[sdiy] Re: bit one/DAC fault tip

[Stephen Lenham]

Hi Ruud/list, 

I can suggest another possibility. I had a very similar fault on a 
Sequential Sixtrak but it could theoretically affect any 80's 
micro/DAC/4051s-based synth. 

When I got the Sixtrak it played, but all the notes were "wrong" - virtually 
random, but repeatable. The sounds were iffy and when trying to set 
parameters their value would jump, sometimes by a substantial amount. 

To condense a LOT of head-scratching into a few words, the DAC was being 
driven correctly but had an internal "stuck bit" (in my case, bit 10 of the 
12-bit device was permanently high). 

This seems consistent with what you have observed. Pretty much *everything* 
in that sort of synth rests on the DAC - pitch, software envelopes (don't 
know if the Bit has those) and - less obviously - the reading of panel pots 
(DAC value is compared with pot voltage to create a crude ADC). So if thing 
seem to be basically working but with the "wrong values" everywhere, it 
might be worth checking the DAC (and its input latches, etc.). 

Your sysex patch dumps may well be arriving just fine, but if they can't be 
properly converted into analogue control voltages... 

With this fault the severity of the symptoms could vary substantially 
depending on which part of the DAC is dead - I was lucky in that it was a 
high order bit so the effects were very pronounced. If you have software 
envelopes, program a long release and listen - any jumps in the volume as it 
decays would confirm this problem. 

Of course, it may be something completely different ;-) But perhaps this 
will be useful to somebody. Either way, best of luck fixing it! 

Steve L. 


Rude 66 writes: 

> 
> ok, here's an update on the bit: 
> 
> cleaned all connectors. didn't make a difference. re-soldered the output
> board. didn't make a difference. after playing around a bit more, i can
> describe the problem a little better: something is modulating everything.
> the oscillators sound almost like they're frewuency modulated: on some notes
> you have a pure tone, on others you have what my mono-poly in fm-mode sounds
> like.
> and the 'sound staying on after the eg has closed' is more apparent when you
> open the filter. when full open there's only a slight drop in volume if the
> key is released with the release at zero. the more you close the filter, the
> more it dies out. also if you first select a patch, you don't hear the
> sound. it's only after it has been triggered that it won't stop anymore.
> there are some things left to try: i haven't taken the chips out of their
> sockets and clean them.. (and there are 6 CEM filter chips..) and i haven't
> re-soldered the 'analogue board' yet. though maybe it could be a digital
> problem.. what i find strange is that the sounds i send it via sysex or tape
> save are corrupted right away. 
> 
> ruud 
> 
>