Joerg Reisenweber wrote: > R54 indeed seems insane with 120R, Sebastien has warned on #milkymist that the rise and fall times of the optocoupler have been a source of trouble in the past, and only switching to this part, along with tweaking of resistor values, has solved the issue.
Looking at figure 6 of [1], this does indeed seem to be a weak spot of the optocoupler. The test circuit appears to be the one in figure 24, although the voltages don't seem to match. The footnote promised by the asterisk on Cl (Cload) does not exist. I'm surprised by two things here: 1) why does Rload have such an impact on the fall time ? I could understand the rise time, because Rload has to load the parasitic capacitance. 2) if the main factor is indeed parasitic capacitance, then the numbers seem way off. E.g., if I consider this a low-pass filter, then I'd get a cut-off frequency of about 1 MHz (MIDI runs at 31.25 kbit/s, so this should be plenty) with R = 3.3 kOhm and a generous C = 50 pF. - Werner _______________________________________________ http://lists.milkymist.org/listinfo.cgi/devel-milkymist.org IRC: #milkymist@Freenode