← Back to team overview

kicad-developers team mailing list archive

Interesting consequence of "subtract soldermask from silkscreen"

 

A warning to anyone making PCBs - and maybe to get some devs thinking, I'm not
sure if there's a better way to do this:

For those who didn't know, "subtract soldermask from silkscreen", for Gerber,
is implemented by plotting the silkscreen, then switching plot modes to 'clear'
with the %LPC*% command and plotting the mask over the top. This is like
"drawing" the mask with an eraser, clearing out the silk layer.

I sent a board off yesterday, to Elecrow, which had an empty back silkscreen. I
sent the file anyway, to make sure they were aware I indeed wanted it empty and
didn't just forget the file.

Got an email a few minutes ago telling me my silk and mask layers were the same
and asking me to check them. Checked the files I sent them and confirmed this
wasn't the case, my files were fine - though I was led on a wild goose chase by
the fact that the silk layer contained the whole mask layer in negative. What I
believe happened is that their software saw a layer with all negative drawing,
and automatically inverted it.  Probably meant to detect gerbers generated by
old-fashioned EDA packages or packages with old-fashioned settings that output
power plane layers in negative.

Something to watch out for - if your silk layer is empty, either don't send it,
or don't use "subtract soldermask from silkscreen". I wonder if we can do
anything about this from the software side, like not generating the mask
negative if the silk is empty.

-- 
Chris



Follow ups