someone upped this on "another" forum...worth a look!!!!!

[Mr Bill]

Gotcha!

[Mr Bill]

Well, MAJOR humbling experience.  When I started troubleshooting this circuit, I originally figured I needed to have the bypass signal path working first and was expecting  a true FLAT bypass.  Working on our companies studio-quality mixers, equalizers, crossovers all day, I was looking for a FLAT bypass!
Of course,  the bypass does not resemble this, it's more akin to the "reverse smiley" you set up on a graphic equalizer as a starting point when you go out to do a gig mix.

  When I originally saw the roll off on the sweep and when I sine swept it and observed it cornering around the 800 Hz region at the low end, I immediately went after the caps.  The only problem was that my generator was on "divide by ten".  OOPS! 80 Hz!  So, naturally when I started paralleling the interstage caps, I was seeing some improvement, but well you get the idea....

So, Jimi, leave those tropical fish in the tank!  Of course, I read R.G.'s Tech Overview" on the way home on the bus, and he comments on it being a buffered-bypass.     

  Oh well, water under the bridge.  I did find a number of bad transistors on the lower deck, my focus tomorrow will be getting the fuzz animation circuit to switch Q21 on/off and get the Q5/Q6 into the equation!

By the way, R.G., that "Tech Overview" is an exemplary piece!

[Mr Bill]

Have a question on Q21 on the lower deck, when the animation fuzz is not driving the base of Q21 (fuzz bypassed) do you have a quiescent  base and emitter voltage in the 2-3VDC region as Jimi had mentioned on his point-point transistor readings or does the base and emitter voltages read zero and depend on the animation circuit to solely switch on the rail to Q5 & Q6?

Lastly, could somebody post or re post the voltages on the upper deck transistors/test points?  I would assume these are all with all effect off.... 

Found a transistor up there (I think Q1, my schematic is at work) just in front of the UJT, that I had cut out of circuit in 1974 and replaced with another part.  When I went to measure the voltages on it, I kind of blinked.....is that....yep, I  had wired the   replacement transistor with the collector and emitter reversed!

Surely this was the ticket to finally go LIVE.  Replaced the tranny with a 2N4123 just to continue reviving the circuit...still no beans.....the upper deck is the key....getting closer!   

[Keppy]

Have a question on Q21 on the lower deck, when the animation fuzz is not driving the base of Q21 (fuzz bypassed) do you have a quiescent  base and emitter voltage in the 2-3VDC region as Jimi had mentioned on his point-point transistor readings or does the base and emitter voltages read zero and depend on the animation circuit to solely switch on the rail to Q5 & Q6?

Not sure what you mean. The animation circuit doesn't drive the fuzz that I can tell, unless you're talking about the fuzz repeat/perc. repeat. The Fuzz stomp switch just selects which point of the circuit to couple to the filters, so as far as I can tell those transistors should look electrically identical in both states, bypassed and not.

Lastly, could somebody post or re post the voltages on the upper deck transistors/test points?  I would assume these are all with all effect off.... 

control board transistor voltages all sliders @ min, bypass on, fx on

q1
e 33.8
b 31.2
c 35.5

q2 is oddball, it has TWO BASES
i don't know what the first terminal is, so...

?   7.2 - 4.9 keeps cycling
b1  14.4
b2  .3

q3

e  0
b  .5
c  10.9 - 0 keeps cycling

q4

e  0
b  .5
c  15.6 - .1 keeps cycling

q5

e  0
b .5
c  6.4 - 5.4  keeps cycling

q6

e  35.5
b  36.0
c  44.8

Found a transistor up there (I think Q1, my schematic is at work) just in front of the UJT, that I had cut out of circuit in 1974 and replaced with another part.  When I went to measure the voltages on it, I kind of blinked.....is that....yep, I  had wired the   replacement transistor with the collector and emitter reversed!

Surely this was the ticket to finally go LIVE.  Replaced the tranny with a 2N4123 just to continue reviving the circuit...still no beans.....the upper deck is the key....getting closer!   

Hang in there! R.G. said in an earlier post that that transistor is responsible for the fast/slow switching on the animation.

[Mr Bill]

Keppy, i just want to point out that your clone is gorgeous!

The last few weeks exploring the forum posts have been extremely exciting!  37 years of hoping SOMEBODY would be able to decipher the poor schematic rendering on the upper/lower schematics, let alone the horrific omittance of the upper deck pot/switch wiring!

[Keppy]

Thanks a lot, Bill! I tried to make the look of mine an homage to the original. I'm really pleased with how it turned out. It's only the second enclosure etch I've done, and it could've been a real mess!

