Hmmmmmm.... I'm still skeptical. You say you let the watchdog timer just time out every few milliseconds, rebooting the machine? An interesting idea... How does a 12C508 know from millisecond to millisecond what is going on? Since you don't have any EEprom, can't rely on the RAM to be correct after a watchdog timer reboot, the processor is "starting cold" every watchdog period. I suppose this might be OK for a toy or a benchtop prototype, but it would not do for a product which can fail in an unsafe way, like a heater or a cooking appliance. I'd still vote for a US$0.03 transistor brownout circuit. Lawrence Lile ----- Original Message ----- From: "Roman Black" To: "Lawrence Lile" Sent: Tuesday, July 31, 2001 5:27 AM Subject: Re: [PIC]: 12C508 BOR > That's not entirely true Lawrence! You can give > brownout proofing to many simple designs done > entirely in software. Maybe not EVERY design, I > agree there. > > But in most cases the watchdog can reliably > restart the PIC and then clean all the ram, > fix all the ports, etc etc and keep going. If > this is done very often, (maybe 1mS?) then > the product is brownout proof. > > In fact, this is better than hardware brownout > protect in some ways as it will will also protect > against very short spikes which can trash ram > and cause latchup etc. It can be a very cheap > and sensible protection for small devices. :o) > -Roman -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads