Ludwig Phase II Clone Tech Notes

[digi2t]

Yeah, that's kinda messin' me up as well. I've looked at them a bunch, and played with them more. They're wired, and work like the description.

Jimi, did you roll that sucker a bit too tight bro? 

[R.G.]

A common problem in the music repair business is what the repair techs sometimes call 'amateur genius' repairs before they got the equipment.  I wonder if while Jimi's unit was out of his hands it was fixed at. It would only take a wire or two being moved in that string of switches to make it do radically different things.

[digi2t]

A common problem in the music repair business is what the repair techs sometimes call 'amateur genius' repairs before they got the equipment.

Hmmm, spillover from the automotive industry?  I'm figuring that maybe the tabs are bent, where the wires are soldered, possibly shorting some connections together. Those tabs are long enough to touch.

R.G., did you get my current reading from my previous post? Does 40.5mA sound right to you? I thought it would be higher, but then again, voltage is higher here, right?

I'm starting to get that melancholy feeling that this adventure will soon be over. What's the next project guys? EMS Synthi Hi-Fli? I found this btw; http://jbemond.free.fr/SDIY/EMS/EMS_Hifli.pdf  

[R.G.]

R.G., did you get my current reading from my previous post? Does 40.5mA sound right to you? I thought it would be higher, but then again, voltage is higher here, right?

I would have guessed a total of 50ma if forced to say, so yes, that seems fine to me. That figure has nothing in it for the lights because they come from low voltage AC switched directly by parts of other switches. And so it doesn't account for any power used by LEDs if you use LED annunciation a lot.

I've been looking at recommended power supplies, and I'm starting to come around to Keppy's very practical view that you can get 48Vdc wall warts for under $20 at Mouser every day. These are uniformly 300ma to 500ma in that price range, so what does another hundred ma for LEDs matter? I'll probably come up with a few alternatives, but for solid works-first-time operations, that's probably the way to go. Things will get ugly if one plugs in the 48Vdc into the 9V daisy chain, but then education is always expensive, no matter how you pay for it.   

I'm starting to get that melancholy feeling that this adventure will soon be over. What's the next project guys? EMS Synthi Hi-Fli?

Sure. Just go track down a couple of "victims" to test on. 

We are close to where someone who's willing to etch a board can do their own. It's a little scary that the first version worked after we figured out the parts value changes and the pinout of the 2N2646. No adventure left, eh?

[pinkjimiphoton]

yah, i think my mind has been smoking the chocolate covered crack or something...looking at the circuit, it seems no way the switches could interact, or SHOULD interact.

(no, i don't smoke crack, by the way.. lol)

one of these days, i will open it up, don't believe it was modified tho in any way...the only thing i found was the damage done originally when the switch had broken.

yes, this adventure may be coming to an end, as all things must...but think about it guys...we cracked a nut that people have tried to crack for around 40 years, and in just a couple months.

we obtained units...we cracked 'em open, we traced the stuff, we repaired one, we built one... that's something nobody else can lay claim to...and in just a couple months. incredible.

now, i dare say, the adventure is over....we reached our destination. or DID we?

what about improving the circuit as we'd talked about? making the formant filters out of a couple chips? making it have ample gain instead of slightly less than unity? making the fuzz actually useful and useable?

i say, the adventure has really just begun. now we know how to build it, we know how it works...and we can update it and improve it, and maybe make something that no one ever envisioned from this humble starting point.

i can see this having a loop added, better fuzz, more control over which formants, yada yada...


what do ya say? is it worth seeing what we can build together, or should we just rest satisfied on our well-deserved seats of achievement?



man...i shoulda been a politician. except i'm not a crook.   

[digi2t]

Hi guys,

Got the other unit today. Esthetically, it's A++. Totally complete, knobs, trim, handle, lights, the works. Extremely clean. Even got the original shipping box, and an original ad  .

So the skinny is this. It had fuzz working, but the pedal was dead, and no animation. You could hear the formant filters come on, but that was it.

Inside of an hour, I had it working. Loose terminal clips on the console board, and the treadle pot had been forced, so the pot had turned, and one of the lugs had grounded to the base  .

The rate slider will have to be cleaned though, I'm only getting action on the top 1/3. Unfortunately, someone had already gone in and really messed around with the trimmers, so there's no useful info to be gleaned here. I did set it up with the same setting as my other unit though, and they both sound exactly the same. R65 was set all the way down (ccw), like my other one when I got it.

