OBERHEIM SYNTH

Hi,

I'm having problems with the autotune on an OB-Xa. The problem is that when pressing autotune, sometimes all voices tune, but equally often one or more (often not more than 3 or 4) refuse to tune in the sense that the machine flashes the corresponding voice number on the patch buttons. However, it seems as if the machine in fact does try to tune the voice; I've tried trimming one voice purposely off by one semitone, cancelling the autotune voltages using the internal slider on the underside of the front panel (the right hand one, can't remember if it's SW1 or SW2), pressing autotune, noting that the voice in question is flashing on the panel, yet when playing the machine it seems fine.

I've tried analyzing the signal produced by the voice cards when autotuning (one way is to short the base and emitter of Q8 on the lower control board which makes the autotune process audible), and also looking at the signal going to U118 on the upper control board (the shift register which is part of the autotune periodmeter), but I can't see anything odd there, in particular, it looks the same no matter whether the machine says it can tune the voices or not.

I've read success stories along the line of 'I recapped all the voice boards and now my machine works fine', but I'd really like to find the offending components rather than just do a shotgun operation.

My latest idea is that something is wrong in the D/A-converter circuitry and some of the low order bits are faulty, as when listening to the autotune signal (and slowed down by a factor of 8 or else it goes very fast) it appears that the tuning takes place in fairly large discrete steps, in fact for long periods on some oscillators there's no perceivable pitch change at all). I haven't looked more into that yet.

The other idea, related to recapping, is that the two decoupling capacitors decoupling the +/-15V lines on the voice boards are not doing their job properly so that the oscillators tend to sync to each other when close in pitch via to the supply lines (most likely the +15V one). But I can't detect this happening to any great extent when just listening to a voice and twiddling the detune knob for instance.

Has anyone else had this problem and would like to share any experiances?

/Ricard

Here I go replying to myself, I haven't had any responses here on the forum, but one member responded privately (you know who you are, thanks!, and it's been a real nice chat afterwards too).

What I'd like to ask is to one extent other people also have problems with the autotune.

In searching for the problem with the autotune, I've analyzed not only
the hardware but also the software program that runs on the Z80 processor.

It turns out to be a fact that the autotune acceptance range is -24 .. 0 cents. First off, the range is rather large, and secondly it is offset in a strange way.The algorithm works like this:

For each oscillator, the synth tries to calibrate the oscillator 50 times. For each time it measures the period of a pulse formed from four cycles of a target tone C4 (1046.502 Hz) tone. The resulting period, measured as 1/4.9152 MHz cycles, should be 4964 (hex). If the period is within +1/-0 of 4964 (hex), then it goes on to the next oscillator. Otherwise, if the last measurement cycle was
within +256 / -0 from 4964 (hex), the oscillator is considered ok, otherwise it is considered failed, which will cause the corresponding LED to flash after the tune procedure is complete. 256 steps (or FF hex) steps from 4964 works out as 24 cents at the measurement frequency.

I've verified that this is in fact the case by breaking up the autotune circuit and injecting a tone from another (digital) synth where I can set the frequency fairly exactly, and also verify the pitch using a tuner.

So that's the reason for several oscillators failing on my machine; probably the general instability in the synth causes the oscillators not to be spot on, andif they're not, the oscillators are considered untuned if they are even half a hertz over the reference frequency, but not if they are up to 24 cents below.

I believe this is a bug in the software, notwithstanding any instability in the machine due to aging components etc.

All this is rather technical I know, so perhaps not really the right forum. Someone might find it useful I hope. My main reason for this post was to try and poll to what extent others also experience autotune problems, both on unrestored and restored machines.

/Ricard

Your caps are bad.  If you don't replace them, eventually the synth will die.  If you don't replace the caps, some CMOS will go out, you'll experience all sorts of erratic behavior, then the Bridge Rectifier or regulators will go. Oberheim's are famous for not tuning if the caps go bad.  

Replace the caps.  The bridge rectifiers in the original design are simply too small and also need to be replaced.  Replacing the regulators won't hurt either; the synth will run cooler.  

An extreme mod has been made on my OB-8, probably has close to the same software, in the autotune loop by removing the caps that DC block the VCF from the VCA and I pass autotune every time because the synth has been restored properly.  

There isn't a bug in the software.

