I’m new here, and I’m trying to learn a bit about my newly-acquired Moog Source. I’m pretty experienced with the electronics of amplifiers and other pro audio gear, but the multiple causes possible for my Source’s problems is overwhelming. I’m hoping one of the experts here can help me pinpoint the issue.
THE PROBLEM:
The Source powers up, the control panel lights up and is functional, the jog wheel is functional, and a very low constant tone is heard on the output. But, the keyboard is completely inactive and no tones are produced.
THE INVESTIGATION RESULTS SO FAR:
All power supply voltages are correct.
The internal battery is still functional, holds a charge, and is not leaky.
There are no burnt/disconnected/missing components.
No components are hot.
When I short random pins on various Op-Amps, I am able to cause bleeps/electronic swells, etc. Therefore, I don’t think there is a broken connection or dead IC in the analog path.
Feeling that the problem is somehow related to the components near the keyboard connection, I began testing things in that area.
Results of those tests:
When measuring the voltage at Pin 3 of U31A on the Digital Board, I see rising/changing voltages when various
keys are pressed on the keyboard. That signals to me that, physically, the keyboard is working. So, it seems
that maybe the way the keyboard’s voltages are transferred further down the “chain” is the problem.
When measuring the output of U31A at Pin 1, I see the voltage go up to approx. 9.286V when ANY key is
pressed. When the key is released, the voltage returns to about 0. This happens no matter which key is pressed,
essentially showing an ON/OFF effect, rather than the increasing voltage with ascending keys on the keyboard I
measured at Pin 3. Does this indicate a blown U31?
Another anomaly I’ve discovered is that both ends of R84 on the digital board (which is also connected to U31) show
only about 2.4V, rather than the 5V the schematic says should be there. The 5V power rail has been checked and is
correct.
Finally, when measuring at the conjunction of R58 and R61 on the digital board (while attempting to set the “Full
Scale” 10V), the R62 trimpot has no effect. But, the voltage is correct at 10V, so no adjustment was needed.
Regardless, I’ve included this detail in case it was related to the current problem.
If both ends of R84 measure 2.4V, then something is wrong with the trace that supply 5V to R84. Work your way back the trace until you find 5V, you may have a broken trace and have to jumper it. If the output of U31 doesn’t go above 2.4v then the cpu doesn’t detect a key change and the Source will not trigger anything.
The junction of R58/R61 isn’t supposed to change. That point is a regulated 10V source. Re-read the calibration procedure for Full Scale adjustment.
I don’t know the Source at all - but have a comment about your fault finding procedure.
Shorting random Op-Amp pins wouldn’t really be recommended…!
You’ve probably shorted out the power supply either to ground or to somewhere where more damage could be done..!
One thing that you wrote disturbed me a little :" - When I short random pins on various Op-Amps, I am able to cause bleeps/electronic swells, etc."
It’s never a good idea to short random pins on chips, especially op amps. Whilst this will surely result in some observable electrical changes, it can do more harm than anything else.
I rather use the “wet finger” approach, which does essentially the same but without the potential damage to the chips, as I don’t cleanly short out pins this way, but just change resistance slightly to see if there’s any reaction or if the chip is “locked/dead”.
It’s not very “scientific” but has often saved me time and trouble when trying to locate approximately the trouble area in analog/linear circuits.
(of course, a follow-up with the schematics, test point values, proper tools and measuring equipment is in order to properly confirm a defective component.)
Yeah guys, I know the op amp shorting thing was not the best of all ideas.
I’ll take all of your suggestions and try them out. Hopefully, I’ll be able to report back with some positive info.
I stopped reading here " When I short random pins on various Op-Amps, I am able to cause bleeps/electronic swells, etc. Therefore, I don’t think there is a broken connection or dead IC in the analog path."
Alright, alright. I know I phrased that a bit recklessly. I’m not as dumb as that comment appeared. The shorts I made weren’t “random” in a true sense and weren’t particularly dangerous. I simply injected small voltages (less than 5V) into a couple of the TL072s at points relatively distant from the output of the synth (per the schematic) to see if a signal was making it all the way to the output. This was just a crude “buzzer” type troubleshooting technique. This has proven effective in a number of other projects and I’ve yet to blow an IC doing so. But, it probably wasn’t too helpful in a mixed analog/digital circuit.
I’m sorry that it exposed my inexperience with synth circuits. I’m more used to dealing with various amplifiers and the like.
All I can say is that it might seem logical to inject 5 volts into the input of an opamp IF that 5 volts is current limited by an inline resistor AND (as inputs follow outputs often enough), you’re sure you’re not actually applying 5 volts to the output of other devices.
A straight 5 volts hot off the power supply can have some real current and cause damage (500ma-1.5a.)
