I've heard something similar -- From that guy who seriously modded his poly-61 on YouTube actually. He say DCO2 is completely digital.
But you can have VCOs generating a waveform, while using a D/A to control and automatically trim the voltage it receives based on a sampled frequency that is compared against a digital clock. Synthesizers like the Jupiter-8 and Sequential Max have VCOs that are digitally trimmed with a built-in auto-tune function.
The Poly-61 is probably similar in that respect, and I would suspect resistors and caps in the area of the A/D or divider circuit they are using to sample the VCO outputs. If these voltages are not correct, or the caps are discharging too soon, the CPU will interpret the frequency incorrectly. That's just my guess... I bet there's nothing at all wrong on the voice board.
Did you have to replace your battery? Did it possibly damage the main board at some point? Mine's completely hosed and it's difficult to tell which components are damaged and which only look bad. Even the traces are suspect. -pc
--- In DW8000@yahoogroups.com, "crustypaul" <crustypaul@...> wrote:
>
> Theres a few places around the net that state that DCO1 is in fact a VCO 'by the signal generation method', whatever they mean by that. I haven't checked it out yet but you are right, if they are 100% digital then its unlikely there would be any way to retune the voices as in theory there should be no way they could drift.