Spoke width > minimum zone track width should be a soft constraint

[Michael Joya]

Related to wishlist #1661963 "Ability to specify number of thermal relief spokes":

Given that there are usually multiple thermal spokes (currently 4, soon user-configurable), should the spoke width really be lower-bounded by the track width of the (zone, net)?

If current follows cool paths through the copper (a la "skin effect"), one should be able to use slightly smaller width "traces" in the spokes since the current will distribute over the spokes as they increase in temperature. Smaller spokes would improve the thermal soldering insulation that they provide.

What I propose is that the hard constraint be converted into a permissive warning, then a new hard constraint could be added that is more loose:

Change

spoke_width > zone_track_width

to:

spoke_width > zone_track_width / spoke_number

Or, a version with a healthier safety margin:

spoke_width > zone_track_width / min(1,spoke_number-0.5)

[Seth Hillbrand (sethh)]

This is a larger request than might be first apparent. We draw the zones with the line width first and then fill. Since we cannot make a section smaller than the line width, we can't make spoke smaller than the line width (they wouldn't be drawn).

This requires re-working how we create/check zones all together. Not impossible but a big change.

[Jeff Young (jeyjey)]

Actually we need to change it to what Seth says. Right now (believe it or not) we draw thermal reliefs by knocking out the 4 pie-slices *around* the spokes. This is why we can't handle spokes on custom pad shapes.

But yeah, big change.

[Jeff Young (jeyjey)]

Minimum zone width and minimum track width are not the same thing. Minimum zone width is as much about peeling copper as it is about current-generated-heat.

The zone algorithm has been rewritten for 6.0 to be closer to #1. The minimum zone width can be set to less than the track/net width, but it still has to be at least as large as the spoke width.