Ludwig Phase II Clone Tech Notes

[R.G.]

You have in a nutshell why the Ludwig P2 died. No matter how good the inspiration might have been, the resulting design (which by definition means design-for-manufacturing, or lack thereof) was too fragile to live in the real world. "Too fragile" is a serious flaw in design.

(Competent) engineers know that there is a hierarchy of cost associated with fixing mistakes. If a design has no mistakes - meaning *anything* which the customer perceives as a flaw - the cost is zero. There are no designs like this, because there are *customers* with flaws who will find fault with even an otherwise-perfect design. But if a flaw/mistake can be found and corrected in the design phase, when no actual products are being built, the cost is, say 1X at that time. 

Correcting flaws when the product is being prototyped means fixing prototypes, correcting orders for parts, reworking, retesting, etc. and so the cost is 10X for each one found. If a problem escapes and is caught during manufacturing, the cost escalates again, to about 100X because now every unit has to be manually reworked while incorrect stock and parts have to be scrapped. If the flaw is only found by the customer, the cost to fix is easily 1000X, and that's only if the customers are not lawyers.

(Competent) engineers know that serious errors in products in the field often cost them their jobs, either directly by angry management, or indirectly by killing off the product and its follow ons or bankrupting the company.

[pinkjimiphoton]

hey dino, i would still seriously advise replacing ALL the electros in it...i mean, 40 year old caps are gonna fail catestrophically eventually.

[digi2t]

Yup, electros and the tropical fish as well. It's on my "honey-do" list.

I'm working all weekend, but I'll check the values vis-a-vis the diagram this week. Once I'm sure everything matches up, I'm going to start ordering any caps I don't already have. All the electros for sure, I only have radials in stock.

I wonder if green chicklets would be alright here? I've got lots of those. They should be fine in place of the tropicals. I'll have to double check the specs.

[R.G.]

Tropical fish caps are metalized mylar. The bands were how they did the value coding, like resistor bands.

Any metalized mylar with a voltage rating over the working voltage should be find. I think all chicklets have voltages over 50V, so go with them.

The magic mojo of tropical fish caps is... um, advertising.

[Keppy]

I finally got around to playing with the trimmers. There's a lot of weirdness in here, but a lot of the settings make one thing sound cool and another sound unuseable. I found that the easiest way for me was to go (sort of) in the order posted by R.G. here:

Keppy,
Got the filter fired back on the simulator, put the circuit mods in. None of the changes made much of a difference either way so far, at least as regards getting a filter sound.

What *was* critical was the setting of R29. Unless this is just right, you get no filtering. So pick a filter, set up points 8 and 9 on the filter board to be about 2.5V, then tweak R29 til you get a filter sound. If that doesn't work, I can read values off the sim for DC voltages that should get you closer. I suspect that as critical as the setting is in simulation, that may be one of the biggest remaining issues.

I can come up with rationales for the other trimmers.
R31 sets the amount of DC on the rocker pedal so it moves Q5 on the console the right amount to make the control voltages for the two filters.
R41 and R62 have to do with either the sensitivity/idle frequency/balance of the two filters as respects the voltage made on R29. They are much less touchy than R29.
R55 and R77 are a balance between the two halves of the filter sides; don't know exactly what they do yet, but again, much less touchy.
R65 seems to be an output level control for the high formant filter. It won't keep it from filtering.

r.g., tell me what to do, i've gotta couple hours free...no idea what to check or where...but if ya tell me say "check at both sides of r56 to ground with a voltmeter" i can probably handle that without screwing it up too badly.

Actually, you've been immensely helpful in running stuff down. What would be very useful is to clip your voltmeter black lead to ground (wires 13 and 14 on the bottom board are handy), then measure the voltage with the red probe on
- all the transistor collector pins; collectors only should be enough. The schemo shows you where the collectors are.
- voltage on the wipers of the trimpots.
- voltages at wire/connections 8, 9, and any one of the places labeled with the funny greek letter in the circle on the schematic. The schemo also shows you where the pick off points for it are.

Given how critical that voltage is, I maybe should put in a test point on the PCB for setting it.

My modified version:
*Note: be careful with your volume, as some of these adjustments (particularly R65 & R77) can cause loud oscillations.

