More stuff on the 303...

[Tony Allgood]

Hi,

Well I have built my first TB, at least the analogue bit anyhow. Some
more musings from me about the circuit. I have already established that
the VCO waveform is very dependent on the selection of the saw buffer
FET. Roland do indeed have preselected SK30s to choose from. However,
Roland have plenty of these (but not at the moment in the UK, because I
bought them all, ha ha) I had five other more common types of fet and
they just didn't give the output I wanted.

The filter resonance is not only dependent on the res pot, but also on
the actual value of the 5V3 power supply. If it is only a little higher,
the gain of the diff amp becomes large enough to cause the filter to
self oscillate with the res pot fully clockwise. The 5V3 is set up to
feed the dac, and hence to give 1v/oct, so I suppose most 303s will have
fairly similar power supplies.

The mixer is greatly affected by output load. When loaded with a 22k
input impedance from my bench amp, it clips on negative swings, and the
positive is affected in an unusual way. This generates some nice
distortion. This may well be a cause of that TB sound. However, the
amount of distortion is dependent on the level of signal from the volume
pot. I am experimenting with additional overdrive circuitry to enhance
the sound. But I can't decide whether to put it before the VCA or after
it. The more I look at the waveforms, the less I am convinced that the
BA662 actually affects the sound... but more tests have yet to be done.
I could really do with a distortion meter or a DSO.

All the Japanese parts are fairly easy to get hold of, once you know
where to look. My thanks to all who helped me in this. The SC1583 is
actually very cheap. 75p compared to 3UKP for the SSM2210. I havn't
tried the CA3046 yet.

Regards,

Tony Allgood, Cumbria, UK

e-mail: oakley at techrepairs.freeserve.co.uk

Rack mounted Moog VCF module. Details to be found at...

http://aupe.phys.andrews.edu/diy_archive/schematics/effects/filter.html