Ludwig Phase II guitar synthesizer calibration procedure

[soapfoot]

UPDATE: After much tail-chasing, I've figured out a calibration procedure that worked for me, on my unit (I can't make any claims broader than that).

I thought I'd share it here for informational purposes, as it diverges in some ways from the procedure Dino posted at DIYstompboxes in 2011 (linked below). This is not intended to replace Dino's procedure, nor to call it into question in any way--but I discovered some things for myself that may prove helpful as a supplement if anyone is, like I was, struggling to get their unit dialed-in.

So here's the procedure I arrived at that worked well for me:

[list type=decimal]
[*]Before applying power, measure precise value of R44
[*]Adjust R65 carefully until it is an exact match for R44
[*]Set all other trimmers on the fall board to center position
[*]Insert cable or dummy plug to unground the input.
[*]Take the cover off of the fall board and disassemble the console board unit. Orinent the console board/switch assembly in a manner that renders it accessible for work but avoids shorts.
[*]Citing the faceplate you just removed for reference, turn "contour" formant trajectory ON and all other formant trajectories OFF.
[*]Apply power and turn "Fuzz" and "Animation" footswitches OFF. Position the pedal fully toe-down. Allow at least ten minutes for the unit's zener-regulated supply to stabilize
[*]Measure and note voltage at CP18 (console board). If it's too far from 7.4V, set it closer to that range using R20 (orange trimmer) on console board.
[*]Back on the fall board, adjust R29 to achieve exactly 3.80 4.0 VDC at TP3. (edit: this differs from Dino's numbers but worked much better in my unit)
[*]As Dino alluded to in his procedure, take a deep breath. The next step requires a lot of tweaky patience.
[*]Measure and note voltages at CP8 (also called TP4) and CP9 (TP5) on the fall board. Shift the pedal to fully "heel down" and measure and note those voltages again.
[*]Ideally, those voltages will be opposingly-balanced and very-well matched. Heel down at CP8/TP4 matches toe down at CP9/TP5 and vice-versa
[*]Using R20 console board and R31 fall board, tweak voltage at these test points until you get a good balance between the two. You're looking for approximately 6.1-6.4 at one point with 2.4-2.7 at the other. The exact voltage matters, but a very good match seems to matter more as long as you're within this range. R20 (console) seems to have the most impact on both targets at CP8/TP4 while R31 (fall board) seems to have most impact on CP9/TP5. However they're very interrelated. Tweak one, check the other, and zero in on it. It will take time.
[*]If you're struggling to get within about a tenth of a volt for both points at both targets, try jumping a tooth in one direction or the other on the pedal's rack gear and start again. I'd consider this a last resort if things are just really not working.
[*]Check voltage at TP3 and make sure it's still 3.80 4.0 v. If not, adjust R29 until it is, and then touch up/reëvaluate your balance at TP4/5.
[*](Here's where my procedure begins to diverge from Dino's a bit) With pedal heel-down, measure DC voltage at the base of Q14 and Q20, and record those measurements. Do the same with pedal toe-down and record those
[*]Using trimmer R55 (fall board), balance the base voltages of Q14 and Q20 at both heel-down and toe-down pedal positions. Do your best to zero-in on the best compromise for both heel-down and toe-down positions
[*]Repeat the above procedure on the other filter--transistors Q13 and Q7, trimmer R77.
[*]Balance on the above step is also very important--an imbalance of 0.2v on one filter in my unit caused audible malfunctions, pops, gurgles and squeals in my unit. Ultimately I had to replace Q13 and Q7 with a pair that's more closely-matched for hFE. If you have a unit that has had some parts replaced, pay special attention to the balance of Q14/Q20 and Q7/Q13 and don't be afraid to change them if you have to.
[*]Ensure that trimmers R41 and R62 are near mid-rotation. Measure voltage at the base of Q8 and Q15 and use trimmers R41 and R62 (respectively) to set them exactly the same.
[*]Connect an input signal to the unit and connect the output to an amplifier. Try filter animation ON with fuzz OFF and see if things make sense and behave normally. Sweep the entire filter, both with animation ON and OFF, and see if everything seems sensible. Watch for squeals and oscillations; sometimes these can indicate an error with the adjustment of R55, R77, and sometimes R41 and R62.
[*]If you have filter ringing on particular notes that cause them to jump out, your voltage at TP3 may be too low. This sent me around in circles because I was going from Dino's number at that location, which did not turn out to be the best fit in my unit.
[*]Evaluate the speed of the LFO with the "rate" slider at minimum and maximum. Adjust trimmer R4 (console board) if you want to speed up or slow down the overall range. Dino recommended 20v at CP28 console board, but I ended up with mine in a different place. It is fine to adjust this step by ear (I'll reiterate others' advice to NOT attempt to adjust the other trimmers by ear--you'll ultimately give up, and only worsen your starting position when you ultimately have to get out the test equipment).
[*]Turn "fuzz" ON, turn "fuzz mix" fully UP, and evaluate fuzz. If fuzz is not working, check Q5, Q6 and Q21 (fall board). If turning fuzz OFF mutes all output, check Q22. If all is good, button it up (carefully!) and enjoy!
[/list]

-----------------------
(original post follows)
-----------------------
I'm working on calibrating this relic from 1970-1971.

