KMD Phaser (small stone OTA style) broken. Not sure where to start...

[Earthscum]

I don't have time right now, but I figured I'd throw this out for 'ponderment' today.

I have a KMD Phaser that is basically a Small Stone OTA (LM13700) type. It crapped on me in the middle of a show... basically, I thought it was a jumper cable, so I switched them out, and finally figured out it was the Phaser.

So, here's how it's gone:
No sound, even when bypassed. I crack the case, check solder joints, connections, grounds... find that I cut the battery (neg) lead when I closed it (it's in Ibanez/Boss type enclosure). Stupid me. I fixed it.

Plug it in and it works... then it craps on me on the second song of our set at the NEXT show (of course, it works fine until I get it on a stage... ML! (murphy's law)). Same thing, all signal is dropped, but at least I got bypassed signal.

So I pick it up last night and hugged it and asked it to work and petted it and bought it candy (j/k... I gotta buy my woman candy first!). Then I plugged it in, and still nothing.
I got the dimmest indicator light, hardly on. I had no LFO, I thought (duh... it's a Phaser, not a chorus... I wasn't waiting for a full sweep).

Well, fiddling with the knobs, basically I discovered that all it does is sweeps volume up and down. I have a REALLY slow volume vibrato.

I'm guessing that I shorted something when I pinched the lead and now I have a cap that has blown, ala the crybaby volume pedal mod. I'll dig in to it tonight, probably, if I can keep away from my other project. Hopefully I haven't blown a chip. I really don't feel like pulling out a factory DIP, ya know what I mean? it's a snug fit to get to the leads. I guess if it goes that route, I'm cloning the damned thing and swapping the entire board, lol!

Anyways, any insight is helpful. I assume this isn't some new phenomenon.  Thanks!

[R.G.]

Pretty much the answer to "what's wrong with my pedal? It doesn't work." is contained in the thread "Debugging. What to do when it doesn't work." That is nearly always where to start. That's why it's a sticky note at the top of the forum where it has to be skipped over to get to the rest of the stuff.

[Earthscum]

Not to be rude, but, umm... yes, I've read that a couple times.

I figured if someone had this problem already, and answered me before I got around to fixing it, it would save me a BUNCH of time, and that's all I wanted. This is a KMD, not one I built, or I probably woulda had it fixed by now.   

Like I said, I already gave this the once-over... all traces are good, all solders were good... I did redo a couple wire ends since they were fraying a bit and a strand or 2 were broken. No obvious capacitor blowouts. No blackened resistors, nothing that looks remotely like it's fried.

I remember reading somewhere about something to do with the small stone not phasing, however I'm not even 100% sure this IS a ST clone... it has equivalent chips in it, but I haven't fully traced it out yet, and lost what I had started (the ONE page in my books I lost... as far as I know. ML I guess).

I guess I start with the caps around the OTA, unless someone thinks it's bad. Check back later. 

[R.G.]

Not to be rude, but, umm... yes, I've read that a couple times.

I figured if someone had this problem already, and answered me before I got around to fixing it, it would save me a BUNCH of time, and that's all I wanted. This is a KMD, not one I built, or I probably woulda had it fixed by now.   

And likewise, I wasn't being snotty. The technique of figuring out whether the DC conditions are correct is deadly effective, since most conditions which make something simply not work at all show up as the DC conditions being incorrect. There are some AC-only failures, but the ones which are purely AC failures and do not affect the DC conditions are the minority.

In many cases, it's quicker to read all the DC conditions than to wait for a response to show up on a forum if you have any reasonable idea what the DC conditions should be. Sorting wheat from chaff is relatively quick if you first see if all the ICs have power (if it has ICs) and all the transistors have reasonable conditions as noted in the debugging thread. Often that tells you all you need to know. Understanding that voltages have to be mostly in the middle of the power supply to get amplification puts another criteria on that. Nothing can amplify if it doesn't have voltage to do it with. And resistor/capacitor failures upset those.

