Cgs synth

...and unfortunately it's not the fun kind.

I have two VCS boards assembled. Neither do what they are supposed to.

The first has voltage and pot control of the slopes, input also works, lin/exp switch is fine, I get slopes out, but the end-out and trigger arrangement do nothing. I hard wired it to cycle just to see if that works, and it also doesn't. And the LED stays on all the time.

The other VCS is even more confusing than the first. Again, input, pots and CV all seem to work, trigger and end out definitely do not. This time, the LED works, but the whole thing seems more like a Source of Uncertainty module! The voltage wobbles all over the place, and seems to be affected by my body resistance/capacitance!

For the wobbling/random voltage problem, I'll check the wiring again. But since (besides the LED), both PCBs have trouble with the trigger/end out/cycle feature, I wonder what the similarity is? I replaced one of the LM3900 ICs, just to see if that was the problem (it wasn't).

ICs seem to be getting power on the correct pins. I don't have a scope, unfortunately.

I've gone over the transistors, and even checked to make sure that the BCE arrangement wasn't funny in that batch, also double checked wiring. Beyond that I'm not so knowledgeable.

I've built tons of Ken's stuff, this is the first time I've ever screwed something up. Suggestions for other possible checks to make are greatly appreciated!

Second problem of the wobbling voltage fixed by removing some other PCBs further downstream (DC Mixers + Analog Voltage, will be debuggered later), so that VCS works fine. The first one is still kaput however. Suggestions welcome!

D.

Hi,

Glad you got #2 working. When you say "removing other PCBs downstream", that could indicate bad grounding practices or a multitude of other issues. Can you elaborate a bit?

I'm not familiar with this particular board but it might be helpful to see clear photos of the board, it's neighbors and the wiring around the area.

Did you use IC sockets? I might try replacing that 3900 with a different one (again).

How good are your soldering skills? Are you comfortable that you can spot a tiny short or bad solder joint. Sorry if that's blunt, but it gives everyone a better idea of your confidence level.

Keep the questions rolling. We're bound to figure it out :-)

Regards,

Doug

Hi Doug,

thanks for your thoughts.

--- In cgs_synth@yahoogroups.com, "Doug" <cgsyahoo@...> wrote:

> Glad you got #2 working. When you say "removing other PCBs downstream", that could indicate bad grounding practices or a multitude of other issues. Can you elaborate a bit?

I had the output of the processor section of the board connected to a DC Mixer board for offset, and then running to the Analog Logic board. They were all part of the same module (i.e. behind same panel) but at some point in the process I gave them all separate power lines. I'll deal with those guys later though...

I've got everything working on the problem VCS board except the Trigger In and End Out, so it can't cycle. All other inputs and the Output work fine, as does the LED now.

I get a faint flicker on the LED from an external trigger, but no noticeable output voltage from the Output. Using a patch cable to connect End Out to Trigger in gives no response at all.

> I'm not familiar with this particular board but it might be helpful to see clear photos of the board, it's neighbors and the wiring around the area.

I can try to get this later on. I made some voltage comparisons between the IC pins of the working and not-working PCBs a while ago, would something like that be useful?

> Did you use IC sockets? I might try replacing that 3900 with a different one (again).

All ICs replaced at least once. Also most of the transistors, espc the one in the End/Trigger section.

One thing I've noticed is that with these boards, it's *very* easy to lift the traces. Other CGS boards haven't given me this much trouble. Ken, is your new PCB company able to do the same thickness of traces as before?

> How good are your soldering skills? Are you comfortable that you can spot a tiny short or bad solder joint. Sorry if that's blunt, but it gives everyone a better idea of your confidence level.

Normally, they are excellent. I've successfully built something like 16 modules from PCBs already, and a handful more on protoboard. Plus I was trained as a silversmith some years ago ;-)

Best!
D.

> Keep the questions rolling. We're bound to figure it out :-)
>
> Regards,
>
> Doug
>
> --- In cgs_synth@yahoogroups.com, "derekatumaticdotnl" <derek@> wrote:
> >
> > Second problem of the wobbling voltage fixed by removing some other PCBs further downstream (DC Mixers + Analog Voltage, will be debuggered later), so that VCS works fine. The first one is still kaput however. Suggestions welcome!
> >
> > D.
> >
> > --- In cgs_synth@yahoogroups.com, "derekatumaticdotnl" <derek@> wrote:
> >
> > > The other VCS is even more confusing than the first. Again, input, pots and CV all seem to work, trigger and end out definitely do not. This time, the LED works, but the whole thing seems more like a Source of Uncertainty module! The voltage wobbles all over the place, and seems to be affected by my body resistance/capacitance!
> >
>

OK, the last problem turned out to be a cracked trace between pins 9 and 11 of the LM3900. Apparently, the force of cutting the lead alone cracked this trace, so I'd like to ask Ken again if this might be a result of his new PCB manufacturer, since I didn't have anything remotely like this problem with other/earlier CGS PCBs.

Also, for the archive... if you can't get the VCS to cycle then setting it in cycle mode and then adjusting the trimpot can help. The effect won't take place immediately, but on the next power-up or trigger it should. My procedure was:

1. Flip the module around, with the Out connected to a VCO so I can hear when it starts cycling.
2. Adjust trimpot a little bit.
3. Power down/power up. Listen for cycling in the VCO.
4. When I got the VCS cycling, ease back on the trimpot until just the point where it stops, then bring it up again so it starts. I'm guessing this gets it as close to a 5V trigger out as possible without a scope.

>OK, the last problem turned out to be a cracked trace between pins 9 and 11

of the LM3900. Apparently, the force of cutting the lead alone cracked this
trace, so I'd like to ask Ken again if this might be a result of his new PCB
manufacturer, since I didn't have anything remotely like this problem with
other/earlier CGS PCBs.

I doubt it was the manufacturer. ALL boards suffer from this, with no
apparent rhyme or reason as to where and when it happens.

>Also, for the archive... if you can't get the VCS to cycle then setting it

in cycle mode and then adjusting the trimpot can help. The effect won't take
place immediately, but on the next power-up or trigger it should. My
procedure was:

A problem in the genuine Serge is that cycle locks up at various frequencies
due to a charge inbalance in the trigger circuit. I solved the problem by
changing the 10n capacitor to 1n . This is the cap near the cycle switch on
the circuit diagram. Perhaps you may consider making that change too.

Ken
_______________________________________________________________________
Ken Stone sasami@... otherunicorn@...
Modular Synth PCBs for sale <http://www.cgs.synth.net/>
Australian Miniature Horses & Ponies <http://www.blaze.net.au/~sasami/>