so i smoked my univibe last nite......

[pinkjimiphoton]

i mean roasted!! there's a 4.7k resistor that lit right up, and it smoked one of the q's in the lamp driver... i mean a small fire that completely melted the transistor!!
i figured oh shit, this one's done.  but decided to see if i could save it. i replaced the melted stuff, and all the components in the general part of the circuit after carefully desoldering and cleaning up all the residue with some denatured alcohol and a q tip. also melted one of the two 10u caps on the board, so replaced them as well as a bunch of the other electros... basically anything where the voltage was weird.
the lamp survived... when i first plug it in, it will flash once and i can hear the phaser sweep... but thats it.
i checked voltages, and the only thing that seems weird is q1, where the c is reading about a volt and a half at the junction with the 1m resistor and b of q2... i expected it to be higher, as everything else is 11 or 15 volts.
its kinda distorted, particularly in chorus mode, so am assuming i gotta bad component somewhere in the audio path, gonna audio probe it and figure that part out tomorrow when i get a day off.
but the oscillator part i can't seem to get going, i replaced everything in it, and the voltages seem ok... am i maybe missing something? a little advice would go far..

i will of course post voltages etc when i go thru it tomorrow... but reading rg's articles, i'm wondering if perhaps for the lamp drivers i should go with 3904's instead of the 5088's i used?

i built it from this kit

https://guitar-electronics.eu/en_US/p/file/a39c59d03eaa77c3632f11e4769a851a/univibe.pdf

[PRR]

> weird is q1, where the c is reading about a volt and a half at the junction with the 1m resistor and b of q2...

B of Q2 "must" be low to leave room for Q2 C to swing.

And there's no 4.7K to smoke in this area.

The Lamp Driver is Q13 shown as BD681. Q11 Q12 sure can be '5088; indeed may have to be a high-hFE device.
WHICH 4.7K did you smoke? That's gotta be A Clue, and might "go far".

Were you poking at it? There's little stress on most of the circuit unless random mis-connections happen.

[pinkjimiphoton]

> weird is q1, where the c is reading about a volt and a half at the junction with the 1m resistor and b of q2...

B of Q2 "must" be low to leave room for Q2 C to swing.

thanks paul, that would make sense. but i'm getting a lot of rfi and noise right at the q1c q2b 1.2m resistor on the low side. should i suspect that resistor is bad? as soon as i take a voltage measurement there, it freaks out with noise, but only in this one spot.
a couple of the electros <1u> were also being weird.. if ya touched the top of the cans, ya gotta buzz like touching the end of an unplugged guitar cord, too. i thought they may have been shorted, so i replaced them.
do you think i should just replace 'em all with mono ceramic ones? i have enough on hand to do that if it may help.
still gotta trouble shoot the audio path. gig tonite tho, so gotta wait til tomorrow.

And there's no 4.7K to smoke in this area.

the 4.k resistor is R49, the one going to ground off the q12e/c22 junction. q12 is what lit up, and the resistor. i replaced q11 and 12, and all the electros associated with that part of the circuit just to be safe. c19-c23 all replaced. i checked for voltage before soldering the q's in, looked good so i went for it. i gotta schematic i can post below to make stuff a little easier.
i DID do RG's reccomendation also for r3 and went with a 4.7m resistor instead of the 47k specc'd, HUGE difference, it goes well above unity gain now. before it was easily 6db below. that is the only mod to the circuit i made. well, that and the speed controller pot. i thought i could get away with tapering resistors and a dual 100kb i had, a nice 2 watt one cuz i mounted it in an old dearmond volume pedal... figured since its clocked 180 degrees off, i should be able to get it to work. yeah... heel down over about 1/10th of the sweep lol... more on that in a minute...

The Lamp Driver is Q13 shown as BD681. Q11 Q12 sure can be '5088; indeed may have to be a high-hFE device.
WHICH 4.7K did you smoke? That's gotta be A Clue, and might "go far".

that one in fact. i had to have the 200r trim pegged to even get the lamp to light up, should i maybe try a darlington in there for more gain? i have more 5088's, but in RG's neovibe project he reccomends 3904's for their ability to handle the current.

