Ludwig Phase II Clone Tech Notes

[Keppy]

keppy, does the fuzz repeat tick?

Nope.

[R.G.]

heck, rg, couldn't kep just use the fuzz repeat to test with? i mean, that thing should be clicking...or "ticking" loud i'd imagine.

keppy, does the fuzz repeat tick? if it does, does the speed of the ticking vary with the animation rate?

If it were ticking, sure, that would do it. My first suspicion on the LFO was that it was latching on and never unlatching, so there would be no ticking. We haven't yet seen anything to prove or disprove that.

as for an audio probe, REAL dumb question, but what about the jack darr method of just using a finger (with the other hand in your pocket) to see if a buzz happens?
sorry for the dumb questions, but again, i am but an egg here...

No, it's not a dumb question. I'm suggesting that he deliberately misuse the audio probe technique to listen to the LFO. Normally that wouldn't make any sense, but this LFO does periodic ticks that would come across on an audio probe. A finger on the signal lines would make the output of the pedal buzz, but wouldn't do anything to the LFO. I'm really trying to figure a way to do what I'd normally do with an oscilloscope; maybe we can listen to the LFO and hear bits of it operating or not. It will at least tell us something about whether there is anything there or not. Your thinking is not off at all - just that I'm trying to do something that's way nonstandard.

rg, can you figure out enough to tell me WHAT stuff would be involved to figure out voltage readings etc that may help?
i'm willing to read every connection on the board if that's what it takes.

Actually, this maybe ought to be called the Pink Jimi clone.    Without your pushing, to get it started, I'm suspecting it wouldn't be here. You've been very willing. Right now, I think the issue is how to get a quirky, unusual part to function by internet debug.

I may have to order myself a 2N2646 to tinker with to provide sane advice to Keppy. Alternatively, I have breadboarded and got running two different substitutes for the UJT, one a 2N6027 PUT and the other a connection of a PNP and NPN transistor that acts very much the same. I'm holding off on that till we can focus on the UJT in Keppy's board for a while. That's the short term goal.

On the other hand, we now know that the PCB layout produces working formant action, subject to some tinkering and maybe more work on the UJT thing. We're nearing the point where we'll have to figure out when it's good enough to go print a few. I know a place that owes me some PCB favors, and based on the area of the PCB, I can get 15 sets of the PCBs (fall and console bds) for about $250 plus shipping, etc. Call it under $20 a set. If you want, one of you can set up a group buy and we'll get them. I'll sift in any changes needed first. Keppy has been very good at making suggestions back into the PCB and BOM. Huge help. I think the last changes I did were to put in several options for different sized trimmers. We may want to put in the 2N6027 PUT based on what we find from the LFO stuff.

That's the longer term goal, I think.

[R.G.]

keppy, does the fuzz repeat tick?

Nope.

I think the next thing to do is to run the repeat trimmer R4 on the console board *up* to maximum resistance. I think I told you to run it down to minimum before. That was probably not a good idea. Or just set it dead middle.  Then mess with the animation rate pot while audio probing for ticks on B2. If nothing happens at any setting of the rate pot, then it's likely a bad component, solder joint, or UJT.

JImi, Digi2t, what are the fastest and slowest rates on the animation speed? 1/second? 10/second?

[Keppy]

I tried the audio probe and got no ticking. I took some resistance measurements of the UJT in the circuit, power off, unit grounded through the amp. Weird things happening here. Interbase resistance is about 4.5k. Resistance from emitter to either base is about 13k, but it takes awhile to settle there. I saw the meter start at 1M+ and travel all the way to 13M. I also saw the meter start high and drop to 13M. Again, this is without power to the circuit.

[Keppy]

If the flipflop is working, there will always be 15-16V between the two collectors, because one will be at 15-16V, and the other will be at nearly ground; they just change positions too fast to see with a meter. Both positions produce the same voltage, but alternating polarities. Here's a thought - measure the difference between the two collectors on an AC scale. This puts a capacitor in series inside the meter to keep the DC out. If you get a significant reading, then they're flopping but too fast to read on DC. Works sometimes.