These are historically difficult to calibrate, and documentation is scant. As far as I know, there is no extant service manual. Fortunately, my unit had schematics pasted inside the chassis (attached).

In lieu of factory calibration instructions, I've been following the somewhat-vague instructions in this decade-old thread on DIYstompboxes, which the user appears to have just figured out through trial-and-error:

https://www.diystompboxes.com/smfforum/index.php?topic=94478.0

I was able to do fairly well, even with the "difficult" step.

The problem is, when I get to the next step and scope TP8 and TP9 with "animation" on, I get wildly-asymmetrical amplitudes (TP8 is much higher in amplitude).

Checking some transistor voltages around TP8 and TP9 (for example Q15 and Q12, etc) I'm getting some asymmetrical readings... lower voltages on the TP9 side, across multiple transistors.

For one example: I have 16.4 on the collector of Q12, but only 9.4 on the collector of Q15. I'm not sure what this may mean.

I'm wondering if this may account in part for some strange behavior I'm seeing... the unit's functions all work, but the filters are very-resonant in a way that's hard to dial out. This does not seem normal from my admittedly-limited experience with these units.

I'll attach the schematics here just in case anyone may want to help me--I'm just stabbing in the dark.

 

[Gus]

Found some stuff
https://www.deadendfx.com/product/ludwig-phase-ii

another threadhttps://www.diystompboxes.com/smfforum/index.php?topic=91498.620

 

[soapfoot]

Thanks, Gus!

I'm familiar with those links--sadly, much of the documentation refers to the clone that RG Keen designed with help from others, which is not part-for-part for the original.

Still, I'll have a stronger look and see if I can manage to trace the schematic of the clone and see if I can translate it to the original--then I can at least look for "in the ballpark" voltages on some transistors and hope to track down the issue that way.

In the meantime, I've hosted the schematic files elsewhere, and added them to the original post of the thread. If anyone is inclined to have a peek, you might have to scroll left-to-right to see the whole thing.

Thanks again!

 

[synthiaks]

Wow!

that thing just sound fantastic! it even got me wondering if its used on the hook in "Da Funk" from Daft Punk

 

[soapfoot]

Okay! I have done some more sleuthing and it appears my unit may have a problem. I'm just not sure how to determine what that problem may be, as I have very little experience with discrete transistor electronics theory and design. So perhaps someone here can help me.

On page 18 of RG Keen's build notes for his clone here, he lists test voltages for nearly all of the transistors in the device.

I went through and compared his schematic to the originals pasted inside my unit to ensure that each transistor part number matched my schematic and layout (fortunately, they did!), and was able to compare these voltages with mine.

I found discrepancies centered around one area, and maybe someone here can help me sort out a potential cause and remedy?

(The "QF" part designations below are RG Keen's terminology; on the fall board schematic above, they're simply listed "Q5," "Q6," etc.)

QF5 bogey values are

• E: 3.3
• B: 2.6
• C: 0.2

My QF5 measured values are:

• E: 0.28*
• B: 0.02*
• C: 0.006*

*values increase with "fuzz" circuit turned up, but not all the way to bogey value

QF6 bogey values are

• E: 0
• B: 0.6
• C: 0.2

My QF6 measured values are:

• E: 0*
• B: 0.02*
• C: 0.04*

*values increase with "fuzz" turned up, sometimes meeting or even exceeding bogey value

QF7 bogey values are

• E: 3.6
• B: 4.2
• C: 8.8

My QF7 measured values are:

• (close enough, within a few tenths)
• (close enough, within a few tenths)
• C: 20

and finally, QF21 bogey values are

• E: 3.3
• B: 3.8
• C: 10.2

My QF21 measured values are:

• E: 0.28
• B: 0.62
• C: (close enough, within a few tenths)

Anyone have any insight? This is on the fall board schematic (the larger of the two). Thanks in advance, if anyone has the time!

 

[soapfoot]

Update (in case someone is following along, or for future people with similar issues):

I've traced most of the transistor discrepancies above to be in the fuzz circuit, which is working fine (and sounds great). Considering the values change somewhat when turning fuzz on and adjusting fuzz level, I assume those are unrelated to my issue.

However, I did find a bad transistor, QF22. I replaced it this morning with a 2N5551 and the unit mostly works great now.

My only issue is that some of the filter sweeps are a bit TOO resonant in spots--doesn't seem quite right.

Having noticed that my unit had been re-capped (electrolytics only), I decided to note the actual values and compare them with the schematic.

I found the following discrepancies:

C22, C23, C32 are specified as 2.5µF and have been replaced with 3.3µF (doubtful this would be enough to really throw things off too much).

C14 is specified as 2.5µF but has been replaced with 10µF. Is there reason to suspect this may have been an intentional change, or could this have been an error?

C31 is specified as 10µF and has been replaced with 47µF. Same question--is this an "upgrade," or a "mistake"?