Here's the interesting bit. My first unit is serial #1277, no piggybacked resistor on the fall board. This unit is serial #1372, and like Jimi's, it's got the added resistor. I think Jimi's is in the 1300's series as well, if memory serves me.

I'm going to look into some different clips, but I think I'm going to just end up soldering all the wires direct. No fuss, no muss. Will work forever.

[Keppy]

So the 1M parallel resistor has no effect on the sound at the settings Dino was able to try. R.G, you said that was part of the output section, right? So any audible effect should happen at all settings, right? Weird. I wonder why they started putting that in, since the unit works fine without it.

[pinkjimiphoton]

i'm telling you guys, DRUMMERS ON LSD!!!   

my friend mr bill has joined the forum, too...he's the guy i mentioned that was so jazzed we did this, as now he can fix his which has been broken since 1973.
be interesting to see which revision his one is!!

[digi2t]

So the 1M parallel resistor has no effect on the sound at the settings Dino was able to try. R.G, you said that was part of the output section, right? So any audible effect should happen at all settings, right? Weird. I wonder why they started putting that in, since the unit works fine without it.

I'll try this tomorrow; I'll disconnect one end of the 1M resistor with the iron, and do an A -  B test of sorts. Just toggle the resistor on and off, and listen to the difference. It bridges between B and C of Q7, so maybe I'll hear what Q7 is up to at the same time.

[R.G.]

I think what happened is that they got a design that was not tolerant of part variation.

In my era of EE schooling, designing cirucits to work pretty much irrespective of whatever junk parts you put in was considered an art form. The idea was to make your circuits sensitive to one or two things that set how you wanted them to work, and insensitive to wide variations on all the rest.

The alternative was a high percentage of finished goods that did not work as designed without individual tweaking. And even trimmers were a sign of non-manliness in design. If you had to put a trimmer in, it was a dead giveaway that you were not really cut out for circuit design. At the very best something you didn't want noticed.

I think they got the design, built them, then found out that half of the first batch didn't work. Then they started patching.

[Scruffie]

what about improving the circuit as we'd talked about? making the formant filters out of a couple chips? making it have ample gain instead of slightly less than unity? making the fuzz actually useful and useable?

i say, the adventure has really just begun. now we know how to build it, we know how it works...and we can update it and improve it, and maybe make something that no one ever envisioned from this humble starting point.

i can see this having a loop added, better fuzz, more control over which formants, yada yada...

Now this might sound crazy but how about... simplifying the circuit in to a standard phaser that'll fit a 1590BB.

Maybe just one modulated filter (okay you'll loose that vocal sound but who knows what it might be like) add in a modern sounding fuzz/distortion option, Opamp LFO and a nice new input and output, true bypass and get it to run off 9V (or maybe use a LT1054 charge pump, not sure how the design works) and have it run off standard transistors.

I'll have to study these schematics to see if that's doable, at the moment i'm just guessing from a brief glance of R.Gs tech notes.

[Keppy]

Now this might sound crazy but how about... simplifying the circuit in to a standard phaser that'll fit a 1590BB.

Maybe just one modulated filter (okay you'll loose that vocal sound but who knows what it might be like) add in a modern sounding fuzz/distortion option, Opamp LFO and a nice new input and output, true bypass and get it to run off 9V (or maybe use a LT1054 charge pump)

It kinda sounds like you want a completely different effect. The Phase II is based off of bandpass filters, not all-pass/phase shifting filters. If you want an LFO-controlled single filter (not phaser) you just might be better off with the Lightwah (a recent project of deadastronaut's, if you haven't seen it yet). Even if you wanted to change the type of filters or use selectable filters, that sounds more like what you're describing than the Phase II does, IMO. That might be a better starting point for what you're after (which sounds like a cool project, BTW).

[Scruffie]

Now this might sound crazy but how about... simplifying the circuit in to a standard phaser that'll fit a 1590BB.

Maybe just one modulated filter (okay you'll loose that vocal sound but who knows what it might be like) add in a modern sounding fuzz/distortion option, Opamp LFO and a nice new input and output, true bypass and get it to run off 9V (or maybe use a LT1054 charge pump)

It kinda sounds like you want a completely different effect. The Phase II is based off of bandpass filters, not all-pass/phase shifting filters. If you want an LFO-controlled single filter (not phaser) you just might be better off with the Lightwah (a recent project of deadastronaut's, if you haven't seen it yet). Even if you wanted to change the type of filters or use selectable filters, that sounds more like what you're describing than the Phase II does, IMO. That might be a better starting point for what you're after (which sounds like a cool project, BTW).

