[c-nsp] heat fins popping loose on WS-X67xx cards

[nick hatch]

On Fri, May 29, 2009 at 6:58 AM, Jay Ford <jay-ford at uiowa.edu> wrote:

>
> Based on the responses I've received it seems that this is a fairly common
> failure due to a design flaw.  I got the usual "that's strange; nobody else
> is having this problem" from Cisco.  I now have ample justification for
> telling them "bull".
>

This is a total design flaw.  I've never seen it on Cisco gear, but came
across the exact same failure while replacing an IBM x3550 mainboard last
night. (MT: 79784AU)

The failure mode looks exactly like cimg1690.jpg posted by Christian:
dry/brittle/crystalline looking joint on the horseshoe jumper. There's
barely any solder there.

I can't imagine it's a good idea to design a joint which serves as a
structural element when the applied force is normal to the board, along the
axis of the joint. Can anyone else name a situation where this is done (and
it doesn't fall apart...)? I can't think of one.

Anyone know who does the ODM work for Cisco for these boards? The IBM board
I mentioned earlier is ca. 2007, made by ASUS. I can't believe that firms
started using this brilliant idea independently of each other.

-Nick