Troubleshooting a Ludwig Phase 2

[stringsthings]

I've got a Phase 2 build going and I've got some formant sounds and fuzz.  Bought the boards from deadendfx.
The problem is I'm not getting any animation.  I can't see any oscillation coming from the LFO section.
I've probed the areas where there's supposed to be a square wave that corresponds to the LFO rate and I'm not seeing anything.

Right now I've got a 2n6027 instead of the 2n2426 and I've tried adjusting Ralt1 without any success.

I've spent several hours going over the notes in the thread and so far, I haven't come up with any solutions.
I'm probably going to get a 2n2426 from smallbear, but I'm wondering if anyone has had success with the 2n6027.
Is there something I'm missing in this part of the circuit?

Dino? 

[digi2t]

Going back through the debugging thread.

Try flipping the 6027 around (exchange A and C legs).

Personally, I've built a couple with the 2N2426 and haven't had a problem. I've never tried the 2N6027.

EDIT: Try flipping the 6027, exchange the G and C legs!! Leave the A pin alone!!

[stringsthings]

Hey! Really glad to see you, Dino!

Tried flipping the G and C legs but had no success.

I did correct a soldering error with trimmer RF29 and now I get clicking/thumping when animate is on and the rate knob is working.
Problems: - No modulation on the filters
              - Repeat knob has no audible effect
              - Intensity knob has no audible effect
              - Fast/Slow switch has no audible effect

I'm going to go back and do some tuning on the filters to see if that helps.  I seem to recall a post that said if the filters aren't tuned,
then animation won't function.  I did check the voltages on each transistor and they're all within acceptable range.

[digi2t]

If you're getting clicking/thumping, then your PUT is working. If it is working with the PUT legs as per the image in the build doc, leave it as is. It was probably your soldering error that was the issue.

The Fast/Slow switch determines how quickly the LFO will come on. It doesn't have any impact on the audio. To hear it's effect, with the LFO (animation) off, set the switch to fast, and turn on the LFO. Now, turn it off, set it to slow, and turn it on again. You should hear a difference in ramp up speed of the clicking/thumping.

Now that I think about it, I have a couple of pictures of the trimmers, and general resistance starting points for the tuning. I should have included them in the build doc, but I didn't think to do so. An oversight on my part. I'll post them here later this afternoon.

I can't stress enough how daunting this build can be. There are a shitload of components, and it's super easy to plunk the wrong part in the wrong place. It's not a project that should be taken on as a bender. You need to take your time, and do it in small bites. Triple check the section you just did, and then move on. Don't take anything for granted, and always question your work. Figure out a validation method that works for you, and the rewards will be great.

[R.G.]

It's worth remembering that the 2N6027 is a Programmable Unijunction Transistor (PUT) and the 2N2646 is a real UniJunction Transistor. They're different animals that can be coaxed to do the same job.

The PUT was developed because real UJTs suffered from the same disease as JFETs - very wide tolerances. UJTs had to be tightly preselected to be awarded a part number, which amounted to a bucket-sorting of whatever came off the silicon wafer, and even then the "buckets" were fairly broad. PUTs are a variant of SCR that can be contrived to have their gate ramped up by a resistor/capacitor, and then have their anodes dump the cap when the gate ramps high enough. The external difference is that the PUT has to have a slightly different circuit than the UJT to do this.

So the instructions for subbing in a PUT for a UJT have to be slavishly followed. They are not direct subs.

On the other hand, I have actually made both variants work correctly, and that's where the instructions on how to adapt the circuit came from.

[stringsthings]

If you're getting clicking/thumping, then your PUT is working. If it is working with the PUT legs as per the image in the build doc, leave it as is. It was probably your soldering error that was the issue.

The Fast/Slow switch determines how quickly the LFO will come on. It doesn't have any impact on the audio. To hear it's effect, with the LFO (animation) off, set the switch to fast, and turn on the LFO. Now, turn it off, set it to slow, and turn it on again. You should hear a difference in ramp up speed of the clicking/thumping.