Sorry, but you started out writing like a seasoned tech and suddenly took a left turn into circuit-bending.
The former is engineering and latter is art. (flames are better for another thread for those passionate about this subject.)
But for all we know, more things are messed up than previously were.
It’s a complex enough instrument and requires proper troubleshooting techniques and deliberate operations. No guessing.
Don’t take any perceived anger from me personally.
If you don’t know me, I’m a synth tech.
So maybe you can understand my attitude, but I apologize if I offended you.
Fwiw. you can just as easily monitor audio from a probe hooked to an amp if it also goes through a resistor. (common grounds obviously.)
Use something reasonable - 47K-220K hooked to an audio probe to see if you hear the VCOs operating, maybe the output from the filter or noise circuit.
Ideally, use a scope after a long study and note-taking of the service manual.
The Source service manual is not really all that good, but it should at least help trace things out.
For a quick and simple ‘touch free’ audio probe, a guitar cable connected to a cheap practice amp can be used. Hold the free end of the cable like a probe, get the tip within a half-inch of an oscillator (or octave divider) and you can usually hear the note it is playing (or not playing). Don’t touch the tip to anything.
As a correction, R84 DOES have 5V on one side of it. Due to the lettering on the PCB being indistinct, I was reading the 2.4V off of R64, R67, and R66.
Hint?: “Mod Rate” LED is solidly lit. It is not blinking. I’ve heard this may be an indication of CPU trouble…
Still concentrating on the “Keyboard” section of the Digital board, I made a number of additional voltage measurements at various points
on that board.
Hopefully, these will enable the experts here to get a better sense of what might be going on:
U31 Pin 1= 4.927V
U31 Pin 2= 0V (no matter if any keys are pressed or not)
U31 Pin 3= 0.0049V with no keys pressed.
0.353V with the lowest key of keyboard pressed.
8.732V with the highest key of keyboard pressed.
U32 Pin 2= .0365V with no keys pressed.
4.926V with ANY key pressed.
U30 Pin 1= 9.054V with no keys pressed.
9.059V with lowest key pressed.
Up to 9.091V when pressing some keys toward the center of the keyboard.
Pin 2= 3mV (no matter what is pressed)
Voltages at the power supply pins on all ICs are correct.
Based on these readings, my amateur mind is thinking U31, U30, or both are non-functional. I’d love to be told otherwise though!
The keyboard voltages are correct, that means U30 is working. But not all of those test points are DC voltages. Pin 2 should be stepped voltages, and pin 1 should be a pulse train when a key is pressed. U31 is a comparator which should be generating a pulse train, R84 converts it to TTL. U32 pin 2 should also be a pulse train.
If you’re not seeing keyboard CV at pin 5 of U5, that could be many potential causes. At this point you really need an oscilloscope as neither a multimeter or DC readings will tell you anything.
Thanks for the information MC! I’ll break out the scope, do some measurements, and report back. To this point I had just been using my Fluke 187. I have been, until now, unsure of when, exactly, the digital pulses were created.
With either a good DMM or scope, you can also check to see if the CPU is receiving a clock signal and if it’s being reset upon power up.
The CPU or crystal aren’t the hardest parts to find (but the eprom might be.)
Most everything else is somewhat common.
Good luck.
FWIW a CPU lock-up can be caused by a lot of things: bad logic gates chips, bad ram, bad eprom (unlikely but still possible), bad CPU itself (I’ve seen some 30 year old flaky Zilog Z80 get problematic inside warm enclosed casings due to heat build-up inside a Prophet 5).
I remember not too long ago, another member had problems with his Moog Source. Turned out to be a defective logic gate chip in the incremental controller section of the digital board that was locking-up the whole synth. But it took him a week to find out the problem (with the help and guidance of some members here, too).
Thanks for all of the tips. I did a little investigation with my scope; perhaps one of you could enlighten me a bit on what I should be seeing. When probing the pins where I was told to expect a pulse stream, I saw what looked like square waves. But, when I pressed a key, those “square waves” (I put in quotes because I couldn’t see the vertical portions of the waves- only the flat tops) turned into a solid horizontal line on the scope. Does this mean something is working in the opposite way it should be?
I’ve gone through the schematics and looked for points on the schematic that specify certain voltages to be present at that particular location; the results conformed to the schematic. Also, at the points on the OSC (1 and 2) portions of the synth board which show triangle, sawtooth, and square waves, I am seeing those waveforms on my scope.
What I probably need is some guidance as to what steps I should take to really diagnose this problem. Given that the power rails, etc. are all checking out as correct, are there certain points on the boards that can be scoped (or the voltage measured) to narrow down the problem to one particular area? For instance, is there a measurement that I can take to be assured that the digital portions are working correctly, or the components are receiving a proper signal from the keyboard, etc.? I feel like I need to create a checklist for myself that will help me divert my attention away from circuits that are already working perfectly, to ones that need investigation…
What you’re seeing is most likely what’s called multiplexed data.