1) Start with R65 all the way counterclockwise and all others at halfway.
2) Turn on Fuzz RPT and adjust R4 (console) for desired range on the Animation Rate pot.
3) Turn off Animation. Have a friend (my wife, in my case  ) play guitar and work the treadle.
4) In Vowel mode, adjust R29 until you hear "yoyoyoyo" as the treadle is moved.
5) Switch to Counter mode, which is more extreme and harder to tame. Adjust R20 (console) to taste. You may wish to recheck this adjustment in vowel mode.
6) Repeat step 5 for R31
7) In Vowel mode, adjust R41 & R62 to intensify the "yoyoyoyo" effect. Switch to Counter mode and readjust R41 & R62 to tame the effect if necessary.
Repeat step 7 for R55 & R77.

Since these adjustments affect one another, you may find yourself jumping back a few steps to refine an adjustment made earlier.

9) After making all other adjustments, set the animation switch to Both, turn the Intens. Fuzz Rpt knob all the way up and begin adjusting R65 while the guitar continues to play. This adjusts the volume of the Fuzz Rpt effect.

Consider this open for comments & modifications.

[Keppy]

Apparently typing "8" followed by ")" gets you

[Keppy]

[digi2t]

Apparently typing "8" followed by ")" gets you

What do you type for DROOOOOOLLLLLLLLLLL!!!

Amazing job Kep! Congratulations! I'm looking forward to a sound byte, or video, when your shit storm settles down.

[Keppy]

Thanks Dino!

Forgot to say, R.G., that you're free to grab those photos for your docs if you still want to. I can also email you higher res if you need them.

[pinkjimiphoton]

wood....

damn, i think i gotta visit the bathroom ....real quick...

[Brymus]

Absolutely beautiful Kep !!!!!!!! 

[Keppy]

[pinkjimiphoton]

aha! well done, and nice demo...but trimmer r61 (i think it is...right corner area of fallboard ) is adjusted wrong...you shouldn't get those oscillations when ya move the treadle. r41 (i don't have the schem in front of me) will also do that. you may wanna back one then the other off just til it stops...then it will be calibrated right.
but sounds and looks great, keppy!!

we done, mate, well done!!

[R.G.]

A beautiful design job! I will use the photos if you don't mind.

A few other questions: what is the size of box you have it in, and what are the relative spacings on the controls?

It's so good that I think I may do an alternate console layout for people who want to eliminate a lot of the wires.

[pinkjimiphoton]

it is indeed beautiful...once he backs off that trimmer, he's got the ludwig nailed!!!!!!!!!! sounds great,. looks even better!

[Keppy]

Yeah, it was tough to dial in because the oscillations happen mostly just in counter mode, but the adjustments all affect multiple modes. I'll give it a shot, though.

R.G., should I just send you the etch file to measure? The original's in Illustrator, but I can save it in a different format.

[R.G.]

R.G., should I just send you the etch file to measure? The original's in Illustrator, but I can save it in a different format.

I'd love it if you could. I use corel draw, so if you can export to something like PDF or whatever.

I've been staring at the circuit for ways to get rid of that mess of wires, at least part of them.

[Keppy]

Cool. I used a 1590D. Forgot to say that earlier.

[R.G.]

Cool. I used a 1590D. Forgot to say that earlier.

That's what I'd have guessed, but asking someone who knows is more certain. 

Do you think the Lo-Z balance and "Bypass Balance" knobs need to be front  panel controls - or even there? I mean, they're interesting for historical accuracy, but may not be all that useful for a modern pedalboard. I can just hear the posts for "I build a Ludwig phaser clone, and it doesn't phase. And is there some way to wire it for true bypass, 'cause I hear that's always better."

[Keppy]

Dino & Jimi talked me into keeping those. I was afraid the effect level would be quieter than the dry signal without the bypass balance pot to make them match. Jimi & Dino also thought the two inputs were cool. Personally, I could see using the second to turn it into an interesting talkbox, but I haven't tried it yet. I say make the second input optional, and don't make a true bypass version until one of us confirms that the levels will match. Honestly though, even in that case I'd leave the build as is and just include instructions for the true bypass mod. A jumper on the board should leave it free for standard off-board bypass wiring.