One can calculate the required trace width. However,
For internal layer trace: How could heat dissipated when traces are enclosed by core and prepreg layers?
For external layer trace: How to account the existence of solder paste over the traces?
It seems to me that these formulas do the calculation for bare copper without conserving core, prepreg and solder paste. Does core, prepreg and solder paste have significant effect on heat dissipation, and should they be ignored or not?
Look up the thermal aspects of PCB's, lots of info out there with todays preference for QFN,s and high powered LED boards and just the general power aspects of PCB design. It is a subject that requires quite a bit of study as there are numerous solutions to thermal problems.
Also look at the history of the IPC-2152 specification and all will be explained: Printed Circuit Design & Fab Magazine Online
How would solder paste affect heat dissipation?, it provieds a good thermal path between a component and the copper pattern.
It seems to me that these formulas do the calculation for bare copper without conserving core, prepreg and solder paste. Does core, prepreg and solder paste have significant effect on heat dissipation, and should they be ignored or not?
Sorry that I used the wrong term. Solder paste eventually melts and would get off the board; I was trying to ask "solder mask", which would be a layer of insulation coating on the board which I thought would cause the heat to dissipate not as good as in bare air, so perhaps the width calculation based on outer layer bare copper needs to be slightly increased to offset for the covering of solder mask?
No the experiments were done with solder mask. BUT if you are affixing a heatsink use BARE COPPER. The heat loss from a track is emmisive so the solder resist does not have that much influence. If it was between an area of copper and a heatsink, where the heat is conducted it would be a big problem.
No the experiments were done with solder mask. BUT if you are affixing a heatsink use BARE COPPER. The heat loss from a track is emmisive so the solder resist does not have that much influence. If it was between an area of copper and a heatsink, where the heat is conducted it would be a big problem.
I was not talking about heatsink. Just consider a track on the surface that is carrying 1A current, and with no heatsink beneath it. If it is a bare copper exposed directly to the air, then Saturn or calculation with other tool for "bare external track" would give the right value; but if there is another layer of solder mask over it so that heat cannot get dissipated freely to the ambient air, how should the track width be determined?
I think we have to make a comparison here on heat conductivity between {solder mask, core/prepreg}:
1. Is external solder mask thicker than core/pregpreg? However, since manufacturer usually vary core/prepreg thickness to match track impedance, and also because there could be variations in soldermask material, this quantitative comparison is not possible until thickness of both core/prepreg and soldermask are determined.
2. Material heat conductivity comparison.
It seems that without these information, it is impossible to give a definitive answer.