> From: Mike Williams
>
> Whilst it's in great cosmetic condition for its age, a
> problem with Voice 2 has just developed. As I have little
> test equipment I swapped chips around between synth voice
> circuits to find that the CEM3372 chip in Voice 2 was
> defective. When I put a good chip in the circuit that didn't
> work either so I think I have a defective chip and a problem
> in the surrounding circuit. By checking voltages around the
> pins of the CEM3372 and comparing these with a functioning
> Voice there are significant differences on pins 17 and 12
> i.e. the VCF out / VCA in across one of the 2.2 uF
> electrolytic capacitors. Consequently I suspect a faulty 2.2
> uF cap and plan to replace that tomorrow to see if that does
> the trick.

It's pretty easy to see if the caps are leaky, since you can remove the
3372 from its socket, leaving the caps floating. Then you can use an
ohmmeter across the caps to see if there's any conductivity. Make sure you
put the + lead on the + terminal, though. It may take many seconds for the
caps to charge up and up until the meter finally says infinity--or not, if
the caps are leaky.

> The synth is just shy of 30 years old so I'm wondering what
> the best advice is regarding replacement of electrolytic
> caps? I suspect they maybe due to give some problems in the
> synth voice circuits as I've noticed some slight tuning
> stability issues along with a slight portamento sometimes ...

It's true that 1980s vintage aluminum electrolytics dry out in time. I
wouldn't go on a mass replacement crusade, though, because replacing parts
is not only time-consuming but subjects the board and nearby parts to
excess heat. I'd try to figure out which caps if any are bad. Tuning
issues are more likely to be the polypropylene sample-and-hold caps.

> If a cap goes can it damage the CEM3372? I'm guessing it can
> because it looks like that may have happened here! The CEM's
> are relatively expensive to replace (£35 each on eBay) so I'm
> wondering if I should re-cap soon ...

No one knows exactly what's inside them, but the block diagram in the data
sheet suggests that leakage, or even a dead short or open, in CX30 or CX37
wouldn't harm anything.

> Does anyone know of a good source of CEM3372 chips? I've read
> that some of the more recent revisions deliver different
> sound qualities to the older revisions, to the extent that
> they are not interchangeable. Is there anything in that?

I've heard that too, but it isn't something we were aware of at the time
we designed the Polaris, and I've not done any tests. They are readily
available on eBay. I have some new old stock, from what I believe are the
first few chip versions, 3160, 3160A and 3160B. Most of what I see on eBay
appear to be more recent, since they say 3160D. I'm guessing that that's
the chip version, because the other numbers like 8418 look like a
conventional date code (year and week).

--

Ciao, Paul D. DeRocco
Paul mailto:pderocco@...