Were you poking at it? There's little stress on most of the circuit unless random mis-connections happen.

yeah, as a matter of fact i was playing with tapering resistors... started around 200 ohm and had it up to 33k trying to improve the sweep. seemed like it wanted a much smaller pot, so i tried dividing the pot with two 100k resistors between input and output of the pot on each side, i was tacking the second one down when it lit up. my own stupidity at 4 am i know and i deserve the lumps for it.
i was kinda surprised i couldn't get it to oscillate if i reversed the outside lugs of the pot. only one way would work, and only fully heel down would the oscillation start. wondering if maybe i screwed up by true bypassing it? when i click the bypass switch, it seems to start the oscillation. its very weird. i figured about 20% of the pot value is 20k, so 22k i thought would work for the tapering, but no dice.  i'll tackle that once i figure out what i may have smoked.
for reference, here's the best capture i could get on my crummy machine for the schematic, screen shot from the pdf

[R.G.]

The LFO in the univibe circuit is an odd one, and is fussy about the values of the pot. If the total resistance of the pots is too big or too small, it will either start and die out or not run at all.

I found this out by subbing in a center tapped LDR for the pot, and changing speed by varying the current to an LED illuminating the LDR - that is, I used a five-lead Vactrol unit for it. This worked well, excepting that it illustrated very well that LDRs light adapt, or I guess dark adapt. When I turned the speed down ( high resistance on the LDRs, = low current/light on the LED), it would go down to verrrrrryyyy slow, pleasingly so. Then when I had it set about right, a minute or two later it would quit. A little sleuthing showed the LDRs dark adapting to higher resistance and the LFO running out of gain and the oscillation dying out. So I tinkered with speed resistors a while.

What seems to be happening is that the LFO speed resistors may not be supplying enough signal >current< back to the darlington pair running the oscillator. That was the genesis of un-loading that base by making the resistor be much higher. And on perhaps changing to a JFET for the first transistor, making a JFET-bipolar follower. That seemed to give a lot of range and be steadier with bigger bias resistors. A purpose-made darlington like an MPSA13 or MPSA14 helps as well, as their base input impedance is higher than a two-5088 discrete version. A JFET/PNP compound follower would do much the same if you set its gain to unity by tying the JFET source to the PNP collector. This gives a follower with a high input impedance from the JFET, as with the JFET/NPN follower. RFI ... might be a problem, and you might have to neuter the gain of the JFET-PNP by putting a resistor in series between the JFET drain and the PNP base, and maybe a base-emitter resistor on the PNP to tame things down a bit. I did run into this situation with the repro circuits of the Thomas Vox Beatles; modern JFETs and PNPs have higher gain than they had in the 60s, so there needs to be some limiting or device selection. Doesn't apply here though.

[pinkjimiphoton]

hi RG thanks for popping in. i found the pot resistance crucial, it only seems to react or start at certain very limited range on the pot... maybe the first 10% or so, if that. thats why i figured maybe faking a 50k dual with two 100k resistors and two 30k resistors between wiper and input would help. wrong way to do it i guess.

i'll socket that darlington pair, and try some higher gain q's, maybe tonite if the tequilla doesn't flow too much. birthday WEEK around here. gig tontite. miserable, i'd rather be at my bench letting magick smoke out.

thanks for the advice on trying a jfet... but i don't understand why a pnp? i'd think it'd want an npn there? forgive my endless stupidity... first rodeo with this kind of circuit. and i have a mind like a sieve

[R.G.]

The LFO transistors - Q11 and Q12 if I remember the numbers right - form a darlington emitter follower. That is, signal goes in the base of the first one, the current is amplified by the current gain of the first transistor, fed to the second transistor base and is then amplified by that second current gain.

Subbing in a JFET for the first NPN makes a source follower feeding the second NPN. The JFET has its drain attached to the positive supply, the NPN has its collector connected to the positive supply. The JFET has nearly infinite "current gain", but this is tempered  by the fact that no current can go in in any case. What it has is transconductance, where a small change in input voltage causes a change in the current of the channel. So you can make a source follower, where the voltage on the source follows the gate. Feed this current to an NPN base, and the NPN now amplifies the JFET source current, giving a compound JFET/NPN "emitter follower" which has lower overall gain, but much higher input resistance because of the JFET. So far, so good.