Yeah I know it wont sound like a Ludwig any more... but for some reason the idea appeals to me, a bit like Beans new project using the original univibe LFO with lamps and all but hooked up to a simple transistor vibrato (think it's Escobedos Wobbletron) kinda like the Run off groove Phozer.

Based off bandpass huh... as I said I didn't read the notes or check the schematic yet, just going off bits i've picked up from glancing at these two threads.

Swept bandpass could be cool though and yeah I could go off deadastronauts project... or I could make a bat shit insane overly complex effect that I can say is based off a Ludwig  even if the results aren't too dissimilar.

You might be right about using just a single filter though... it's probably more worth it to just have both... sure it'll complicate it a bit but I think it'd still fit a 1590BB judging by the pictures of the original board I saw.

[R.G.]

It is entirely possible to start with the concepts in the Ludwig and come up with a more modern pedal. I have actually done that already.  It was more direct than getting the original to work as it turns out - no long-ago quirks to figure out. I also left out many of the functions of the animation circuits, as those seemed to be overly complex for my taste.

I had always been interested in using formant filters to make voice imitators; that's what led me to put the stuff about wahs and voice formants at geo. But with one thing and another, it didn't get done until I started messing with the Ludwig again. Choice of filters is one quirky issue. Multiple feedback is quirky, wah style needs one inductor per filter, and state variable does everything you want and so much more that it's hard to throttle them down to what you do want.

The Ludwig is an interesting archeological widget. It probably makes more sense to design from principles for today's technology than to simplify it down. But tinkering for the purpose of intellectual curiousity is always a valid thing to do. Go for it.

If I were redesigning the Ludwig today, I'd probably use a CMOS logic or a 555 for the animation LFO, which has a square wave output after all. A single 556 might do all of that, including the opposite-phase output. I'd use some other filter format, at least partly because I still don't understand how the Ludwig's filters work, and don't feel inclined to dig it out because they require a lot of hardware and real estate, and there are other ways to do it. I'd be tempted to provide the control/animation as as microcontroller with PWM outputs for the control voltages, as this would go directly from an input control to output voltages for the filters in one step with a cloud of programming.

[digi2t]

Hey R.G., is there a CMOS chip that reproduces mojo?   We have an expression here in Quebec. If we didn't love you, we wouldn't tease you.  

I did the test on the piggy-backed resistor at R33 on the fall board. It helps to squelch some to the high end frequencies. If you really slow down the "yoyoyoy" to "yooooooiiiiiiiii", the "iiiiiiiii" part sounds less harsh with the added resistor. Try it Kep, just take a 1M resistor, and jump it on manually while the unit is modulating, you should hear some of the high frequencies get killed off. I'll probably throw that mod onto unit #1 as well, it sounds better. Probably explains why my unit always sounded a tad brighter that Jimi's.  

R.G., quick question (like I expect a straight answer after my little shot before  ), on the second unit, I've got everything working, but the rate is not kosher. I'm getting modulation, but slow to fast, only in a small range of the slider. Adjusting R4 only seems to move that range either up or down on the slider pot. Also, the modulation is veeerrrrry slow to come on, even in fast mode. To my noob eye, I'm thinking that maybe C2, or C3, must be close to dead, not allowing the 2N2646 flip to last long enough to spread it out over the 500K range. Is my assumption correct?

Other than that, everything on this unit seems to be OK now. Soldering the clips to the pads is a +1, and I lost track of how many cold solder joints I had on the console board. To the point that R20 was doing nothing. Now, you can hear the frequency change as you move the dial. Must have been a Friday afternoon unit off the line. It would explain why the previous owner messed with all the trimmers, probably trying to figure out what the hell was wrong with it. Except for the rate range problem, I've got it sounding exactly like unit #1, yoyoyoy and all.

