CAN Bus Supply Current Limits Clarification

[bakerstu]

Currently, the CAN Physical standard states:

If a node provides power, it shall provide at least 9VDC and no more than 15VDC from zero current draw up to its specified maximum current. Its specified maximum current shall not exceed 500 mA.

This could be interpreted, as written, that 500mA is a hard limit. In this case, using a PTC would mean that a device could not ever practically sized to guarantee supply of 500mA since PTCs can have a lot of variability in their trip points.

I suggest a minor tweak to make the wording more obviously PTC friendly:

If a node provides power, it shall provide at least 9VDC and no more than 15VDC from zero current draw up to its specified maximum current. Its specified maximum current shall not exceed 500 mA. It shall be overload protected to limit supplied current at or below 1A.

Additional discussion can be found here: https://groups.io/g/layoutcommandcontrol/topic/standards_compliance/99496160 https://groups.io/g/openlcb/topic/testing_current_compliance/99507345

[bobjacobsen]

I think this is progress, but there's likely to be concern that simple protection circuits will need some time to limit a short to 1A or below, hence be above 1A for a little while.

Since the goal is to protect the UTP and connectors from heating, perhaps add a time specification? "limit supplied current above 1A to no more than (nnnn) msec"? The time can be taken, with some margin, from typical protection-component data sheets.

A slightly safer approach would be to put the ultimate current limit at twice the specified maximum current, but I think that's not really needed: Most of the supply nodes are going to supply 500mA anyway.

[balazsracz]

PTC's are usually specified to trip after 8 seconds of overload at 2x the rated current.

So this addendum will surely not be in the milliseconds range.