Re: How about a wild departure, an actual electronic design question!

From: N. Thornton (bigcat_at_meeow.co.uk)
Date: 10/16/04 

Date: 16 Oct 2004 09:06:22 -0700

"Dave VanHorn" <dvanhorn@cedar.net> wrote in message news:<UoSdnVt0AJqSSu3cRVn-jg@comcast.com>...
> > 2.25 is a lot of amps for a single pin, especially with the same on
> > its neighbor. It's easy top get an oxidation-based thermal runaway
> > thing going.
>
> Check the ratings, Samtec appears to be saying that these are good to 2-3A
> at any reasonable temperature.
>
> See also my other note, I was just informed that only units in vehicles are
> failing, regardless of age, and that "deskbound" ones are not failing,
> regardless of age.
>
> >>Only these pins in the connector are failing.
> >
> > Yup!
>
> Well, these do have the largest current/time product on the connector, but
> others carry roughly the same current levels, in pulses of a few seconds
> when running.
>
> > Don't trust those numbers. The pins heat each other, and may have
> > local hot spots. I don't run small pins like this above 1 amp, and
> > prefer lots of power and ground pins in parallel, with equalizing
> > traces to ensure current sharing if you're really serious.
>
> True, but at this point I don't see why they would heat enough to be a
> problem.
>
> > I think gold/tin may be bad news, too.
>
> It must also take current, and maybe something else, because we aren't
> getting these failures on the other pins of the same connector. 24 pins
> total.

It looks to me like youve almost answered your own question. They only
fail when current is consistently high and in auto environment. Whats
the difference?
current heating
vibration
damp
dirt
grit & muck

If other pins are not failing in auto environment, parallelling power
pins is the simplest workaround, though greater contact forces would
be a more rock solid solution.

NT