-Mike


--- In oberheim@yahoogroups.com, "Ricard" <ricard2010@...> wrote:

>
> 
> 
> Here I go replying to myself, I haven't had any responses here on the forum, but one member responded privately (you know who you are, thanks!, and it's been a real nice chat afterwards too).
> 
> What I'd like to ask is to one extent other people also have problems with the autotune.
> 
> In searching for the problem with the autotune, I've analyzed not only
> the hardware but also the software program that runs on the Z80 processor.
> 
> It turns out to be a fact that the autotune acceptance range is -24 .. 0 cents. First off, the range is rather large, and secondly it is offset in a strange way.The algorithm works like this:
> 
> For each oscillator, the synth tries to calibrate the oscillator 50 times. For each time it measures the period of a pulse formed from four cycles of a target tone C4 (1046.502 Hz) tone. The resulting period, measured as 1/4.9152 MHz cycles, should be 4964 (hex). If the period is within +1/-0 of 4964 (hex), then it goes on to the next oscillator. Otherwise, if the last measurement cycle was
> within +256 / -0 from 4964 (hex), the oscillator is considered ok, otherwise it is considered failed, which will cause the corresponding LED to flash after the tune procedure is complete. 256 steps (or FF hex) steps from 4964 works out as 24 cents at the measurement frequency.
> 
> I've verified that this is in fact the case by breaking up the autotune circuit and injecting a tone from another (digital) synth where I can set the frequency fairly exactly, and also verify the pitch using a tuner.
> 
> So that's the reason for several oscillators failing on my machine; probably the general instability in the synth causes the oscillators not to be spot on, andif they're not, the oscillators are considered untuned if they are even half a hertz over the reference frequency, but not if they are up to 24 cents below.
> 
> I believe this is a bug in the software, notwithstanding any instability in the machine due to aging components etc.
> 
> All this is rather technical I know, so perhaps not really the right forum. Someone might find it useful I hope. My main reason for this post was to try and poll to what extent others also experience autotune problems, both on unrestored and restored machines.
> 
> /Ricard
>

Well, there definitely is a bug in the software, no doubt about that, but that's not saying that there's something else wrong in the synth which causes this bug to be exposed, e.g. the oscillators should tune tightly enough that the bug is never exposed, which probably explains why Oberheim themselves appearently never noticed it. (The fact that they didn't points to sloppy engineering but that's another issue).

Many posts refer to replacing the electrolytic capacitors on the voice boards, which I might very well do, however, no one has really explained why they would cause pitch instability. There are 7 caps on each voice board, 5 of which are in the signal chain, and 2 of are used for power supply decoupling. The latter two can affect pitch stability to some extent I suppose, but the other five? They might affect other things, such as the frequency response or other effect, but I can't see how they can affect the tuning. 

I agree that components can and will age, but as a result of that there will be some parameter (capacitance, leakage current or whatever) that will degenerate and which can be measured and diagnosed. Agreed, if the leakage current goes up, power drain will increase, putting a larger load on the regulators which can cause premature failure or instability due to thermal issues etc.

That said, there is another problem in this particular machine, on all voices there are a couple of control voltage values which cause the 12 dB to generate a lot noise at its output. Could very well be a bad cap in the middle somewhere causing that phenomenon.

The very act of removing the voice boards, even potentially cleaning the contacts, and then replacing them I would think would tend to have a larger effect on stability.

Yeah, the power supply rectifiers are underdimensioned. Someone has already had a go at these in this particular machine though.

Well, I'll see. Perhaps I'll have a go at recapping one voice board, taking the others out, and see if it actually helps.

/Ricard

--- In oberheim@yahoogroups.com, "Mike" <mborish_2000@...> wrote:

