Importance of the Sample Point Configuration in a CAN FD Network
The correct sample point configuration is very important with CAN FD networks. The reason is the Bit Rate Switch (BRS). In order to indicate a switch of the bit rate the BRS bit is sent recessively. The first part of a CAN FD frame, until the BRS bit is transmitted with the NOMINAL BIT RATE. The bit rate is switched if the BRS bit is recessively, until the CRC DELIMITER is reached or until the CAN FD controller sees an error condition that results in the starting of an ERROR FRAME.
When the bit rate is switched at the BRS bit or at the CRC DELIMITER bit, it shall be
switched immediately after the SAMPLE POINT, causing the length of these two bits to be intermediate.
Now let’s have a look what happens if two CAN FD nodes are interacting in a CAN FD network with different sample point configurations.
ECU1 -> Sample Point at 60% is sender of a CAN FD frame and indicates a BRS
ECU2 -> Sample Point at 75% is receiver of the CAN FD frame
In addition to that the relation of the NOMINAL BIT RATE and the DATA BIT RATE is very high: 65 kbaud to 1000 kbaud.
In this case the BRS bit is much longer for the receiver than for the transmitter. For the receiver the BRS bit could look like this:
This means that within one bit the receiver sees a recessive and a dominant part. Now the bit is being sampled at the dominant part (in this moment the transmitter already switched to the faster bit rate and continues sending other bits). This leads to the assumption for the receiver that there is no BRS at all and he continues to sample with the lower bit rate which automatically will lead to errors.
See also article 856.