There's another way to do JFET/bipolar. You hook the JFET up as a source follower, but hook its drain to the base of a PNP. The PNP's emitter is tied to the positive power supply, so all of the current running through the JFET goes through its base-emitter, and is amplified by the PNP's current gain to come through its collector, given that the collector of the PNP is hooked up to something that would let it work. The easy way is to just connect the collector of the PNP to ground. Now the gain from JFET gate to JFET emitter is Yfs (that being the transconductance of the FET) times the source to ground resistor. That current is pulled through the base-emitter of the PNP by the JFET drain, and is current amplified again and expressed on the collector to ground resistor of the PNP. The voltage gain is quite large.

If you think of this n-channel JFET > PNP as an "opamp", you see that the voltage gain is non-inverting from the JFET gate to the PNP collector. But, like an opamp, if you feed some of the output voltage on the PNP collector back to the JFET source, the source acts like an inverting input, as voltage on it removes some of the gate-source differential that made the voltage in the first place. If we connect a resistor between the PNP collector and the JFET source, it acts like an opamp feedback, with the resistor between collector and source as the feedback resistor and the source resistor to ground acting like the inverting input resistor. In the special case where the "feedback resistor" is zero - that is, collector connected to JFET source - the voltage gain from JFET gate to PNP collector is unity, just like an opamp with its output connected to its inverting input. It's a voltage follower. It has high voltage and current gain, but is fed back to a gain of one, and so acts just like the NPN/NPN and N-JFET/NPN followers in this situation.

[pinkjimiphoton]

thanks for the explanation, i will read it more in depth when i get home tomorrow.

one thing i should note with my problem, that may be germaine...

when i'm reading voltage on  q12 b, as soon as i pull my test lead away, suddenly the light will flash, which makes me suspect that maybe the meter is triggering the 5088 somehow?

very weird. more tomorrow, thank you both again!!!

[PRR]

It would take 35V or 50V to smoke a 4.7K resistor.

It's a mystery what happened.

[R.G.]

Yeah, what Paul said.

You haven't offended any minor deities recently have you?

[pinkjimiphoton]

just the usual ones, rg, i make bool do my laundry n stuff...

how the hell could i get that much voltage off a 16vac wart? it does have 1500ma current available, i'm wondering if when i tacked on the resistor...

holy fudgepackage, ya think my iron might be leaking enough voltage to have done that to it? i've been wearing it out, as i gotta soldering station, its a cheap harbor freight 30 watt pencil type. its about the only way i could think of to introduce enough voltage to melt that transistor. i mean it was smoked right off the legs and crumbling, the resistor was still intact but black, and one of the 10u caps was melted to the point the covering was melted on to it and almost impossible to read.

ya saying i should start putting garlic around the entrances to the house, or what? lol

[R.G.]

holy fudgepackage, ya think my iron might be leaking enough voltage to have done that to it? i've been wearing it out, as i gotta soldering station, its a cheap harbor freight 30 watt pencil type. its about the only way i could think of to introduce enough voltage to melt that transistor. i mean it was smoked right off the legs and crumbling, the resistor was still intact but black, and one of the 10u caps was melted to the point the covering was melted on to it and almost impossible to read.

You've identified a possible culprit. Back in my wet-behind-the-ears engineerling days, I once melted the ground clip and all the insulation off a $100 oscilloscope probe. I was working on an off-line switching power supply and tagged the ground to a hot terminal inside it before plugging it back in. Fortunately the $9000 (and 1970s dollars!!) Tek scope still worked though, as a small favor. AC power line leakage can certainly do things like this.

We are normally very careful not to touch the tip of a soldering iron - even cold! - so we don't notice if it's got some leakage. Touch this to a circuit that has any path to ground and you get AC line leakage. It's a very plausible explanation.

ya saying i should start putting garlic around the entrances to the house, or what? lol

Given how much I like garlic, I should have thought of that earlier! 