It may not even be necessary to look at this data to repair the instrument, but I’ll give you a really quick explanation…
In synths like these, they have a thing called a data buss.
Obviously what goes on the data buss is data. +5 or zero (1 or 0s.)
Where it all gets crazy is that another buss called the address buss can choose where the data above comes from or goes to.
So in a literal fraction of a second, the CPU (which controls both the address and data buss) could tell it to check if any buttons are pressed as an input, send out to the display, read the jog wheel and send out data to control the pitch.
In reality, they usually do many more operations as well.
But the big issue is knowing where to look for what data and when.
That requires advanced knowledge of a scope and understanding gates, clocked logic and other deep subjects.
The important thing is that you’re seeing some data. This can suggest the CPU is working and you should look elsewhere.
I personally would check to see what you have and what you don’t and find some commonality to it all.
For example, do you have CV hitting the VCOs? Envelope voltages going to the filter and/or VCA?
Do certain buttons on the top panel work, but others don’t?
Does the wheel work? (LED count up or down?)
Are the VCOs putting out waveforms?
Document everything you can to see if you can narrow it down to something that might cause all of the faults you’re experiencing.
After that, look at the schematics to see if there are parts common to whatever failures you’ve experiencing.
One bad 4051 chip, for example, could take out 8 different ouput signals.. or refuse to read 8 different input signals.
Look for things like that.
For those of you just joining us, here are my non-functioning Source’s symptoms:
No output when keys are pressed.
Panel functions appear to be operating correctly.
All power rail voltages are correct.
The synth emits a quick “woop” sound upon turn-on (see below), but then only a VERY low constant tone thereafter. Keyboard has no effect.
Jog wheel fully functional.
No burned/missing components or traces, no loose/corroded connectors, no leaking caps, functioning battery with no leaks.
Some troubleshooting I’ve done since my last post:
I did a few temporary IC replacements to see if any particular chip was malfunctioning. U32 on the Digital Board was swapped with no effect on the problem. On the Synth Board, U3, U4, and U5 and were swapped with no effect.
Another hint that may (or may not) be related to my problem: the Mod Rate LED will respond differently every time I power up the Source. Sometimes it remains steadily lit, sometimes it flashes rapidly (regardless of the positioning of the Mod wheel), and sometimes it flashes slowly (regardless of the position of the Mod wheel). When I first power up the Source, I hear a rapidly descending “shooowwwooop” sound coming from the speaker and the Mod Rate LED “sputters” as the aforementioned sound is present (2-3 seconds), then returns to a steadily-lit state when the sound stops.
I used my scope to trace backwards from my output jack. I noticed that with the scope on a very sensitive setting, I can see some miniscule output from the synth which changes slightly when a key is pressed. However, that tiny signal is not heard through my amplifier at all (yes, the amp is functioning properly and is turned up loud).
Tracing back, I can follow that incredibly small signal back through to the oscillators. At a certain point though, I reach the perfect sawtooth/triangle waves produced by the oscillators before they were robbed of all recognizable features, culminating in the synth output I am seeing.
Here are some pictures of my oscilloscope screen while testing the points at which the signal changes:
The OTHER end of R117 shows NOTHING. Is this proper? I, at first, thought R117 might have failed, but perhaps there is another reason for not seeing an attenuated triangle wave on its other side?
As a result of these pictures, I’m curious to know what is causing the nice big triangle and sawtooth waves I’m seeing coming out of the OSC circuits to be reduced to (seemingly) random noise.
Some other questions:
What kind of “reset pulse” should I look for on power up? What should that look like on a scope? Is it long enough (in duration) to see?
Does the inconsistent “Mod Rate” LED flashing indicate a problem with the CPU (even though all of the front panel controls seem to be working)?
Thanks for your help. I’m sure I could take this synth down to a local repair shop, but I’m enjoying the learning process and appreciate all the insight I get from the experts here.
Short spike that ends usually in a ground. (low)
Doesn’t matter though.
You don’t need to check it.
If your controls work or the keyboard plays, the cpu is working.
I can’t say. Could be something loose or something taking a lot of the CPU cycles as a broken state.
Normally it should just flash.
I’m not as dumb as that comment appeared. 
The shorts I made weren’t “random” in a true sense and weren’t particularly dangerous. I simply injected small voltages (less than 5V) into a couple of the TL072s at points relatively distant from the output of the synth (per the schematic) to see if a signal was making it all the way to the output. This was just a crude “buzzer” type troubleshooting technique. This has proven effective in a number of other projects and I’ve yet to blow an IC doing so. But, it probably wasn’t too helpful in a mixed analog/digital circuit.