> > Mark, can you explain to Vitaly how to produce a complete dump, as > > you just did? > > Look again, he already did that. Oh, you're right. Sorry. > > I'd also need at least a third user's dump, together with the model > > information (A or not). Without that, there will be too many > > registers that could be the one we are after. Of course, more dumps > > would be welcome. The more we have, the easier the work will be. > > OK, if the opportunity comes up. Really I'm not sure the effort > would be worth it... my ASB100 !A ignores writes to pwm1 with no > harm done AFAICS. You're right. I sure have better things to do, and actually it's simply impossible to come to a conclusion with only two dumps. Let's keep them in a corner until the day we have at least two more, then we'll see. After all, we don't have a datasheet for the ASB100 so we shouldn't even be supporting it. That said, had I to bet on a register, I would choose 0x85 in bank 3. But I admit I must have less than 1% chance to have it right ;) -- Jean Delvare http://www.ensicaen.ismra.fr/~delvare/