Thanks!

 

[soapfoot]

I just wanted to add one more post of information here. I arrived at a calibration procedure that worked well for me and wanted to share.

What I ultimately arrived at has a few differences from Dino's approach on DIYstompboxes (which is the most commonly-cited cal procedure for this unit).

I'll edit the OP of the thread to include this procedure for anyone seeking this information in the future.

So here's the procedure I arrived at that worked well for me:

[list type=decimal]
[*]Before applying power, measure precise value of R44
[*]Adjust R65 carefully until it is an exact match for R44
[*]Set all other trimmers on the fall board to center position
[*]Insert cable or dummy plug to unground the input.
[*]Take the cover off of the fall board and disassemble the console board unit. Orinent the console board/switch assembly in a manner that renders it accessible for work but avoids shorts.
[*]Citing the faceplate you just removed for reference, turn "contour" formant trajectory ON and all other formant trajectories OFF.
[*]Apply power and turn "Fuzz" and "Animation" footswitches OFF. Position the pedal fully toe-down. Allow at least ten minutes for the unit's zener-regulated supply to stabilize
[*]Measure and note voltage at CP18 (console board). If it's too far from 7.4V, set it closer to that range using R20 (orange trimmer) on console board.
[*]Back on the fall board, adjust R29 to achieve exactly 3.80 4.0 VDC at TP3. (edit: this differs from Dino's numbers but worked much better in my unit)
[*]As Dino alluded to in his procedure, take a deep breath. The next step requires a lot of tweaky patience.
[*]Measure and note voltages at CP8 (also called TP4) and CP9 (TP5) on the fall board. Shift the pedal to fully "heel down" and measure and note those voltages again.
[*]Ideally, those voltages will be opposingly-balanced and very-well matched. Heel down at CP8/TP4 matches toe down at CP9/TP5 and vice-versa
[*]Using R20 console board and R31 fall board, tweak voltage at these test points until you get a good balance between the two. You're looking for approximately 6.1-6.4 at one point with 2.4-2.7 at the other. The exact voltage matters, but a very good match seems to matter more as long as you're within this range. R20 (console) seems to have the most impact on both targets at CP8/TP4 while R31 (fall board) seems to have most impact on CP9/TP5. However they're very interrelated. Tweak one, check the other, and zero in on it. It will take time.
[*]If you're struggling to get within about a tenth of a volt for both points at both targets, try jumping a tooth in one direction or the other on the pedal's rack gear and start again. I'd consider this a last resort if things are just really not working.
[*]Check voltage at TP3 and make sure it's still 3.80 4.0 v. If not, adjust R29 until it is, and then touch up/reëvaluate your balance at TP4/5.
[*](Here's where my procedure begins to diverge from Dino's a bit) With pedal heel-down, measure DC voltage at the base of Q14 and Q20, and record those measurements. Do the same with pedal toe-down and record those
[*]Using trimmer R55 (fall board), balance the base voltages of Q14 and Q20 at both heel-down and toe-down pedal positions. Do your best to zero-in on the best compromise for both heel-down and toe-down positions
[*]Repeat the above procedure on the other filter--transistors Q13 and Q7, trimmer R77.
[*]Balance on the above step is also very important--an imbalance of 0.2v on one filter in my unit caused audible malfunctions, pops, gurgles and squeals in my unit. Ultimately I had to replace Q13 and Q7 with a pair that's more closely-matched for hFE. If you have a unit that has had some parts replaced, pay special attention to the balance of Q14/Q20 and Q7/Q13 and don't be afraid to change them if you have to.
[*]Ensure that trimmers R41 and R62 are near mid-rotation. Measure voltage at the base of Q8 and Q15 and use trimmers R41 and R62 (respectively) to set them exactly the same.
[*]Connect an input signal to the unit and connect the output to an amplifier. Try filter animation ON with fuzz OFF and see if things make sense and behave normally. Sweep the entire filter, both with animation ON and OFF, and see if everything seems sensible. Watch for squeals and oscillations; sometimes these can indicate an error with the adjustment of R55, R77, and sometimes R41 and R62.
[*]If you have filter ringing on particular notes that cause them to jump out, your voltage at TP3 may be too low. This sent me around in circles because I was going from Dino's number at that location, which did not turn out to be the best fit in my unit.
[*]Evaluate the speed of the LFO with the "rate" slider at minimum and maximum. Adjust trimmer R4 (console board) if you want to speed up or slow down the overall range. Dino recommended 20v at CP28 console board, but I ended up with mine in a different place. It is fine to adjust this step by ear (I'll reiterate others' advice to NOT attempt to adjust the other trimmers by ear--you'll ultimately give up, and only worsen your starting position when you ultimately have to get out the test equipment).
[*]Turn "fuzz" ON, turn "fuzz mix" fully UP, and evaluate fuzz. If fuzz is not working, check Q5, Q6 and Q21 (fall board). If turning fuzz OFF mutes all output, check Q22. If all is good, button it up (carefully!) and enjoy!
[/list]