Well, if it was me, and I wanted close to 100% assurance that my data was good, I guess I'd go a little bit crazy and do something like have a 'packet' of X number of bytes, each packet for one temperature reading for each 'location'...Originally Posted by ruijc
packet start ID (example of $FF)
address X.lowbyte
address X.highbyte
data byte
same data byte (inverted)
address X.highbyte (inverted)
address X.lowbyte (inverted)
packet end ID ($00, which is the packet start ID inverted)
So, I'd end up with an 8 byte packet, everything is sent, then it's all sent again but inverted at the bit level. The packet start ID might be an $FF (because the temp's will never be that high), the addresses might be that high, but if you got that then the packet end ID wouldn't match up.
It's a bit overkill, but I don't see why it would work. And really, I think I'd trust my wiring enough to just send a couple of $FF's as a start marker, then the whole recorded memory from start to finish, maybe with a couple of $FF's as an end marker.
It's all up to you. You get to make RSP... ruijc's serial protocol.
Bookmarks