[pinkjimiphoton]

i have had problems in the past with the wiring in my workshop and ac leakage. i had my old drummer rewire it, he's a certified electrician. thought the problem was solved.. i discovered it cuz i noticed the tingling i was feeling sometimes only happened when i was using electronics on my bench, not constant... i had figured it was from nerve damage, but was leaking enough voltage to be felt from the outlet itself. i forget, something like 30 volts or so... 30 + 22 from the power supply <as you sussed i was lazy and tacked on the resistor with the circuit powered>  = 50 volts, which appears enough to smoke that resistor !!!!

when i get down the shop later... right now its about 102 here in ct, and its plain too humid in there to work right now... i will check with my meter and see if i've got the problem again. i will assume its with the cheap harbor freight soldering pencil. i bet its leaking.
thanks RG for helping me suss this out. i've been wearing it down til its dead, i guess its time to throw it out and go with the station i bought a while back.

i love garlic too, on everything. very strange for a nocturnal vampire like me, but hey, i guess its them lithouanian roots

look forward to making this thing live!!

[pinkjimiphoton]

i am kinda brain dead, too much running around the last week

seems like the problem is the oscillator isn't running. if i touch b of q12, the lamp will flash momentarily as i pull the meter lead away, so i'm asuming the driver transistor is good, tho the voltage around it seems weak.

i temporarily have a 500k trimmer tacked in for the 47k to ground at r50 to try and get a little tuning room but that didn't do anything either. lamp will flash if i touch the base, or if i short any terminal of the dual pot to ground. as soon as i release the connection, it will flash.

i'm thinking a shorted cap i might be missing? or perhaps i put one in backwards?
anyways... the phase shift stages all have like spot on predictable voltages and i can phase it by blocking ambient light with my hand a bit. the audio seems to be passing thru.

just no blinky blinky.

voltages in the lfo and lamp driver stages

q11
c 21.9
b 6.9
e 21.2

^^^^ this seems a little weird to me, but i understand <i think> its cuz of the way
they're wired together. unless my silly hippy brain needs to understand this right now, i'm taking the
free pass on darlingtons being darlingtons.

q12
c  22.00
b  21.1
e  20.5

and the driver
q13
b 1.5
c  21.1
e  0

to me, it looks kinda reasonable. but it won't flash.

thanks for guiding me thru these waters

[pinkjimiphoton]

i'm stumped still. i socketed q11 and 12, and tried some mpsa13 in there and some different 5088's. not much change. r50 read open... i desoldered one side to check, out of the circuit it read 47k. i replaced it with a trimmer as i said this morning.
decided to check the bd681. looked weird on my little chinese gizmotron, said npn one orientation, and pnp the other. looked like a diode connected e to c. never seen that before.
but it said the gain was like, 34. i wasn't expecting that so assumed it must be cooked? the voltages didn't seem right to me, but i don't think i'm quite wrapping whatever bit of mind i have left around the problem.

i looked up the bd681, and had nothing even close. so i took a gamble and tried another mpsa13. now, with the combo of trimmer i added and the stock 200 the filament lights which it wasn't doing before. still, if i touch my meter lead to the b of q11/12 it will flash when i release it. but i can't get this bloody thing to wiggle. no oscillation.
i checked for continuity, heck, i even tried a second speed pot. absolutely nada.
the thing still has audio issues i gotta deal with too, the chorus mode the dry signal is a bit distorted <the vibe side sounds good tho>... but i think first i gotta solve the blinky light dillemma.
gonna order up a couple more bd681's just in case.
the mpsa13 is running a mite bit warm. i twisted the legs to match the bce sockets.

voltages have changed, but barely:
q11
c 21.9
b  6,8
e  21.2     <dumb question, but do 5088's have more than one pinout? seems like b and e should be reversed, and it would be more normal, right??>

c12
c  22
b  21.1
e  20.4

q13
b  1.5
c   21
e   0

any ideas?

[Perfboard Patcher]

the thing still has audio issues i gotta deal with too, the chorus mode the dry signal is a bit distorted <the vibe side sounds good tho>...

What is the voltage at the collector of Q3? Your best chances for a clean sound are when you socialize the headroom, 0.75 x 15 = 11.25V (?)
On the other hand, if the voltage is too close to 15V or 7.5V one of the halves of the output signal will clip.

[R.G.]

You're probably running into more than one problem, just as a guess.

First off, the LFO in the univibe circuit is tricky to troubleshoot. The base of Q11 is at such a high impedance that it's difficult to measure the voltage because all the common meters load it down. The series resistor from the two bias resistors is so big that your meter's input impedance is comparable, and the meter itself loads down the base. The actual input impedance of modern digital meters is high compared to the old swinging-needle meters, but it's not infinite. I don't know if your series resistor is still 2.2M or has been changed to 4.7M, but the bigger one would make it worse.