>
> Your caps are bad.  If you don't replace them, eventually the synth will die.  If you don't replace the caps, some CMOS will go out, you'll experience all sorts of erratic behavior, then the Bridge Rectifier or regulators will go. Oberheim's are famous for not tuning if the caps go bad.  
> 
> Replace the caps.  The bridge rectifiers in the original design are simply too small and also need to be replaced.  Replacing the regulators won't hurt either; the synth will run cooler.  
> 
> An extreme mod has been made on my OB-8, probably has close to the same software, in the autotune loop by removing the caps that DC block the VCF from the VCA and I pass autotune every time because the synth has been restored properly.  
> 
> There isn't a bug in the software.
> 
> -Mike
> 
> 
> --- In oberheim@yahoogroups.com, "Ricard" <ricard2010@> wrote:
> >
> > 
> > 
> > Here I go replying to myself, I haven't had any responses here on the forum, but one member responded privately (you know who you are, thanks!, and it's been a real nice chat afterwards too).
> > 
> > What I'd like to ask is to one extent other people also have problems with the autotune.
> > 
> > In searching for the problem with the autotune, I've analyzed not only
> > the hardware but also the software program that runs on the Z80 processor.
> > 
> > It turns out to be a fact that the autotune acceptance range is -24 .. 0 cents. First off, the range is rather large, and secondly it is offset in a strange way.The algorithm works like this:
> > 
> > For each oscillator, the synth tries to calibrate the oscillator 50 times. For each time it measures the period of a pulse formed from four cycles of a target tone C4 (1046.502 Hz) tone. The resulting period, measured as 1/4.9152 MHz cycles, should be 4964 (hex). If the period is within +1/-0 of 4964 (hex), then it goes on to the next oscillator. Otherwise, if the last measurement cycle was
> > within +256 / -0 from 4964 (hex), the oscillator is considered ok, otherwise it is considered failed, which will cause the corresponding LED to flash after the tune procedure is complete. 256 steps (or FF hex) steps from 4964 works out as 24 cents at the measurement frequency.
> > 
> > I've verified that this is in fact the case by breaking up the autotune circuit and injecting a tone from another (digital) synth where I can set the frequency fairly exactly, and also verify the pitch using a tuner.
> > 
> > So that's the reason for several oscillators failing on my machine; probably the general instability in the synth causes the oscillators not to be spot on, andif they're not, the oscillators are considered untuned if they are even half a hertz over the reference frequency, but not if they are up to 24 cents below.
> > 
> > I believe this is a bug in the software, notwithstanding any instability in the machine due to aging components etc.
> > 
> > All this is rather technical I know, so perhaps not really the right forum. Someone might find it useful I hope. My main reason for this post was to try and poll to what extent others also experience autotune problems, both on unrestored and restored machines.
> > 
> > /Ricard
> >
>

What is your objective?  

You aren't going to be able to measure a messed up cap unless you have some serious equipment like a very high bandwidth 4 channel DSO scope with 16 channels of logic and huge memory.  Perhaps if you took the cap out of the circuit and did some sort of pulse tests to it you could find some irregularities with cheaper equipment, but why waste your time?

The caps are failing.  I can tell you that because replacing them has fixed the same problem for me in the past.  I don't need to troubleshoot a cap on the molecular level to know that it is messed up.  Caps do all sorts of strange and unexplainable stuff when they are beyond their lifespan that you won't learn in school.

You have classic faulty cap symptoms:

1.) The Synth tunes sometimes which mostly rules out CMOS with the exception of wiggling (see below). 

2.) You probably just took it out of storage or woke it up from a very large period of dormancy.  If it starts acting up after a few weeks to a month, the caps are bad; they are trying to reform or have dried out.

If it isn't the caps it is probably CMOS, a trimmer with a partially lifted wiper, a bad jumper, or a bad solder joint on the jumper socket that has been cracked from removal stress.  

Even if you find some silver bullet problem unrelated to the caps, the synth will actually work better if you replace the caps.  Your leds will run brighter, patches will load faster and it will tune faster.

I forgot to tell you in the previous post that you should always reset the chips in their sockets before troubleshooting; this clears up lots and lots of problems because Oberheim used crappy sockets that patially "wiggle" the chips out of them when the board flexes from movement.

Good luck.

-Mike   

    

--- In oberheim@yahoogroups.com, "Ricard" <ricard2010@...> wrote:

