TCAN4550: Power Consumption calculation

Hi Tomer,

Your understanding is mostly correct.  There are a couple of things to note, though.

One is that in the dominant state, the transceiver is sourcing current to the bus in order to establish a differential voltage across it.  So, some of the power dissipation will occur outside the transceiver in the load resistance.  Since the maximum recessive-state ISUP is specified to be 15 mA, we can estimate that the current flowing to the bus in the dominant state when the ISUP is 90 mA (max) is 75 mA.  Since the load at this condition is 50 Ohms, the dissipation occurring in the load resistance is ~280 mW.  This should be subtracted from the computed sourced power (VSUP * ISUP) to find the proportion dissipated in the transceiver.

Another is that in most applications, a given node will spend much less time driving a dominant output state than it would in the recessive state.  For thermal calculations, the average power dissipation (rather than instantaneous power dissipation) is typically more relevant.  In normal mode, this average dissipation is the weighted average (based on relative durations) of the dominant-state and recessive-state dissipation levels.

To give a simple example, say a node's output frames are ~100 bits long (this is roughly what an 8-byte classical CAN frame would be without significant bit stuffing).  At 500 kbps this would take 100 bits * 2 us/bit = 200 us.  If the node transmits a frame every 100 ms, then, the percentage of time it is transmitting is just 200 us / 100 ms = 0.2%.  If the data is random in nature, the percentage of this segment where the output state would be dominant rather than recessive would about half.

We don't have a specific MCU recommendation since any MCU capable of supporting the SPI interface could be used.  So, there are probably other application-level factors that would influence the choice of MCU rather than the TCAN4550's requirements.

Regards,
Max