Hello,
I am implementing a 3.3V CAN bus using SN65HVD233 transceivers. There is a possibility we will have a 5V transceiver on the bus for brief periods of time. Some of the boards I am laying out are very dense, which leads to 3 or 4 mil trace widths to keep the 120 ohm differential impedance on 0.5 oz copper on internal layers.
This should be fine for normal operation, but the fact that 250mA is called out as the short circuit current alarms me a bit. On the 3 mil 0.5 oz traces, a continuous 250mA current could induce a 60C temperature rise! There are some areas of some boards where two busses intersect in X-Y position on different layers, and if there were a short on both at the high extreme of our temperature range, this could push the limits of FR-4.
But is this actually something this bus would see in a short? The 250mA spec looks like it corresponds to a 12V system, but we'd be looking at a 5V transceiver on the bus in the worst case. Would this reduce the short circuit current to a maximum of ~104mA? That would help the thermal situation dramatically (basically a non-issue, ~10C rise on the 3 mil trace). I know duty cycle plays a role too, but I'd rather not rely on software behavior to ensure the safety of the equipment.
