Participants: Paul, Larry and Andrew
Got the old PIC16F877 + SJA1000 CAN node talking to the MCP2510 development board, hooray. Hopefully the problems we had last week doesn't mean the FC CAN circuitry is toast... my guess is that it was a wiring problem. I added a small connector on the FC for the CAN bus so that it's harder to screw up the wiring now.
Discovered that the MOPS520 has two 120ohm resistors soldered in parallel on the board. Whoops!
Found out that because the CAN protocol chip is not in the default on state in the PIC18F458, it's possible to complete kill the CAN bus by turning the PORTB pins to grounded outputs. That's bad, we may want to do something in hardware to prevent a locked up node killing the bus. But using watchdog timers may do that for us.
Larry worked on the Twiki logistics team site. Go Larry go.