So here's a way around that. Measure the voltage at the junction of the two biasing resistors, R39 and R40 if I read your schemo right. Jot down that voltage. Then ignore the base of Q11 and measure the voltage at its emitter. This has to be the voltage at the biasing resistors minus the base-emitter voltage at Q11, minus any voltage generated by current through the series bias resistor. This is the base current of Q11 plus any leakage through C19.

You've done the emitter voltage test already, and you're finding it is way too high. The buffering effect of Q11 (if it's working right!) eliminates that pesky problem measuring the base voltage, so you can trust this measurement more. Let's assume that your meter is past its silly games with you and the emitter voltage really is 21+ volts. How could that happen?

Q11 emitter has no path to ground except through the base of Q12. So if it's path to Q12 is open, or if Q12 is back-feeding current into it, those guesses might pull the emitter high. It might also be too high if current is leaking in from the collector of Q11 - an internal short or high leakage. Then there is the usual panoply of problems with solder shorts, board errors, etc. And you made a good guess there - what if the pinout on the devices you're using is incorrect. That might be the case if they're ebay special counterfeits.

Looking at Q12, we see all three pins up at nearly the power supply voltage. But here the emitter is a bit lower than the base, so it might (might...) be working OK, and just have its base pulled too high by a problem with Q11.

How to proceed? It's OK to modify your circuit temporarily to try to figure out what's working and what's not. So here are some things that come to mind, in no particular order.
Pull out Q11, then measure the voltage on the base and emitter of Q12. You're open circuiting the base of Q12 doing this, so we would expect the voltage on the emitter of Q12 to be very low, maybe even fully zero if Q12 has low leakage like it's supposed to. If that's what you find, temporarily bridge a 220K or so resistor from the emitter socket hole for Q11 to the junction of the three bias resistors. The bias resistors are 3.3K and 4.7K, so you ought to find a voltage of your power supply, 22V, times 4.7/(4.7+3.3), or about 13V at the bias resistor node. The base of Q12 will be a little lower because  it's dropping some base current, but it should not be dramatically lower. It might drop by one to three volts, maybe, and this ought to result in something like 10V on the emitter of Q12, of Q12 is working and hooked up correctly.

If that checks out, pull out that temporary 220K resistor and stick Q11 back in, first checking for whether the pinout is correct for the actual manufacturer and type of transistor you're using. All 2N5088s >should< have the same pinout, but it is a good idea to check. Or use your DMM set to read low resistance to figure out which way the internal diodes conduct on the actual transistor.

Now temporarily bridge that 220K ( or 100K, or 47K, or ... ) resistor across the 2.2M/4.7M bias resistor and measure voltage on all three pins of Q11. The resistor being lower now, you can actually read the base voltage.

Give this a try and let us know what happens.

[pinkjimiphoton]

i ordered more parts last nite for it, should be in in a couple days... i will try these implementations and reccomendations later when the sun goes down, its so humid in the basement where my shop is right now my glasses just fog up.

i will def check around q3, PP

rg, thanks i will try that and see whats what. its been nuts here lately, thanks for staying tuned and always having my back!!

[pinkjimiphoton]

boy rg, you were right!! boy-howdee this things about as messed up as a soup sandwich

its still passing audio... and i got to write down voltages n stuff finally...  but ldr 4 is reading about 1.6 v. the others are reading 4.7. all the caps in the phaser stages read the same ... they have about 3.6 v on one side, and 4.7 on the other.
i've read the resistors with one side disconnected and everything off and discharged, and they all read good. the values are good. the voltages are right on with all the other stages. but c14, the 1u going to the ldr from the junction of q9e and r31 is off. i'm getting 3.6v on the negative side as expected, but not getting 4.7v on the positive side like the other stages.... getting 1.6 or so. this makes no sense to me, maybe i'm missing something obvious. i thought i'd shorted a cap and replaced it, same exact thing. unsoldered one side of ldr 4, no change with voltage, and the ldr seems to be fine, and makes no difference in or out of the circuit.

stumped. could one of the q's be weak? could the momentary flame out have roasted more stuff than i thought it did? i can replace all the caps and q's on the board, all the resistors too, but it seems like most of its working.

