18 Nov
2006
18 Nov
'06
1:48 p.m.
Well I'm not sure if that's better. Granted
there
would be less of a shock, but wouldn't it be even
better to insure no heat gets to the chip at all?
Heat will get to the chip. You can not stop it travelling up the metal
pins to the die. The thermal resistance of the metal is so much less
than that of the surrounding material (typical epoxy for DIPs) that
heat will mostly ignore anything but the metal - and the metal goes
straight to the die.
Remember that we are talking about tiny parts here with tiny
tolerances, so even a tiny force can have a big effect. The big effect
that can happen is the hairline fractures I mentioned earlier. Warming
up a chip first and slowly (obviously do not go overboard) will give
all the internal parts a chance to expand easily, and when the heat
from the solder operation hits those parts, the shock
will not be as
bad.
Even Radio Shack used to sell these little aluminum
clips that presumably would get clamped onto a pin for
desoldering. That's what I seem to remember anyway.
Heat sinks are good as well, just most IC packages can not accept them
well, and when they do, they often get in the way. About the only
place you can really use them is on the old TO- cans - and there it
would be better to just clamp on a normal thick heatsink for the case.
The advantage to using these is twofold - the heatsink is designed for
the case so it works well, and it gives you something to grab hold of.
--
Will
QA auditor U S Robotics circa 1996