure, do it the simple, straightforward way!  

Not flipping. Got some weirdness, though. I'm reading a stable 15.7v difference between these, but only .06v to ground from either one.  

Hmmm. DC and not DC. Almost sounds like they're flipping madly faster than a meter can follow. That could mean that the two-transistor flipflop circuit is hosed. I'll try working up a diagnostic test for it.

Hmm. What happens if you pull out Q2, leaving them with no trigger? That should leave them forever in the same state they come up, with one collector high and one low.

Well, turned too far clockwise based on R.G.'s layout, anyway. So I got that going for me.

Yeah. Be suspicious of the direction the pots turn. I have a classic mistake I make of getting pot rotation backwards.

Just noticed that my earlier post failed to say that I did try this, and got no result.
[/quote]
Bummer. Well, it does eliminate the "resistance too big" from being the cause of the animation UJT not running.

[Keppy]

Hmm. What happens if you pull out Q2, leaving them with no trigger? That should leave them forever in the same state they come up, with one collector high and one low.

I get the same readings. Doesn't matter if the Animation footswitch is on or off, Q2 in or out. I've noticed that those joints are a bit touchy when I'm reading. I'll reflow the solder and check again.

[Keppy]

Hmmm...

So I reflowed the solder at 25 & 26, and now they appear to be cycling. With Q2 still out of the circuit.

[Keppy]

And now they've stopped. Maybe I need some fresh air.