AW: Expo conv. heater (tomg et al.)

[Martin Czech]

>> Martin,
> 
> I think it just means that the compensation does not work to 100%
> (Nothing works to 100%, after all !). Temperature changes slowly
> inside the enclosure when you power the synth up, so there you have
> the slow change. Most of it is compensated by the chip, and the remaining
> error is what we hear. As temperature rises over time, it approaches the
> temperature at which you have made your last tuning adjustment.
> Makes sense for me, even if there would be nothing such as temperature
> gradients across the chip.
> 

The first mail seemed to imply that the frequency shift is quite large,
this would require a large gardient.

If the observed error was very small, ok, this would explain the time lag.

But why do some voices lag, and others don't. Is there perhaps a large thermal
gradient also on the boards? I guess PSU.  Are the "failing" voices naer to 
areas of large power consumption ?

m.c.