In any case, good luck.

[Earthscum]

I understand where you're coming from now. That's why I said I was throwing this up here... if someone has a "check this part" while I'm at work, and while I can't work on my pedal, then I've wasted no time waiting for a response. If I get no response, then I figure I'm still at nil, just like I was the entire day at work anyways. Honestly, I read and I search. I've looked for over 3 years for schematics or any info on this pedal, and I think Mark was the one who told me in brief about KMD's history of basically putting out clones. So far I think that was the most information I've got on it the entire 21+ years I've owned this thing, lol.

So, yeah... I'm not really waiting for a response... more like procrastinating... keep getting sidetracked by the circuit on my breadboard.

Oh yeah... anyone know where to get a pinkish-red push on knobs? I lost one years ago... also can't find the screw-in hinge pins. Have a feeling I'ma gonna have to make those myself    None of the typical sources have quite the right knobs... not sure if someone has a link for some kind of knob superstore or something (that would be kinda kewl).

[Earthscum]

Well, I'm at a loss. I have (cable plugged in)  0.8MOhm resistance between the positive and negative battery leads, no circuit resistance to the traces from the battery leads. Fresh battery gets sucked down to about 6.5V across the rails and steadily drops.

Positive voltage hits the board and goes across a diode to the V+ rail. From this rail there is a zener, 2x 51k, 2x 1M, 100uF, 78L05, V+ on the LM13600's (2), and one of the 2 dual op amps. The 78L05 goes to the oscillator section.

Anyone have any ideas? It sucked a brand new battery down (albeit a cheap enercell) just in the time I was starting to probe for voltages about a full volt, and the thing got really warm... like I left it in the sun, about. The battery recovered, of course, but still... It just does a volume sweep, and I don't even want to plug my power supply into it just because of the possible current from it.

I'm trying to find a schem that is at least similar to this, but I'm not seeing anything quite yet. When/if I find one I guess I'll link it.

And decided to give it another go on the voltage metering before I hit post... battery didn't warm up this time (that good), but I figured out that every time the LFO sweeps, the voltage drops, then stabilizes, then drops, then stabilizes. It doesn't go up, just keeps dropping. I think at about 4.5V is when it must've started getting warm last round. This time I only monitored it down to about 5.6V.

That enough clues to help anyone point me in the right direction?

[R.G.]

Once DC conditions have told you where to look - in this case, something about the LFO or what it drives - and that it's an excessive current condition, you're ready to start isolating.

Overcurrent conditions in the power supply require a divide-and-conquer approach. And this one will require you to do at least some digging out of the schematic. What you have to do now is isolate what is pulling too much current by finding what causes the current drain to stop. Voltage problems are easier than current problems in this regard because you usually have to open up circuit paths to deal with overcurrents. Either that or make more-or-less enlightened easter-egging forays.

I personally would pull out that 78L05 and see if the drain continues. It's about the point that you've eaten up your third or fourth battery that you realize you're sooo ready for a current-limited power supply so you quit buying a $2.00 battery every couple of minutes. I was lucky enough to find an old HP 6214A power supply for cheap. This gives me 0-12Vdc, and 0-1.2A current limits, both independently settable and independently metered. It's probably worth the time of anyone who does much debugging to make a variable current limited power supply for situations like this. It is possible to make a variable current-limit clamp for low currents with two NPN transistors and three resistors, and I can tell you how to do that if you're interested. It ruins voltage regulation, but that's not the big issue at the moment with this pedal. Sorry, I'm wandering around in the back-story.

From the complement of ICs, this thing uses the two LM13600s as the phase line; that uses up all four sections. A phaser needs two amplifiers, one a buffer/driver in the front and another for mixer/output at the end. That's one of the two dual opamps. The other dual is almost certainly a two-opamp integrate-Schmitt LFO. This has to drive the Iabc inputs of the OTAs somehow. Probably through resistors from the integrator section of the LFO. Also, the failure was sudden. That's different from slow death in a closet. It's pretty sure that the 2x51K and 100uF are a bias supply, maybe the 1Ms too; Either the zener or more likely the 78L05 power the dual opamp that's not the signal opamp.

