<whitequark>
other than the ESD issues i think everything that needed to be fixed for revC1 seems to work
<tnt>
Well, backward soldering shouldn't be an issue wit PnP.
<whitequark>
well no
<whitequark>
there were two issues
<whitequark>
one is a backward placed MG2040
<whitequark>
and the other is *something else* that also traced back to an MG2040
<whitequark>
that impacted A1A3A5A7
<whitequark>
it was not soldered backwards...
<whitequark>
seems like there was a short to ground
<whitequark>
i should have x-rayed it i think, but i got impatient :/
<tnt>
Oh, ok, I though only 1 side was bad.
pke has joined #glasgow
<pke>
@tnt Depends on the packaging. Tubed parts can go 180 degrees. Also square parts can go 90 degrees over if the pickup is bad and vision corrects in the wrong direction. So yeah fab is hard. Also a board as complex as glasgow human visual inspection easily misses errors.
<pke>
Does anybody know something about open source automated visual inspection? Newer seen an open solution before.
<tnt>
pke: those are in reels.
<whitequark>
pke: and they are not square. they are VERY much not square
<whitequark>
the aspect ratio is like 1:4
<whitequark>
as for human visual inspection... yes. unfortunately.
<tnt>
if the gnd part weren't soldered/exposed at all, it wouldn't even have soldered 180deg appart, since it's not symmetrical.
pie_ has quit [Remote host closed the connection]
<pke>
Sometimes at work i have to fix failed boards after testing and at the end of batch the cursed ones remain. Are cases when IC-s get placed upside down, legs in the air, and remain glued to the board because of the flux. From far seems fine. Until i check up-close and FCK.
<pke>
My point is mistakes happen even if a robot does the job.
<tnt>
Sure, but hopefully not 1 in 4 :p
<whitequark>
well... i may have just been unlucky
<whitequark>
4 is a very small sample size to make any conclusions from
<tnt>
Yeah sure, and glasgow is also not fun to assemble by hand, _lots_ of parts and bom lines :p I'm very thankful to esden to get one :)
<whitequark>
indeed
<gruetzkopf>
does unit 4 work?
<whitequark>
. o O "unit-04"
<gruetzkopf>
2019-05-12T00:04 didn't happen :(
<whitequark>
i'm confused
<whitequark>
oh
<whitequark>
oh shit
<whitequark>
gruetzkopf: i think esden didn't make enough for that
<whitequark>
but, i asked greg to make two more for some people
<gruetzkopf>
yeah 4 is a bit too little for that
<whitequark>
i think he might be fine making three? try dm'ing him on twitter
<electronic_eel>
esden said he might make one for me to test the test jig, but he wasn't sure if he'll have the time for it
<tnt>
somehow unit-04 meant something else to me :)
<whitequark>
yes :p
<gruetzkopf>
we're not doing S² powered glasgows anytime soon though?
<tnt>
what could possibly go wrong
<whitequark>
S2?
<whitequark>
oh yeah
<electronic_eel>
marcan picked the wrong silkscreen picture for that