I finally got around to doing this. There's no significant AC between points 25 & 26. However, measuring DC, pin 26 now reads the supply voltage, while pin 25 reads ground. This is a change. Before, I read 15v between the two, but both read .06v to ground. Not sure why that changed.

[pinkjimiphoton]

hi guys...gonna be out for a little bit, so will look at voltages or what ever when i get home.

diggit....

trimmer r77 controls the fuzz repeat intensity, looking from the back, (as on the layout) it should be somewhere between 1/3rd and 3/4 way up...try turning up the fuzz repeat control full, the animation rate about 1/2, the animation intensity full. then f around with that trimmer.

trimmers r 41 and r 62 also seem to have something to do with the ffm....mark where it's working now, and try cranking r62 full and r41 almost off. as i recall there's a sweet spot in there where everything comes together, the phase II is a real juggle of a balancing act.

worst case scenario, i will go beg my old oscilloscope off my old guitarist that i gave it to, and ship the sob to ya keppy!!!
when i stopped building tube amps when my wife got sick i gave all that stuff away, if he still has it (and he's a packrat, so he should) i will try and get it out of him.


i gotta jet for now...great work, all around, thanks guys for believing in this thing...it's like real-life jurassic park, that don't eat people...or suck, for that matter!!

again, i apologize for not reading all the damn voltages and stuff in the first place, i'm realizing now my laziness cost alot of time that could have been saved..

talk soon brothers...peace and blessings to ye..

[R.G.]

I tried the audio probe and got no ticking. I took some resistance measurements of the UJT in the circuit, power off, unit grounded through the amp. Weird things happening here. Interbase resistance is about 4.5k. Resistance from emitter to either base is about 13k, but it takes awhile to settle there. I saw the meter start at 1M+ and travel all the way to 13M. I also saw the meter start high and drop to 13M. Again, this is without power to the circuit.

It's hard to tell, but that *might* be because the UJT is starting in the off condition; the meter resistor measuring circuit puts a current out, then measures the voltage it causes. On capacitors, this charges up the cap. It might be that the meter was charging up the timing cap and then the emitter turning on. Maybe.

Do you have a breadboard? I'm thinking that we may be able to test the UJT in a breadboard circuit.

All the descriptions so far are consistent with the UJT tripping on once, then never un-locking.

[Keppy]

Yeah, I have a breadboard. That's how I do my audio probing, actually. I never bothered with altering a cable; I just run a wire to a cap on the breadboard.

Let me know what kind of test circuit you're thinking. In the meantime, I might order another 2646 just in case it DOES turn out to be bad. I didn't want to do that earlier because I'm moving next week, but I should be able to get it shipped to the new place now.

[R.G.]

Yeah, I have a breadboard. That's how I do my audio probing, actually. I never bothered with altering a cable; I just run a wire to a cap on the breadboard.

Let me know what kind of test circuit you're thinking. In the meantime, I might order another 2646 just in case it DOES turn out to be bad. I didn't want to do that earlier because I'm moving next week, but I should be able to get it shipped to the new place now.

Ordering a new one isn't all that bad an idea. When you do, get a 2N6027 too. The 6027 is only about $0.35, and I'm **sure** I can make that work in this thing. I almost just ditched the 2646 at first.

For a test circuit, I'd try the first schemo here: http://www.allaboutcircuits.com/vol_3/chpt_7/8.html, the one without the transformer. I think it'll work at 9V.

In my (long ago!) experience with UJTs, I learned that they're quirky to get working reliably. Any of the resistors involved can make them not work, and there are both upper and lower limits on the timing resistor for the timing capacitor. If one or more of the resistances wasn't right, it would latch up and not oscillate or never trip and discharge the timing cap.

While you're at it (in your copious spare time, getting ready to move    ) think about what other changes to the PCB might be useful. Right now, you have more experience on that layout than any other living person. 

[digi2t]

sorry guys, i'm out of town right now. i'm texting from my buddys phone. i'll be back sunday night. my bet is on the ujt, can't be that messed up. my fingers are crossed for you guys!

cheers,
dino

[pinkjimiphoton]

k guys, i'm here and will check in periodically if anyone needs any info or anything. just say the word, i should be home all day now.

i did talk to mike, so will have my old oscilloscope back soon. don't know what the heck to DO with it, but... 

