ground-planeheatpcbpowerthermal
If you want the best thermal exchange with the air, is it preferable to leave a thermal/ground plane raw or to cover it with solder mask? And how important is the difference?
(Dave from EEVblog is pretty self-confident saying it's more efficient without solder mask: https://youtu.be/q3RhWuXFixU?t=782)
Recognizing that there are a lot of assumptions here about the material specifications involved...
I spoke with two of our partner factories and they both said the same thing:
This is typically a result of rework, where the original soldermask was partially removed for some reason and needed to be reapplied post-hoc.
Should still meet spec -- meaning, there are no anticipated concerns as a result of the mixed process. Some overlap between the masked areas was suggested to ensure that you maintain complete coverage over the full temperature range.
If needed for a "special" application (life/safety critical, aerospace, etc) then recertification/testing is required.
Most of the datasheets I saw had a recommendation, how to use the ground planes around the RF IC.
So my recommendation is:
PS: regarding digital signals - it is not the SPI frequency that matters regarding signal integrity, but the rise and fall times of those signals. Rule of thumb: if your track is longer than 1/10-th of the way the signal travels during the rise/fall time, you might get reflection issues.
Best Answer
The short answer is no it won’t make a significant difference. The long answer is that conduction will be slightly worse with solder mask but radiation will be better, assuming that the bare metal is in good condition - the emissivity of a shiny metal surface is generally low, around 0.1 while that of a coated surface is 0.9 or more. Thus bare metal may paradoxically be a worse heat sink at higher temperatures. This assumes that the bare metal doesn’t oxidise or corrode, unlikely if it’s copper but possible if it’s tinned or gold plated.