anyways, voltages
15.1vdc after the rectifier, under load the adapter is putting out 8v on one side, 8.6 on the other for 16.6vac

q's 1-12 2n5088

q1
c  1.6
b  1.3
e   1

q2
c  4.5
b  1.6
e  1.4

q3
c  11.2
b  4.5
e  3.9

q4,6,8 all the same
c  15.1
b  4.7
e  4.3

q5,7,9 all the same

c  11.2
b  4.3
e  3.7

q10
c  15.1
b  1.7
e   1.1  <<< this on looks fishy, and should cuz ldr 4 isn't working i think, or else its a buffer, right?

q11
c  21.0
b  17.2
e  20.6

q12 
c  21.3
b  20.6
e   2.0

q13, i subbed in an mpsa13 darlington for a lamp driver until i can get the proper bd681 in in a couple days, ordered 10 in case i let the magic smoke out again... legs twisted awkwardly to match the proper pinout

b  3.1
c  18
e  2.5

right now, the lamp stays just lit enough to see the filament. still, hitting the b of q12 makes it flash brightly when i pull the lead away, so i'm assuming thats good? maybe?

i dunno... its quarter to four am, i gotta get some rest before the next onslaught begins again. i feel like i live in a perpetual groundhog day. where's bill murray when ya need him.

so... any ideas what may be killing that ldr's voltage? bad resistor? maybe one that opens up when it gets warm or something?

stumped. again, deeply indebted for the help!
i printed out all the advice so i can have it on my bench before writing this out.
tomorrow... i mean today is another day

oh yeah, forgot, i tried the 100k dual log pot that came with the kit, no difference. still doesn't sweep. wondering if the depth circuit got fried, too.

peace out, and thanks!!!

[R.G.]

its still passing audio... and i got to write down voltages n stuff finally...  but ldr 4 is reading about 1.6 v. the others are reading 4.7. all the caps in the phaser stages read the same ... they have about 3.6 v on one side, and 4.7 on the other.
i've read the resistors with one side disconnected and everything off and discharged, and they all read good. the values are good. the voltages are right on with all the other stages. but c14, the 1u going to the ldr from the junction of q9e and r31 is off. i'm getting 3.6v on the negative side as expected, but not getting 4.7v on the positive side like the other stages.... getting 1.6 or so. this makes no sense to me, maybe i'm missing something obvious. i thought i'd shorted a cap and replaced it, same exact thing. unsoldered one side of ldr 4, no change with voltage, and the ldr seems to be fine, and makes no difference in or out of the circuit.

If I understand correctly, you're seeing about 4.7V on the + side and 3.6V on the - side of C8, C11, and C16; but 1.6 on the + side  and 3.6 on the negative end of C14?

If that's right, what can we make up in our heads that would make it so? Well, first off the negative end is connected to the emitter of Q9, so it ought to have the same voltage. A quick look at your transistor pin voltages shows 3.7V, close enough. That one checks. The + end of C14 goes through the LDR and  R28. You've replaced C14 so we'll assume for a moment that it's good. C14 is a DC blocker, so its + side is the same voltage as what it connects to – which is the base of Q10.

You also show Q10 voltages which you properly picked out as suspicious. Even though the biasing resistors should be setting up Q10 just like the other transistors, this one is suspiciously low on the base and emitter. Ahah! Q10 could be dragging C14 down! But how?

You can make a transistor's base too low two ways. One is to drag down the bias network that's trying to hold it up; another is to drag down on the emitter. The base has to follow the emitter if the thing dragging the emitter can pull it down by eating a lot of current. If you've checked the values of R30 and R33, and the supply voltage on the collector pin is at 15V (the measurements show this) then the bias resistors are trying to make things right. So maybe something is pulling the emitter down.

What could that be? Solder shorts and board faults are always possible, but you've checked for those. Let's look at what else it could be. The emitter connects to C16 and C17 as well as R34. You've checked R34, so it's not a completely wrong value there. That leaves possibly C16 and C17 being leaky or some such as the obvious suspects. A really good next step in debugging (IMHO) would be to remove C16 and C17 and see if the base and emitter voltages of Q10 pop back up to where they are supposed to be.

stumped. could one of the q's be weak? could the momentary flame out have roasted more stuff than i thought it did? i can replace all the caps and q's on the board, all the resistors too, but it seems like most of its working.