>
> 
> Well, there definitely is a bug in the software, no doubt about that, but that's not saying that there's something else wrong in the synth which causes this bug to be exposed, e.g. the oscillators should tune tightly enough that the bug is never exposed, which probably explains why Oberheim themselves appearently never noticed it. (The fact that they didn't points to sloppy engineering but that's another issue).
> 
> Many posts refer to replacing the electrolytic capacitors on the voice boards, which I might very well do, however, no one has really explained why they would cause pitch instability. There are 7 caps on each voice board, 5 of which are in the signal chain, and 2 of are used for power supply decoupling. The latter two can affect pitch stability to some extent I suppose, but the other five? They might affect other things, such as the frequency response or other effect, but I can't see how they can affect the tuning. 
> 
> I agree that components can and will age, but as a result of that there will be some parameter (capacitance, leakage current or whatever) that will degenerate and which can be measured and diagnosed. Agreed, if the leakage current goes up, power drain will increase, putting a larger load on the regulators which can cause premature failure or instability due to thermal issues etc.
> 
> That said, there is another problem in this particular machine, on all voices there are a couple of control voltage values which cause the 12 dB to generate a lot noise at its output. Could very well be a bad cap in the middle somewhere causing that phenomenon.
> 
> The very act of removing the voice boards, even potentially cleaning the contacts, and then replacing them I would think would tend to have a larger effect on stability.
> 
> Yeah, the power supply rectifiers are underdimensioned. Someone has already had a go at these in this particular machine though.
> 
> Well, I'll see. Perhaps I'll have a go at recapping one voice board, taking the others out, and see if it actually helps.
> 
> /Ricard
> 
> --- In oberheim@yahoogroups.com, "Mike" <mborish_2000@> wrote:
> >
> > Your caps are bad.  If you don't replace them, eventually the synth will die.  If you don't replace the caps, some CMOS will go out, you'll experience all sorts of erratic behavior, then the Bridge Rectifier or regulators will go. Oberheim's are famous for not tuning if the caps go bad.  
> > 
> > Replace the caps.  The bridge rectifiers in the original design are simply too small and also need to be replaced.  Replacing the regulators won't hurt either; the synth will run cooler.  
> > 
> > An extreme mod has been made on my OB-8, probably has close to the same software, in the autotune loop by removing the caps that DC block the VCF from the VCA and I pass autotune every time because the synth has been restored properly.  
> > 
> > There isn't a bug in the software.
> > 
> > -Mike
> > 
> > 
> > --- In oberheim@yahoogroups.com, "Ricard" <ricard2010@> wrote:
> > >
> > > 
> > > 
> > > Here I go replying to myself, I haven't had any responses here on the forum, but one member responded privately (you know who you are, thanks!, and it's been a real nice chat afterwards too).
> > > 
> > > What I'd like to ask is to one extent other people also have problems with the autotune.
> > > 
> > > In searching for the problem with the autotune, I've analyzed not only
> > > the hardware but also the software program that runs on the Z80 processor.
> > > 
> > > It turns out to be a fact that the autotune acceptance range is -24 .. 0 cents. First off, the range is rather large, and secondly it is offset in a strange way.The algorithm works like this:
> > > 
> > > For each oscillator, the synth tries to calibrate the oscillator 50 times. For each time it measures the period of a pulse formed from four cycles of a target tone C4 (1046.502 Hz) tone. The resulting period, measured as 1/4.9152 MHz cycles, should be 4964 (hex). If the period is within +1/-0 of 4964 (hex), then it goes on to the next oscillator. Otherwise, if the last measurement cycle was
> > > within +256 / -0 from 4964 (hex), the oscillator is considered ok, otherwise it is considered failed, which will cause the corresponding LED to flash after the tune procedure is complete. 256 steps (or FF hex) steps from 4964 works out as 24 cents at the measurement frequency.
> > > 
> > > I've verified that this is in fact the case by breaking up the autotune circuit and injecting a tone from another (digital) synth where I can set the frequency fairly exactly, and also verify the pitch using a tuner.
> > > 
> > > So that's the reason for several oscillators failing on my machine; probably the general instability in the synth causes the oscillators not to be spot on, andif they're not, the oscillators are considered untuned if they are even half a hertz over the reference frequency, but not if they are up to 24 cents below.
> > > 
> > > I believe this is a bug in the software, notwithstanding any instability in the machine due to aging components etc.
> > > 
> > > All this is rather technical I know, so perhaps not really the right forum. Someone might find it useful I hope. My main reason for this post was to try and poll to what extent others also experience autotune problems, both on unrestored and restored machines.
> > > 
> > > /Ricard
> > >
> >
>

> What is your objective?  

It's really twofold, I want to get this machine fixed, but I also want to learn what the root of the problem is so it can be applied to other machines.