Now that I think about it, I have a couple of pictures of the trimmers, and general resistance starting points for the tuning. I should have included them in the build doc, but I didn't think to do so. An oversight on my part. I'll post them here later this afternoon.

I can't stress enough how daunting this build can be. There are a shitload of components, and it's super easy to plunk the wrong part in the wrong place. It's not a project that should be taken on as a bender. You need to take your time, and do it in small bites. Triple check the section you just did, and then move on. Don't take anything for granted, and always question your work. Figure out a validation method that works for you, and the rewards will be great.

That's good to hear that the PUT is working.   I probably made a silly error or two somewhere that I've yet to discover. 
I did have some good success in going back and tweaking the filter trimmers RF31 and R20 in counter mode.  Was able to get a balanced 1.5V-6.7V
with the "pedal" ( 10k pot for now, real pedal on the way ).  So that was very encouraging.

Seeing as how the smallest adjustments to the trimmers are significant, I'm going to swap them out for multi-turn. 

It's worth remembering that the 2N6027 is a Programmable Unijunction Transistor (PUT) and the 2N2646 is a real UniJunction Transistor. They're different animals that can be coaxed to do the same job.

The PUT was developed because real UJTs suffered from the same disease as JFETs - very wide tolerances. UJTs had to be tightly preselected to be awarded a part number, which amounted to a bucket-sorting of whatever came off the silicon wafer, and even then the "buckets" were fairly broad. PUTs are a variant of SCR that can be contrived to have their gate ramped up by a resistor/capacitor, and then have their anodes dump the cap when the gate ramps high enough. The external difference is that the PUT has to have a slightly different circuit than the UJT to do this.

So the instructions for subbing in a PUT for a UJT have to be slavishly followed. They are not direct subs.

On the other hand, I have actually made both variants work correctly, and that's where the instructions on how to adapt the circuit came from.

After researching the threads, I found the tip for how to check the flip-flopping of Q3/Q4 collectors and Q7/Q8.
I was able to get good results with Q3/Q4 although the high voltage was around 17.0V which seems high.
Q7/Q8 didn't work as well.  And the voltages were way off.  I was gettting readings of 40V and 20V.  So that looks like a problem.
Will troubleshoot more in this area.

[stringsthings]

Question:  I've isolated the main board from the filter board and the pots and toggle switches are installed.
I'm getting a reading of about 44V at the 35V point.  Is this because the main board is isolated?

Voltages on Q6 (TIP112 )
C: 46.5
B: 39.4
E: 44.5 ( this is the 35V point )

D3 reads correctly, 36V at the cathode

When I had everything connected, I was getting normal voltages on Q6, so I'm wondering if the high voltages
are because the main board is isolated, or if I've made an error.


I was getting "normal" voltages with everything connected.

[digi2t]

The voltages should be with everything connected. All voltages listed are with everything connected.

Here are the initial trimmers set up. Set the trimmer resistances with the power off;





Then power up, and tweak RF29 to get the 3.8v.

[stringsthings]

I've redone the offboard wiring and have everything hooked up and I'm still getting odd readings on Q6.
Also, R22 ( 100R ) is getting hot.  It's not smoking but it's pretty hot.  I'm measuring a voltage drop of 6V over R22.
I'm wondering if I borked the TIP112.

I'm using the 48V DC supply from Mouser that was recommended and using a 1N4004 to connect to the +48V rail.
I didn't think the rectifier was needed since I wasn't using an AC supply.  Will the rectifier make a difference?

Q6 voltages:
E: 40.6
C: 41.7
B: 37.8

40.6V seems to be significantly higher than the 35V spec.

