ard(a)p850ug1.demon.co.uk (Tony Duell) wrote:
I wonder if you're actually slightly melting the plastic package and then
disturbing the bondout wire from the LED 'chip' to the other terminal,
perhaps by pulling on the leads while it's still hot. That could make it go
intermittent.
Hi Tony
Yes, I suspect this was the failure method. I was
working for a company that made boards with 60 or so
status lights. Before heat sinking, we would get about
1 in 200 or 300 failures. After heat sinking, we had no
failures. Even 1 in 200 to 300 is too high. If in fact this
is the failure mode, it means that some of the tested
units are getting into the field with flaky connections.
With a 3 year warranty, this is a pain.
Dwight