I'm a bit wary of things like 'I replaced all X and Y and it worked' because it doesn't really say what was actually wrong or what you actually did, you may very well have done Z in the process which was the actual fix, but without knowing what was wrong you don't know what you put right, and since you didn't really you did Z you might not have done it properly anyway.

It borders on audiophile things like 'I demagnetized my CD's and now they sound much better, the highs are clearer and there's more presence to string sounds, ...'.

I've had interesting experiences with 50+ year old vintage equipment. Bottom line with a lot of tube stuff is electrolytics don't necessarily go bad, but paper caps do, and carbon composition resistors tend to wander upwards in value. I've got a fair amount of experience with vintage synths to and have never really come across anything like the OB-Xa, not even in a CS-80. Then again, maybe Oberheim used really crappy components which Yamaha would not have dreamed of (indeed, the trimmers and IC sockets quality would point to this).

> You aren't going to be able to measure a messed up cap unless you have some serious equipment like a very high bandwidth 4 channel DSO scope with 16 channels of logic and huge memory.  Perhaps if you took the cap out of the circuit and did some sort of pulse tests to it you could find some irregularities with cheaper equipment, but why waste your time?

Well, because replacing 50+ is also time consuming it if happens to be only 16 of them that need replacing.

Also, I don't the failure mode is rocket science. It's probably that the capacitance has gone down, the ESR has gone up, or the leakage current has gone up. Most likely the first or third neither of which are hard to measure.

I'm not really interested in taking out every capacitor, measuring it, and putting it back if I can't find anything wrong with it, rather I'm interested in learning which ones fail and why. For instance, I can't believe that a coupling cap in the filter would affect pitch stability, although of course it can have an effect on the sound. I know for instance that electrolytics that are operated at very low voltages tend to fail with time (because the polarizing voltage which maintains the dielectric is too low).

> The caps are failing.  I can tell you that because replacing them has fixed the same problem for me in the past.  I don't need to troubleshoot a cap on the molecular level to know that it is messed up.  Caps do all sorts of strange and unexplainable stuff when they are beyond their lifespan that you won't learn in school.

I'd rather think that circuitry does all sorts of strange and unexplainable stuff when some components in the design go out of spec.

But I've heard this recommendation from several sources, the only thing I've been wary of is that most of them have been rather unscientific. Yes, replacing all caps might help, but did you really have to? Of course you could argue that once the voice boards are out and your soldering iron hot you might as well replace them all, but I'd be a lot happier knowing why I'm actually replacing them.

Which caps did you replace in your machine? I've heard various bids on this one, from 'all electrolytics on the voice boards', to 'all electrolytics everywhere' to 'the power supply electrolytics are the ones that need replacing'.

> You have classic faulty cap symptoms:
> 
> 1.) The Synth tunes sometimes which mostly rules out CMOS with the exception of wiggling (see below). 

What exactly do you mean by 'CMOS'? The analogue demultiplexers, the digital logic, the battery backed up memory, or all CMOS logic chips?

> 2.) You probably just took it out of storage or woke it up from a very large period of dormancy.  If it starts acting up after a few weeks to a month, the caps are bad; they are trying to reform or have dried out.

That's true, the machine hasn't been used for quite a while (and wasn't stored in favorable conditions either).

> If it isn't the caps it is probably CMOS, a trimmer with a partially lifted wiper, a bad jumper, or a bad solder joint on the jumper socket that has been cracked from removal stress.  

Yeah, those trimmers aren't the best to put it mildly.

> Even if you find some silver bullet problem unrelated to the caps, the synth will actually work better if you replace the caps.  Your leds will run brighter, patches will load faster and it will tune faster.

I have noticed some differences in sound between voices which I could put down to bad coupling caps.

> I forgot to tell you in the previous post that you should always 
reset the chips in their sockets before troubleshooting; this clears up lots and lots of problems because Oberheim used crappy sockets that patially "wiggle" the chips out of them when the board flexes from movement.

Yeah I've noticed that some of them have crept up slightly. Re-seating is good practice anyway as it clears up minor corrosion that could creep in between the IC pins and the sockets.

> Good luck.

Thanks!

/Ricard

Your missing the big picture.  