My first shot once I figured out that a board is eating enough current to make the battery warm is to try to figure out what else is getting hot. If nothing is obviously burned, it's probably safe to lay your hand on the components, plug in a battery, and see if one part gets hot. If it does, that tells you very closely where to look next. A dead short will not get hot, of course; that forces all the power to be dissipated in the power supply, which is a possibility, but the burning-palm trick is worth trying. It's fast. One thing you might consider is plugging in your power adapter and letting it cook. Whatever cooks is already damaged, and chain failures are unlikely in pedals. This is not a good idea with a solid state power amplifier, but in pedals, the part(s) that get hot are probably not going to kill their brothers. If something starts getting too hot, you've found a suspect.

If there's no joy there, a little judicious easter-egging is in order. Pull the 78L05. Still too much current? Look up the datasheet for the LM13600 and find the Iabc pins. Then open them up by pulling up one leg of any series resistors to them from the LFO. Check current. Still too much?

At this point you have to decide whether you want to get smart or get mad.  They are both effective, and it's a toss-up as to which is better. "Getting smart" means tracing out the circuit, understanding how it works, and then using measurements and logic to figure out what is wrong, then replacing part or parts.

"Getting mad" means realizing that (a) time is money and (b) parts are cheap. If you're choosing the "getting mad" approach, order a fresh complement of every single semiconductor on the board. While they're coming, look at geofex, FX skills, "removing a DIP IC". http://geofex.com/Article_Folders/remove_IC/remvdip.htm and remove all the ICs, diodes, etc. This is not traumatic if you realize that removing one pin from a PCB pad is trivially easy, and that the hole is then easily opened with a thin splint of wood. When the order comes in, put parts back in *one at a time* checking power drain between each part. If you get to all parts in, then one of the parts you removed was bad and you now have a working pedal. If you get to too much power, then one of the passive parts involved with the last part you put in is bad too, and you now know where to look.

Remanufacturing can make sense when 90% or more of the parts cost in a box is in the enclosure, and trivially little in the parts that actually do the work.

Getting smart is good for your brain, getting mad gets a working pedal faster in many cases. It's a trade off, and a judgement call. Experience is what makes that call easier; unfortunately, experience is what you have left when you've forgotten her name.

[Earthscum]

I''m on lunch right now and doing photo shoot with my girlfriend after work, so won't get to it til tonight...

The 51k's (should've mentioned) go to a pair of transistors that are part of the LFO section (that may help on topology). The 1M's seem to be simple pull-up resistors, without looking up the actual transistor and pinouts, but yeah... the 1M's are still in the LFO/Transistor section.

I was contemplating yanking the VRegulator out for a good part of the reason you mentioned. I was going to pull out the 2 filtering caps before that to at least test them (probably just replace since I have some on hand). Thanks a bunch for the rundown. I'll still start with the filtering caps since it's easy and cheap insurance, if I still have the problem (most likely), I'll start running through your steps.

I've had pictures overlayed forever now for tracing the PCB and getting a schem from that, just haven't got around to it with all the band stuff going on this summer. I figured out a neat trick is to copy the block layout or pinout of IC's and paste them over the top of the actual IC in the picture for reference... saves a lot of time jumping back and forth, especially if you find one like your typical OP amp diagrams that has the block inside the IC layout.

Oh yeah, I couldn't feel any heat (haven't actually probed temps tho) from any components, but my battery was getting pretty warm. Discolored components were the first thing on my list to check. Usually the board seems to show it first when it's resistors, but nothing apparent on this board. The resistors all look fine as well. I'll definitely keep my eye open though, thanks again.