For some reason, this project just caught my eye and I've been happy to contribute. It was fun running down all the omissions/errors to get it working. It took four of us, two original units, and some random quasi-related schematics to get it done, but we got it!


...


Now what can I do with it? I think between this, the Synthbox, and the Basic Audio-inspired Squarewave I'm planning, I'm going to need a different kind of band than I'm used to being in.

[Mr Bill]

So, Keppy , do you have an original?  If, in fact you built one without having one of the old beast's present, that's absolutely incredible... The (lack of) documentation and trying to remember what I did to this pedal 37 years ago is frustrating.
  What's wonderful is knowing that I had NOT built that 11th hour gain stage on the lug strip.  I kind of thought I had.  That was obviously a Ludwig, quick kludge.  I had PLANNED to duplicate it and cascade it to make up for the lack of gain on the fuzz circuit.  Back in '74 I assumed there was something wrong with the box when in fuzz mode, I mean who designs a pedal where when you go wet you can't even get to unity gain (Ludwig Electronic Division)  ..LOL! 
  You guys are providing me with what was the most frustrating experience, having the beast, but not able to bring it back to life!

[pinkjimiphoton]

like i said before, drummers on lsd....lol


so much more than just stage levels... (you know, that's how ya know if the stage is level, the drool runs out of both sides of their mouth evenly)

[Mr Bill]

Hey, Jimi!  Leave those tropical fish in the tank!  I'm SO close to getting this beast flying!  Now that I'm on the upper deck, I have nothing hooked to I believe  #12,  I think you stated that some people don't have anything connected there, five in from the edge of the board?  Looking at the copper trace there it actually is a dual connection on an adjacent trace, so they could've wired this connection "off board" at the pot/switch connections.  When the hundreds of technicians were building these harnesses, after about 10,000 they probably wanted to introduce a little poetic license to break up the monotony....

[pinkjimiphoton]

hi bill, yah, on mine, no connection there.

i don't think they had that many techs, lol...probably like 5 asian children in a sweat shop somewhere..lol.

can't wait to hear it's running bro...

gonna use mine today live...should blow a few minds!!

[Keppy]

No, I don't have an original, but I might as well have since Jimi & Dino were giving me so much info.

Funny story: I left the clone on a rug in my basement after making that video. That was the day we put down that rug. My dog started sleeping there. Now my wife tells me he hasn't slept there since I moved the pedal a couple days ago. I guess he likes it!

[pinkjimiphoton]

a bump to check in, guys...any news?

[Mr Bill]

Back in the pedal tonight.  One thing to point out is that what we call a "clean" signal coming off the lower board, Q3 collector prior to the fuzz section is indeed a pre-distortion network.  Q5, Q6 then provide the amplitude modulated "super fuzz via Q21 being AM'd.
Because Q3 has it's emitter tied to ground, the incoming waveform, will have the negative clipped off, as Q3 gets turned off during that period and the positive limited by the base-emitter drop. The collector waveform is a very asymmetrical waveform as you would expect, with the emitter grounded.  If there was an emitter resistance present, you'd have something to work against to reproduce the entire waveform.  But that isn't what we want here.  
R.G., you might want to amend your "tech Review" on this, as I kept assuming the signal must be clean at Q3's collector when it can not be.  Also when you get to the fuzz/non fuzz page, the "non fuzz" signal is really the pre-distorted signal.  You did a hell of a job on the analysis, R.G., just on this aspect I was spinning my wheels on assuming the Q3 collector waveform was supposed to be linear. I don't other people to get snagged on this.
Anyhow, I've now gone ahead into the two BPF sections and found that I do have drive into pin 7 on the lower board but no output on the filters, pin 6.  All transistor voltages are right on and as all capacitors have been changed, I was a little suspicious on how both filter paths could have "died." I double-checked pin 7 and pin 6 to make they were indeed making continuity to sections of the board.  The quiescent trajectory voltage points at the two trimmer wipers were just as Jimi had chronicled.  Just before I left I looked at R44 which is one of the filters output heading to the pin output.  I saw on the filter side a low level, band passed signal with nothing on the other side, which is hooked directly to pin 6.  I measure 5K to ground at pin 6. This is the point where we drive our little off board kludge, if I'm right.  I had a blown 2N4401 there, which I replacedand it then had I think .6V on the base and if I remember right, about a 10VDC for the collector reading.  I plan to pull the wire off pin 6 tomorrow to see if it sudenly pops up.
 