If the PSU is messed up and not handling electrons properly under a variable load, it's going to mess up your CPU, binary words, etc.  Don't forget that there are millions of switches in the synth that are switching millions of times a second which cause millions of little voltage and current ripples.  If a cap chokes on some electrons for a split second, it will probably ruin a word or cause a voice failure in the autotune process: caps don't necessarily fail in a uniform or linear manner.  The failing caps in the PSU and rails are wrecking havoc on your binary switches and they're fudging words as a result.  Just replace ALL of the caps in the synth and don't worry about how long it takes  because your synth will sound better.  Replace the bridge rectifiers too, preferably with 6A 200v reverse FRED's or something with a very fast switching time >200ns because you will get a better sonic response, more bass and treble, and the whole thing will run cooler.  

Older 4000 series CMOS chips have a life expectancy of about 15 years.  All CMOS made in a 6um process (all from 70's some from 80's) suffers from 'metal migration,' a slow reaction where ions from the metallized gate material of each MOSFET in the chip flow down to the substrate of the chip over time and BOOM!  The result is a short circuit in one or more MOSFET gates to the substrate layer.  Don't replace them if they're not bad though because they have unique electrical properties that can't be duplicated easily with modern equivalents; some are unbuffered, they have high propagation delay, they all have way higher internal resistances, etc.

On another note, you wanna help me out with my CS-80?  I've got some issue where the filter envelopes on random voices stay open when they keys are played in a rapid staccato manner and I've got IL set to -5 and AL set to + 5.

It's never the same voices where the filter gets stuck either.

I haven't gone through the whole calibration yet.  I'm always scared to move old open trimmers because sometimes the wiper will mount on accumulated dirt and throw the whole thing out of wack. 

-Mike






--- In oberheim@yahoogroups.com, "Ricard" <ricard2010@...> wrote:

>
> 
> > What is your objective?  
> 
> It's really twofold, I want to get this machine fixed, but I also want to learn what the root of the problem is so it can be applied to other machines.
> 
> I'm a bit wary of things like 'I replaced all X and Y and it worked' because it doesn't really say what was actually wrong or what you actually did, you may very well have done Z in the process which was the actual fix, but without knowing what was wrong you don't know what you put right, and since you didn't really you did Z you might not have done it properly anyway.
> 
> It borders on audiophile things like 'I demagnetized my CD's and now they sound much better, the highs are clearer and there's more presence to string sounds, ...'.
> 
> I've had interesting experiences with 50+ year old vintage equipment. Bottom line with a lot of tube stuff is electrolytics don't necessarily go bad, but paper caps do, and carbon composition resistors tend to wander upwards in value. I've got a fair amount of experience with vintage synths to and have never really come across anything like the OB-Xa, not even in a CS-80. Then again, maybe Oberheim used really crappy components which Yamaha would not have dreamed of (indeed, the trimmers and IC sockets quality would point to this).
> 
> > You aren't going to be able to measure a messed up cap unless you have some serious equipment like a very high bandwidth 4 channel DSO scope with 16 channels of logic and huge memory.  Perhaps if you took the cap out of the circuit and did some sort of pulse tests to it you could find some irregularities with cheaper equipment, but why waste your time?
> 
> Well, because replacing 50+ is also time consuming it if happens to be only 16 of them that need replacing.
> 
> Also, I don't the failure mode is rocket science. It's probably that the capacitance has gone down, the ESR has gone up, or the leakage current has gone up. Most likely the first or third neither of which are hard to measure.
> 
> I'm not really interested in taking out every capacitor, measuring it, and putting it back if I can't find anything wrong with it, rather I'm interested in learning which ones fail and why. For instance, I can't believe that a coupling cap in the filter would affect pitch stability, although of course it can have an effect on the sound. I know for instance that electrolytics that are operated at very low voltages tend to fail with time (because the polarizing voltage which maintains the dielectric is too low).
> 
> > The caps are failing.  I can tell you that because replacing them has fixed the same problem for me in the past.  I don't need to troubleshoot a cap on the molecular level to know that it is messed up.  Caps do all sorts of strange and unexplainable stuff when they are beyond their lifespan that you won't learn in school.
> 
> I'd rather think that circuitry does all sorts of strange and unexplainable stuff when some components in the design go out of spec.
> 
> But I've heard this recommendation from several sources, the only thing I've been wary of is that most of them have been rather unscientific. Yes, replacing all caps might help, but did you really have to? Of course you could argue that once the voice boards are out and your soldering iron hot you might as well replace them all, but I'd be a lot happier knowing why I'm actually replacing them.
> 
> Which caps did you replace in your machine? I've heard various bids on this one, from 'all electrolytics on the voice boards', to 'all electrolytics everywhere' to 'the power supply electrolytics are the ones that need replacing'.
> 
> > You have classic faulty cap symptoms:
> > 
> > 1.) The Synth tunes sometimes which mostly rules out CMOS with the exception of wiggling (see below). 
> 
> What exactly do you mean by 'CMOS'? The analogue demultiplexers, the digital logic, the battery backed up memory, or all CMOS logic chips?
> 
> > 2.) You probably just took it out of storage or woke it up from a very large period of dormancy.  If it starts acting up after a few weeks to a month, the caps are bad; they are trying to reform or have dried out.
> 
> That's true, the machine hasn't been used for quite a while (and wasn't stored in favorable conditions either).
> 
> > If it isn't the caps it is probably CMOS, a trimmer with a partially lifted wiper, a bad jumper, or a bad solder joint on the jumper socket that has been cracked from removal stress.  
> 
> Yeah, those trimmers aren't the best to put it mildly.
> 
> > Even if you find some silver bullet problem unrelated to the caps, the synth will actually work better if you replace the caps.  Your leds will run brighter, patches will load faster and it will tune faster.
> 
> I have noticed some differences in sound between voices which I could put down to bad coupling caps.
> 
> > I forgot to tell you in the previous post that you should always 
> reset the chips in their sockets before troubleshooting; this clears up lots and lots of problems because Oberheim used crappy sockets that patially "wiggle" the chips out of them when the board flexes from movement.
> 
> Yeah I've noticed that some of them have crept up slightly. Re-seating is good practice anyway as it clears up minor corrosion that could creep in between the IC pins and the sockets.
> 
> > Good luck.
> 
> Thanks!
> 
> /Ricard
>