[pinkjimiphoton]

i'd be interested in seeing if we could create a modernized version of the phase ii...maybe call it the phaseIII...but don't wanna see it lose it's mojo.

once i get a clone built, i may just send you my unit rg so you can check it out for a while. it  is addictive enough where i'd worry you wouldn't send it back...

but i believe that it is a little cooler than it may appear once you actually play one for a while.

and i think, tho we've cracked the nut and are seeing it's guts in a whole new light, it may take a while to get a true grip on all that it does.
jmo.

[R.G.]

Hey R.G., is there a CMOS chip that reproduces mojo?   We have an expression here in Quebec. If we didn't love you, we wouldn't tease you.

Actually there is. I've applied for patents, copyright, *AND* trademark protection as well as general deityhood for a new Mystic Mojo (c) (tm) (r) (pdq) chip that I make myself out of leftover dual opamps crossbred with industrial programming controllers. The real secret is pickling the chips after crossbreeding by burying them at the crossroads at midnight on the night of a full moon in a mixture of lime juice and leftover coffee. A little salt doesn't hurt, either, and is useful for an ocean tone for playing sea chanties.

Fortunately, I was able to work out a way that didn't require Icelandic virgins. 

...on the second unit, I've got everything working, but the rate is not kosher. I'm getting modulation, but slow to fast, only in a small range of the slider. Adjusting R4 only seems to move that range either up or down on the slider pot. Also, the modulation is veeerrrrry slow to come on, even in fast mode. To my noob eye, I'm thinking that maybe C2, or C3, must be close to dead, not allowing the 2N2646 flip to last long enough to spread it out over the 500K range. Is my assumption correct?

Interesting question. The speed to come on is determined by R1 and R2 on the console board charging C1. Until that ramps up from zero after the animation switch lets it go, no animation happens. C2, R4, R5 and the animation speed pot determine the actual speed that the animation runs once C1 comes up to the voltage that makes it run. C3 isn't really involved in the animation speed. It's either big enough and good enough to trigger the flipflop of Q3/Q4 or it isn't.

I think you have a good intuition on the cause, not enough voltage for the animation LFO, but ought to also be suspicious of C1. If it were mine, I'd replace both C1 and C2. Well, OK, I'm a bit of a nut about electros, I'd replace all the electros, but for these symptoms, those two particularly.

[pinkjimiphoton]

lol...+1 on the electros...don't risk catastrophic failure!!!!

rg...i'll take them chips, i gotta effect that could use them for the output stage..

it's basically a micro leslie run by hamsters of a secret steroidal cactus juice i devised while mixing ammonia and bleach with...

well...i can't give ALL the particulars...but let me tell ya what, it creates one heck of a stomach churning effect!!

[digi2t]

DAMN!!!! IT WAS THE LEFT OVER COFFEE!!!! I can't imagine how many Icelandic virgins you've been through to date  

I think you have a good intuition on the cause, not enough voltage for the animation LFO, but ought to also be suspicious of C1. If it were mine, I'd replace both C1 and C2. Well, OK, I'm a bit of a nut about electros, I'd replace all the electros, but for these symptoms, those two particularly

+1 on that. I believe C6 might be acting up as well. Now I'm starting to lose the yoyoy vowel sound somewhat, and I'm getting popping when I switch between different formants. Also, moving R20 yeilds no frequency changes at all. I think a complete cap job is in order here. No point f**king around with the puck here, might as well shoot and (hopefully) score.

Thanks R.G.!

[digi2t]

To the point that R20 was doing nothing. Now, you can hear the frequency change as you move the dial.

Found the problem here. Bad solder joints at the junction points of R16, R18, and R19. Q5 is now reacting properly to R20 input.

The soldering on this puppy is a freakin' trainwreak. Reminds me of the first shuttle, Enterprise, when they flew it piggyback from California to Florida on a 747. When they touched down in the Cape, half of the heat resistant tiles had come off in flight. Had everybody wondering/laughing how the thing was supposed to survive re-entry, when it couldn't even make it across the country in one piece. Unbeknownst to the public at the time, General Dynamics had put a community good will program into place at the time. They were giving ex-con's jobs, part of a government program, to get them off the street, and reintegrate into society.

Well, these guy's weren't that dumb. Sorta...

They saw an opportunity to "milk the cow" so to speak. They figured that if they only put half the required amount of glue on some of the tiles, they would have a source of job security. Some tiles fall off, shuttle comes back, glue some more on, and so on. Needless to say, after the investigation, the program was killed off.

I'm beginning to think that, that work force must have come from the Ludwig plant