MMMmmmmaybeeee. A frying voltage and current big enough to roast a resistor always carries the possibility of collateral damage. But I think your transistor pin voltages suggest that most of them are OK. Which 4.7K resistor exactly was fried?

anyways, voltages
15.1vdc after the rectifier, under load the adapter is putting out 8v on one side, 8.6 on the other for 16.6vac

I think you mean after the regulator you get 15.1Vdc. Is that right?

q10
c  15.1
b  1.7
e   1.1  <<< this on looks fishy, and should cuz ldr 4 isn't working i think, or else its a buffer, right?

q11
c  21.0
b  17.2
e  20.6

q12 
c  21.3
b  20.6
e   2.0

Good work. You homed right in on the ones that are suspicious.

q13, i subbed in an mpsa13 darlington for a lamp driver until i can get the proper bd681 in in a couple days, ordered 10 in case i let the magic smoke out again... legs twisted awkwardly to match the proper pinout

b  3.1
c  18
e  2.5

right now, the lamp stays just lit enough to see the filament. still, hitting the b of q12 makes it flash brightly when i pull the lead away, so i'm assuming thats good? maybe?

I think what this means is that Q13 is working OK (to be verified later when we get Q11 and Q12 running) and that any disturbance on its base makes it wobble. This is especially true if Q12 is OK as well, and the disturbance on Q12's base is current-amplified and sent through to Q13. I view this as good news that Q12 and Q13 are probably OK.

oh yeah, forgot, i tried the 100k dual log pot that came with the kit, no difference. still doesn't sweep. wondering if the depth circuit got fried, too.

I'm guessing that the depth circuit is more OK than the issues with Q10 and Q11, and may be just fine. Let's dig into the transistors first.

[pinkjimiphoton]

first things first, thanks rg!! man my life is so upside down... just got up, gotta leave for tonites gig in less than 2 hours so it may take a couple days to catch back up, and i appreciate the help and guidance <btw, the first time i "talked" with you was in 1998 or so, when i modded and broke my crybaby... you helped me thru it then, and i still use the same one, and am grateful for the continuous help over the years >

its still passing audio... and i got to write down voltages n stuff finally...  but ldr 4 is reading about 1.6 v. the others are reading 4.7. all the caps in the phaser stages read the same ... they have about 3.6 v on one side, and 4.7 on the other.
i've read the resistors with one side disconnected and everything off and discharged, and they all read good. the values are good. the voltages are right on with all the other stages. but c14, the 1u going to the ldr from the junction of q9e and r31 is off. i'm getting 3.6v on the negative side as expected, but not getting 4.7v on the positive side like the other stages.... getting 1.6 or so. this makes no sense to me, maybe i'm missing something obvious. i thought i'd shorted a cap and replaced it, same exact thing. unsoldered one side of ldr 4, no change with voltage, and the ldr seems to be fine, and makes no difference in or out of the circuit.

If I understand correctly, you're seeing about 4.7V on the + side and 3.6V on the - side of C8, C11, and C16; but 1.6 on the + side  and 3.6 on the negative end of C14?

yessir, exactly right

If that's right, what can we make up in our heads that would make it so? Well, first off the negative end is connected to the emitter of Q9, so it ought to have the same voltage. A quick look at your transistor pin voltages shows 3.7V, close enough. That one checks. The + end of C14 goes through the LDR and  R28. You've replaced C14 so we'll assume for a moment that it's good. C14 is a DC blocker, so its + side is the same voltage as what it connects to – which is the base of Q10.

i think more than a univibe light bulb  is going off suddenly  

You also show Q10 voltages which you properly picked out as suspicious. Even though the biasing resistors should be setting up Q10 just like the other transistors, this one is suspiciously low on the base and emitter. Ahah! Q10 could be dragging C14 down! But how?

so THAT is actually the final phase shift stage? i thought i was a buffer!! my god. if i had a modicum of intelligence, it would still be a day late with too much coffee

You can make a transistor's base too low two ways. One is to drag down the bias network that's trying to hold it up; another is to drag down on the emitter. The base has to follow the emitter if the thing dragging the emitter can pull it down by eating a lot of current. If you've checked the values of R30 and R33, and the supply voltage on the collector pin is at 15V (the measurements show this) then the bias resistors are trying to make things right. So maybe something is pulling the emitter down.

