S?bastien Bourdeauducq wrote: > It increases PCB routing problems, maintainance, and design > complexity. I'd rather see those resources spent elsewhere...
We already have a ton of signals going that way. It/they can probably just blend in with the rest. The main question in my mind is whether we need any protection there at all. Let's see what scenarios we have: Protection Upset: transient Overcurrent Short to GND --------------- --------------- --------------- --------------- Switch w/ fault ignored DCC5V down, with indication Just switch ignored DCC5V down, no indication None M1 may reset ignored (*) M1 main fuse (*) "Overcurrent" would mean more than 50 mA but less than, say, 500 mA. (*) DVI connector may be rated at only 1 A. Thus, overheating may occur. Now, my problem here is that I don't know what the probabilities are for all these events. It may very well be that DCC5V just never causes trouble in real life. But if it does, we almost certainly want to know about it, so that we don't waste time trying to find other explanations for why some device doesn't work. How about this: let's try to have these signals, but allow them to be sacrificed in case they make things difficult. With enable to be sacrificed before fault. Either should be happy in any of the FPGA's voltage domains, so we can let routing pick any pins they like. - Werner _______________________________________________ http://lists.milkymist.org/listinfo.cgi/devel-milkymist.org IRC: #milkymist@Freenode