[R.G.]

Got hacked off at UJTs, PUTs and all of their ilk.

The CMOS 555 does the job nicely. Can be dropped into the design easily, leaves the flipflop unchanged, and replaces Q1, Q2, and the junk around them. Works with a single 1M timing pot, and has an inhibit pin to serve the function that Q1 did.

Of course, now I want to replace that 'flop with gates, too.   

[Keppy]

Honestly, the layout's pretty good as is for the components that are involved. Everything has worked wonderfully once the components were the right values. I didn't even have to tweak the trimpots to get it to work, although there's a lot of dialing to taste that I haven't gotten around to yet. What I mean by that is I don't know that there's any need to specifically make room for test points, unless we find out what the original ones were for. The components fit just fine.

In fact, I think the only real suggestion I have to the layout is removing all the redundant grounding points, since they're joined on the two boards anyway. I jumpered the console ground to the fall plate ground so I could connect it to the chassis with one wire. Even then, having all those ground points let me choose the most convenient connection, so maybe they stay but with a note that it's not necessary to wire them all up (which anyone building something this complicated should already know!).

Actually, now that I think about it, the socket for the UJT is bumped up against some nearby components. Given my issues, it might be a good idea to make some extra room around that so future builders can socket that particular part.

Anyway that's it for layout suggestions. I have a list of documentation issues that I'll be sending you once we have it all sorted out, R.G. Mostly a list of errors/misreads we found plus some suggestions for the wiring diagram.

Back to work wrapping up bookkeeping for the business I just sold so we could move...   

[pinkjimiphoton]

so....did ya get it working, kep? 

[toneman]

UJTs are easy...PAIA used them in their early vco and (i think) sequencer.

But, I would 1000% agree with R.G. that 555's are the coolest bad. 

So, Keepy, did U build some clone pcbs??    Or, are U working on your original unit??

[Keppy]

Jimi & Dino (digi2t) have originals. I'm building a clone with a layout of R.G.'s. If I get a working unit I believe he intends to post the documents.

[Keppy]

Ordering a new one isn't all that bad an idea. When you do, get a 2N6027 too. The 6027 is only about $0.35, and I'm **sure** I can make that work in this thing. I almost just ditched the 2646 at first.

The 2646 is available at Small Bear. The 6027 is available at Mouser. Anyone know a place that has both?

For a test circuit, I'd try the first schemo here: http://www.allaboutcircuits.com/vol_3/chpt_7/8.html, the one without the transformer. I think it'll work at 9V.

What will that circuit tell me? There's no timing cap, so it doesn't appear to be an oscillator. Also, there are no values on that schem, so I don't know how to set it up for 9v. I'll definitely need more info to use that as any kind of diagnostic, because I don't know what I'd be looking for.

[mrmoo1337]

dino, I think he means this schemo; http://www.allaboutcircuits.com/vol_3/chpt_7/8.html#03506.png 

And Newark.com seems to have both the 2646 and 6027!

[R.G.]

What will that circuit tell me? There's no timing cap, so it doesn't appear to be an oscillator. Also, there are no values on that schem, so I don't know how to set it up for 9v. I'll definitely need more info to use that as any kind of diagnostic, because I don't know what I'd be looking for.

Sorry. I wasn't clear. Scroll further down, below the first graph of peak point/valley point, to the two circuits with waveforms between them.

The left hand circuit has a 100K and 10nF timing circuit, RB2 is 470 ohms and Rb1 is 47 ohms. The circuit on the right has a transformer in it.The left circuit is what you want.

You'd be putting the signal into an amp and listening for a (very loud) oscillation.

[Keppy]

Gotcha. First, though, can you tell me if I'm missing something in the wiring diagram? the way I read it, the animation footswitch, when off, grounds the animation speed switch. When on, it lifts the ground. How can the animation speed switch function if it is just switching a wire with no connection between two points? What am I missing? Do I have the right version of the wiring diagram? Here's the one I have:

(sorry about the size)