A Hacker's Guide to the Ludwig Phase 2

[R.G.]

Fall Plate PCB:
Q1 => low impedance (probably sucks treble) but gainy input
Q2 => high impedance input, no gain here
Q3 => Mixes output of Q1 and Q2, adds some gain
Q4 => An emitter follower going somewhere. I can't find what happens to this with the info at hand. Possibly a "clean" output or a bypass output; stereo output?
Q22 => 'nother gain stage. Bypass Balance connects to base, probably attenuates the signal into this path down to the uneffected signal level. Output may or may not be distorted, still working that.
Q5/6/21 => voltage controlled fuzz gain stage. Percussion repeat and Fuzz Mix controls set the amount of modulation and static level of the distortion by varying the DC level on the base of Q21.
Q7-13 => formant filter 1. Input is Fall Plate pad 7, from fuzz selector switch pole. Input to formant filter 2 is paralleled with formant filter 1. Output is fall plate pad 6, also paralleled with formant filter 2.
Q14-20 = formant filter 2
Both formant filters are (probably) current controlled by the currents into pad 8 (formant 1) and pad 9 (formant 2). The filters are almost identical, with cap values and other slight changes to cause about a 1 octave difference in frequency for the same input voltage, formant 2 being the higher. I don't (yet) know exactly how the filters work, but at this point, that's not too important.

Formant trajectory switches "parallel", "counter", and "vowel" do just what the name says. Parallel moves both filters up and down in frequency in sync. Counter makes them play see-saw. Vowel causes some change in the path of F2 to make it hit human vowel hot spots as F1 varies.

The entire audio path is on the fall plate - good design for keeping extraneous clicks, hum, and other junk off the audio.

Console Plate
The console plate PCB contains all the control junk to move the two formant frequencies and the amount of fuzz around.

Q1 and "2N2646" form a variable frequency blocking oscillator. This thing is UJT based, and causes a spike every so often when C2 charges up to the peak point of the UJT. The charging speed is determined by the animation speed pot and the animation speed switch. It is not (yet) clear how the animation switch works, as the only info shows it with only two terminals, "slow" and "fast". That doesn't make sense, so I suspect that there is some other info missing.

When the UJT fires, it causes the bistable flipflop made up of Q3 and Q4 to flip to the other state. Opposite phase square waves appear at pads 25 and 26 and are fed into the "Fuzz Rpt Switch", which has two outputs. It's not clear what that switch does; possibly it only attaches/detaches the square wave to the "perc rpt" pot to diddle the fuzz level.  In any case, it causes some change in the basic fuzz operation.

One of the outputs also goes to the Q5 circuit that makes the formant control signals. The formant control signal from the fuzz rpt switch goes to the FFM amp pot, which is a level pot to adjust the amount of change the square wave makes in the formant controls. This is added to the signal from the foot rocker pedal and this drives, either directly or through Q5, the switch matrix of the formant trajectory switches to make the formant filters move around.

This design is an interesting example of late '60s design style filtered out to the periphery of the profession in the early '70s. No one would today do it this way.

Today, I would use two OTA's in a chip for each of the formant filters, and use opamps to make an LFO that didn't need so much circuitry. I'd try to get it into a 9V supply, or at least a +/- 9V supply, and probably would be successful.

The variable fuzz circuit is interesting; I have already thought of some interesting things to do with that as a basis.

I don't think this thing is a good candidate for pure cloning, certainly not with a couple of questions remaining on how the few places where I could not ferret out of the schemos the actual connections. However, it's probably a GREAT thing to use as a concept, much like the wondering I had done about using vocal formants as a pedal.

I'll have a usable schematic in a couple of days.

[JonC]

Hmmm, didn't Anderton do a voltage controlled fuzz in Device? I'm thinking as part of the AMS-100.  Might be a place to start.

[Ry]

Wow, cool!  I never got any farther than figuring out that the input transistor was probably a treble killer.  Thanks a lot for your analysis on this!

I may still try to build one up and try to figure out the question portions of it...maybe.  I'm honestly just interested in the formant/vowel portion of it rather than the fuzz.

Ry

[R.G.]

I'll still go ahead and flesh out the details as best I can, and if you can get any info on the points I'm missing, I'll finish up the drawings. I think we can do a two-formant setup pretty easily out of a couple of dual OTAs though. I wish I knew the values of some of the pots that the schemo refers to but doesn't state, like the footpedal and the fuzz mix pot, that kind of thing. And what that pesky pad 19 on the fall plate goes to.

I'll post the schemos when I get them finished. They're big, even redrawn.

I suspect that from the schemo and the parts layout, one could do the underlying copper traces to replicate the PCB's pretty easily if you wanted to do an exact repro.

I ..love.. archeofexology.

[Ed Rembold]

R.G.
This is very interesting, nice work!
Makes for great reading.
Thanks, Ed R.

[Ry]

I was actually working on the Seek Wah about a month ago, tuning the capacitor values a little bit more close to you specified values.  I also tracked down a couple TDK inductors to replace 3 inductors in series that I had.  It sounds like a nice, deep stepped wah, but it still doesn't quite do what I wanted...the Ludwig formant sounds...I am VERY excited to get the formant portion of that pedal working and at my feet

thanks again (the exclamation point key on my keyboard is broken or there would be one here).

Ry

[BillyJ]

This is a cool week for weird wah like stuff :O)