SLLA651 April 2025 TCAN2845-Q1 , TCAN2847-Q1 , TCAN2855-Q1 , TCAN2857-Q1
The CAN bus is monitored for 7 different fault conditions on the TCAN28XX line of devices; these are all captured in the CANBUS interrupt register (54h).
| Bit Field | Flag | Description |
|---|---|---|
| 7 | UVCAN | UV Event on VCAN |
| 6 | RESERVED | N/A |
| 5 | CANHCANL | CANH and CANL shorted |
| 4 | CANHBAT | CANH shorted to battery |
| 3 | CANLGND | CANL shorted to GND |
| 2 | CANBUSOPEN | CAN Bus Opened |
| 1 | CANBUSGND | CANH shorted to GND or both CANH and CANL shorted to GND |
| 0 | CANBUSBAT | CANL shorted to VBAT or both CANH and CANL shorted to VBAT |
From the interrupts that are monitored this needs to be clear that the fault types are not exhaustive and an interrupt alone generally does not give all the information an end designer wants or needs to know. To highlight this concept, ten different fault cases can be used to show what interrupts are registered and what information can be gathered from each fault at the SBC.
| Fault # | CANH | CANL | Fault Detected | Comment |
|---|---|---|---|---|
| 1 | Open | Open | CANBUSOPEN | N/A |
| 2 | Open | Normal | CANBUSOPEN | SBC cannot tell difference between fault 1 and 2 |
| 3 | Normal | Open | CANBUSOPEN | SBC cannot tell difference between faults 1,2, and 3 |
| 4 | Shorted to CANL | Shorted to CANH | CANHCANL | N/A |
| 5 | Shorted to VBAT | Normal | CANHBAT | N/A |
| 6 | Shorted to GND | Normal | CANBUSGND | N/A |
| 7 | Normal | Shorted to VBAT | CANBUSBAT | N/A |
| 8 | Normal | Shorted to GND | CANLGND | N/A |
| 9 | Shorted to GND | Shorted to GND | CANBUSGND | SBC cannot tell difference between faults 6 and 9 |
| 10 | Shorted to VBAT | Shorted to VBAT | CANBUSBAT | SBC can’t tell difference between faults 7 and 10 |
With interrupt map explored for CAN faults this needs to be clear that the SBC can tell the controller through an interrupt that there is an issue – and give some information on the fault – but this does not necessarily give all pertinent information about location and does not even give enough information to determine the pin that is at fault depending on exact fault scenario.
One fault that hasn’t been talked about is the dominant timeout fault. At the logic input pin, CTXD, if the level is set near GND and the device doesn’t detect a rising edge on this pin for 1ms (min) to 5ms(max) the CAN driver is disabled to prevent stuck bus conditions.