I finally nailed this after much searching.

This particular machine has a Kenton MIDI retrofit installed, and as it turns out, they way the power was being supplied to the MIDI board, it caused the whole MIDI board ground to have a slight 100 Hz (twice the mains frequency in this country) ripple superimposed on it. That in turn would be superimposed on the bend CV driving the bend/modulation panel, which then caused the oscillators to 'warble' ever so slightly. It's almost inaudible, at least you don't notice it until you know it's there.

Since the OB-Xa does not disconnect the bend lever during autotune (try bending and hit AUTO for some interesting fun), that meant that the pitch simply was not stable while the machine was trying to autotune, causing the resulting tuning to be rather random. In the BEND NARROW mode, it would work much better though, and since that function comes up randomly when switching the machine on, it would sometimes tune nicely and sometimes not for no appearent reason.

Re-routing the power supply to the MIDI board cured the problem, the machine now consistently autotunes fast and correctly. Pitch stability seems ok after an all-night soak test. Pitch accuracy could still be better, but I'm putting that down to the Oberheim design for now.

/Ricard

I think I should mention that 're-routing' the power supply involved scrapping the local +5V regulator on the MIDI board and taking the 5V from the OB-Xa power supply. That way I didn't get any voltage drop in the connecting cable due to the local reservoir capacitor being charged every cycle of the mains frequency. Since the MIDI board only draws 25mA and the power supply in the synth itself is more like 1A I figure the 5V supply can handle it. 25mA is just little more than a single LED so there's bound to be some margin there.

On the subject of CEMs, doesn't Doug at www.synthparts.com have a selection?

/Ricard

Yes, I have CEMs in stock. Some OB-Xa voice cards too. Doug synthparts.com
--- In oberheim@yahoogroups.com, "Ricard" <ricard2010@...> wrote:

>
> I think I should mention that 're-routing' the power supply involved scrapping the local +5V regulator on the MIDI board and taking the 5V from the OB-Xa power supply. That way I didn't get any voltage drop in the connecting cable due to the local reservoir capacitor being charged every cycle of the mains frequency. Since the MIDI board only draws 25mA and the power supply in the synth itself is more like 1A I figure the 5V supply can handle it. 25mA is just little more than a single LED so there's bound to be some margin there.
> 
> On the subject of CEMs, doesn't Doug at www.synthparts.com have a selection?
> 
> /Ricard
>