[Keppy]

Q3 is biased by R12, R13 & R14, giving it room to swing on the negative side. It might be under-biased, as I've noticed some distortion when I play hard in bypass mode. It's not half-wave rectified or anything, though.

[Mr Bill]

Hi, Kep - Not trying to draw any controversy here (anything but), just making an observance. My boss noticed this immediately when taking a glance at the schematic, I mean the base-emitter is a shunt diode with the emitter tied directly to ground (clip).  It produces a haversine (old school) at the collector.  Unless you're swinging a tiny signal way below the peak-peak voltage of the diode junction voltage how would you expect to reproduce the whole waveform? If I can find a scope available this morning, with capability to capture the waveforms to file I will post the Q3 base/collector waveforms.
  My reason to post was that I was spinning my wheels trying to figure out how "my" clean signal wasn't clean.  It's not supposed to be at that point, it's by design..   

[R.G.]

Back in the pedal tonight.  One thing to point out is that what we call a "clean" signal coming off the lower board, Q3 collector prior to the fuzz section is indeed a pre-distortion network.  Q5, Q6 then provide the amplitude modulated "super fuzz via Q21 being AM'd.
Because Q3 has it's emitter tied to ground, the incoming waveform, will have the negative clipped off, as Q3 gets turned off during that period and the positive limited by the base-emitter drop. The collector waveform is a very asymmetrical waveform as you would expect, with the emitter grounded.  If there was an emitter resistance present, you'd have something to work against to reproduce the entire waveform.  But that isn't what we want here.  

I'm a little hamstrung in this area because I've never seen a real Ludwig Phase II, nor even a clone, only the schematics; so what follows is based on theory.

In my experience, the circuit setup for Q3 matches a voltage-feedback gain stage. The 100K from the collector not only biases the stage with a current into the base, but also provides AC feedback to the base. There are series resistors into the input, so the actual DC voltage on the base is almost immaterial. The base balances currents and sits at a relatively fixed DC voltage, much like the inverting input of a more normal opamp sits at a "virtual ground".

Your post made me wonder if I'd missed something, so I popped the circuit into the simulator. It biases up with the collector at about 9Vdc on a 17V supply. I don't know exactly what the power supply voltage is there; if you could measure the DC voltage at the high end of the Q3 collector resistor, I could get some more accurate sims - or breadboard it. Anyway, the output is undistorted up to several volts of signal in the simulation.

That makes me wonder - is your board not a match to the factory schematic, or is is possibly faulty? Could you measure the DC levels on base, collector, and the top side of the collector resistor and post them?


 

[Mr Bill]

%^&*pit error on my end, R.G.,  I was hitting the front end too hard.      When you turn up the drive, Q3 is the first transistor to go non linear, with the output of Q2 looking fine. Of course it is a unity gain follower so you're not running out of swing as on Q3.
Turns out my formant filters are working.  What I don't have is any voltages on Q5, Q6, and Q21 other than Q21's collector voltage. The fuzz animation circuit obviously supplies the pulse necessary to turn on Q21.  But with the animation off, there must be a source to supply Q21's base to bias it on to some level, so the fuzz circuit will still pass signal. That's where I'm at right now.

[digi2t]

%^&*pit error on my end, R.G.,  I was hitting the front end too hard.

Would that explain why the formants are much cleaner (clearer), with the fuzz off, and you roll back on the guitar volume? It certainly is the case with both my units. If I turn off the fuzz, the formants are louder, but not as articulate.

[Mr Bill]

I'm going to unplug (I think wire connection #15, no schematic here) tomorrow that drives the base of Q21 and place maybe a 47k resistor in series with a variable power supply on the base of Q21 and walk the fuzz circuit through it's paces. If it comes alive, then the trouble is all upper deck.  

[Mr Bill]

Hi, digi2t -

My fuzz circuit is currently dead in the water due to Q21 not being "driven" correctly.  I'm going to jump start it tomorrow per my last post.  Once I get the fuzz fully working,  then the animation, I intend to run this through the ringer on our Audio Precision analyzer with gain, FFT, distortion +noise floor readings. Checking the formants through their range both dry and wet should be very interesting.  Just a quick response on your question, the fuzz circuit was always weak on gain in comparison to the dry signal.  This is a drum company and having the wet signal at least start out at unity gain and then progress from their, appeared to be foreign to them.
I would also like to get some captures of what I'm talking about in the previous posts with Q3 on the lower deck.  Back in the 70's, I always assumed the pedal was malfunctioning gain-wise on the fuzz circuit.  It wasn't until I saw posts maybe five years ago on Harmony Central that this was the norm.