[stringsthings]

Update:

Animation is working!  After more studying of the debugging/misc thread, I saw that Keppy suggested bypassing R7 (220r ) to ground.
That did the trick for me.  I'm now able to get the signature yoy-yoy sounds and others like in the demo vids.
I've still got some tweaking to do on the filters since I swapped out the trimmers on the main board for multi-turns.
I may do the same for the filter board trimmers.  Especially R29.
I'll check my results against the demo videos.

Right now, I'm messing with the zener diode that goes between R64/39/40 to ground.
I've got a 12V in there now and I'll probably try a 13V or 15V on my next parts order.

The voltages on Q6 are still about 5V over the +35V spec.  I'll swap in a BD679 at a later date.

[Keppy]

Sorry I didn't see this earlier. I'm glad past me was of some use!

I'm using the 48V DC supply from Mouser that was recommended and using a 1N4004 to connect to the +48V rail.
I didn't think the rectifier was needed since I wasn't using an AC supply.  Will the rectifier make a difference?

No, it should be fine.

Q6 voltages:
E: 40.6
C: 41.7
B: 37.8

This is impossible with a working NPN transistor. The base should be 1.2ish volts higher than the emitter. If I had to guess, it looks like you've got the collector and base swapped. That would explain the voltages, as well as the excess current across R22. If I remember right, the fabbed boards for this build had two sets of pads for Q6 to make it easy to use both ECB and EBC pinouts, so hopefully you just ended up on the wrong pads. Fingers crossed it's that easy!

[stringsthings]

I think I damaged the TIP112.  When I first starting troubleshooting, I had messy wires all over the place
and I accidentally shorted the power from the wall wart.  I heard a pop but the circuit was still working so I didn't
double-check the TIP112 voltages.  Later, when I re-tested all the transistors, I saw the high voltages.
I was lucky that more transistors weren't damaged.

I've got a BD679 on the way. 

[digi2t]

I think I damaged the TIP112.  When I first starting troubleshooting, I had messy wires all over the place
and I accidentally shorted the power from the wall wart.  I heard a pop but the circuit was still working so I didn't
double-check the TIP112 voltages.  Later, when I re-tested all the transistors, I saw the high voltages.
I was lucky that more transistors weren't damaged.

I've got a BD679 on the way. 

Well, that definitely puts a different slant on things. Let us know how things pan out when you get the new power transistor in there.

[stringsthings]

Got tired of waiting for fedex to deliver the supposedly overnight delivery from arrow.
Luckily the local electronics supply store had an NTE261 ( sub for TIP112 ).
Plopped the new one in and the voltages returned to a normal level.

Q6 approximate voltages:  ( rises very slowly until it levels out )
E: 34.3
B: 35.5
C: 42.8

The 100r resistor is still getting warm so I might put a 1/2 watt in there for reliability.

All the functions are currently working on the build.  All 3 filter modes, fuzz, animation for the filters and fuzz/percussion.
All main pots are functioning like they're supposed to.  So now I'm going to tweak the trimmers to get the filters set up,
minimize oscillations at the 2 extremes of the treadle.  I've got a few suggestions for the build pdf if you're up for it, R.G.
Keppy pointed out some important tweaks, but I don't think they made it into the most recent pdf.

Dino, I don't think the build pdf that the deadendfx links to, has the info on the optional RF26 trimmer. I just noticed that
while checking geofex.  I haven't tried a trimmer but I don't hear any excess distortion.

This thing is quite the beast.  Like a wah-wah on steriods.  The filter sounds can be quite vocal sounding.  Probably the most like the human voice
that I've heard out of a pedal.  And to think that they did it back then with just a bunch of NPN's and PNP's.  I was wondering if there's anyway to
tweak the resonance of the filters.  This would be handy to tame the rolloff.  A kinder, gentler Phase II.

[digi2t]

Dino, I don't think the build pdf that the deadendfx links to, has the info on the optional RF26 trimmer. I just noticed that
while checking geofex.  I haven't tried a trimmer but I don't hear any excess distortion.