What could that be? Solder shorts and board faults are always possible, but you've checked for those. Let's look at what else it could be. The emitter connects to C16 and C17 as well as R34. You've checked R34, so it's not a completely wrong value there. That leaves possibly C16 and C17 being leaky or some such as the obvious suspects. A really good next step in debugging (IMHO) would be to remove C16 and C17 and see if the base and emitter voltages of Q10 pop back up to where they are supposed to be.

i will try that when i get home tonite, if i'm not too tired, or tomorrow. it makes a lot of sense. i didn't think the caps would be a likely culprit, but it sure makes sense if the resistors are ok. its hard to tell, some of the resistors have been reading open when i disconnect one side, but out of the circuit completely seem to be right on spec.
ghosts in my basement or something?

stumped. could one of the q's be weak? could the momentary flame out have roasted more stuff than i thought it did? i can replace all the caps and q's on the board, all the resistors too, but it seems like most of its working.

MMMmmmmaybeeee. A frying voltage and current big enough to roast a resistor always carries the possibility of collateral damage. But I think your transistor pin voltages suggest that most of them are OK. Which 4.7K resistor exactly was fried?

i had mistakenly said it was the 4.7k to ground <r 46> but was actually r49, also 4.7k  feeding right into the base of q13. the voltages there seem a lot better, so i assume i melted it when the flame out happened... the transistor, i mean. cuz now it reads pnp in my meter one orientation, and npn the other, with ah hFE of 34 either way. so i am assuming the rest of the stuff around it probably survived and the one resistor took the hit <along with the bd681 and maybe a couple passives i'm still digging up> all 1/4 watt metal oxide flame proof <hah> resistors at 2% tolerance i think.  the one that flamed i replaced with a 1/2 watt 4.7k resistor, thinking the extra beef isn't a bad idea with the amount of current in this part of the circuit.

anyways, voltages
15.1vdc after the rectifier, under load the adapter is putting out 8v on one side, 8.6 on the other for 16.6vac

I think you mean after the regulator you get 15.1Vdc. Is that right?

yessir. to ground, from each side, i get 8vdc at the power jack, with 16.6vac off the unloaded wall wart.
after the regulator, 15.1v strong  ... how do i get 21 volts elsewhere in the circuit is beyond me too, unless the diode bridge and big caps are stepping up the voltage some... but that can wait for now

q10
c  15.1
b  1.7
e   1.1  <<< this on looks fishy, and should cuz ldr 4 isn't working i think, or else its a buffer, right?

q11
c  21.0
b  17.2
e  20.6

q12 
c  21.3
b  20.6
e   2.0

Good work. You homed right in on the ones that are suspicious.

i've been trying to pay attention here when i can

q13, i subbed in an mpsa13 darlington for a lamp driver until i can get the proper bd681 in in a couple days, ordered 10 in case i let the magic smoke out again... legs twisted awkwardly to match the proper pinout

b  3.1
c  18
e  2.5

right now, the lamp stays just lit enough to see the filament. still, hitting the b of q12 makes it flash brightly when i pull the lead away, so i'm assuming thats good? maybe?

I think what this means is that Q13 is working OK (to be verified later when we get Q11 and Q12 running) and that any disturbance on its base makes it wobble. This is especially true if Q12 is OK as well, and the disturbance on Q12's base is current-amplified and sent through to Q13. I view this as good news that Q12 and Q13 are probably OK.

well, they are now after being replaced hopefully the bd will go in and it will work properly. then the final challenge is gonna be tapering the pot correctly.  i think i'll do my experiments with tapering with the power off next time, what do ya think? 

seriously, thanks RG!! this and a couple other projects are getting me ready to tackle the ludwig circuit. if i pull that one off, its going as a gift to steve hunter in spain, and i'll keep the one dino built for me. i treasure that thing!!

again, thank you!
peace!!

oh yeah, forgot, i tried the 100k dual log pot that came with the kit, no difference. still doesn't sweep. wondering if the depth circuit got fried, too.

I'm guessing that the depth circuit is more OK than the issues with Q10 and Q11, and may be just fine. Let's dig into the transistors first.