This thing is quite the beast.  Like a wah-wah on steriods.  The filter sounds can be quite vocal sounding.  Probably the most like the human voice
that I've heard out of a pedal.  And to think that they did it back then with just a bunch of NPN's and PNP's.  I was wondering if there's anyway to
tweak the resonance of the filters.  This would be handy to tame the rolloff.  A kinder, gentler Phase II.

You are correct, that was the final version of the build doc, and some of the tweaks, like the trimmer, were implemented after the fact. While I can't go editing what R.G. wrote, I can however add an addendum to the build doc. I hadn't dabbled with the Luddy in a while, so some of this stuff is getting dusted off in my brain. If you can PM me a point list of what you think should be added, I would appreciate it. I'll add them to the end of the build doc. I'll also add the primary tuning settings that I posted earlier.

Yes.... it is a beast. I personally never minded the... ferocity? of the filter section, it is what makes it so special. Looking over the schematic, if you wish to attenuate the filters a bit, my test point would be resistor RF+2 (47K). This resistor goes to ground, post the amplifier section that boosts the filter section. Maybe try reducing the value here, which should bring down the overall output of the  filter section volume. As for reducing the actual resonance, I'm hoping someone smarter will chirp in here, because that is (to quote the Great Photonic One) above my pay grade.

[stringsthings]

The filter circuit is pretty complex; beyond my understanding for sure.  If there is a way to reduce the resonance, I'm wondering
if the fuzz sound would be a bit stronger.  I think the filters have such a pronounced effect on the fuzz signal, that you lose some
of the "good" parts of the fuzz. 

I might look into a way to bypass the filters. 

I agree that one of the special parts of the whole circuit is that "on the edge of oscillation" sound.  It's quite characteristic.

My build is out of the box and all sprawled out ( made it easier to do the offboard wiring ) When I get it tuned up and boxed up, I'll post some pics.
Many thanks Dino, and the rest of gang, for bringing this circuit to the light of day.  It's really quite a feat of engineering for 1971.
A quick peek at reverb.com and I see an original selling for $2500(!).  That's just crazy.  Supply and demand at work.

[R.G.]

Shoot, there was an original at the Dallas Guitar Show in may for $600 asking. The guy at the booth let me poke around in it to see how it went together.

[stringsthings]

Experiments with the raw fuzz sound:  ( pinkjimiphoton, you might like this )

Took the signal from the junction of QF5 collector and QF6 collector and brought it straight to the breadboard to see how it sounds.
Super fuzzy but had too much extra noise.  Tacked on a simple filter.  Borrowed the basic idea from a univox superfuzz mod that Mark Hammer did.




Nice strong fuzz that you can control with the fuzz/voice switch and the fuzz mix pot.  Now, the idea is to take the unused 3rd pole of the fuzz footswitch
to select between the dry out and the fuzz out to go to the stereo out jack.  That way it won't mess with the main out of the circuit and I don't have to drill
any new holes for another jack. 

That fuzz circuit is pretty unique in the way that the fuzz mix pot sends a voltage to Q21 to control the size of the square wave.
I've not seen that before.  Thanks R.G. for the circuit explanations.  The engineers at Ludwig didn't exactly design a simple fuzzbox.
( and that's why I've spent a couple of hours so that it sounds just like a fuzzbox!  )

[Keppy]

Glad you got it working!

The 100r resistor is still getting warm so I might put a 1/2 watt in there for reliability.

Yeah, this thing draws significant current, and all of it goes through that resistor and Q6, so those components get warm. It's normal.

[pinkjimiphoton]

A quick peek at reverb.com and I see an original selling for $2500(!).  That's just crazy.  Supply and demand at work.

my original one, with replaced switches, and re-fabbed bottom plate etc sold on ebay for 4000$ plus 100 shipping literally the day i listed it. insane! that was... jesus... years ago now. analog tom actually mentioned it in one of his PG columns. he had no idea it was my one.