A motherboard short in a Miami server room changed my whole testing approach

Last month I got called to a small law office in Miami after they had a power flicker. Their main server wouldnt post, just fans spinning and no beeps. I spent 3 hours swapping RAM and PSUs before I noticed a tiny scratch on the board near the CPU socket, right where a standoff screw had ground into the trace. Turns out some tech before me had used an oversized screw and it was slowly eating the board. That short was intermittent, only showed up when the case got warm enough to flex the board. Now I pull every board out and inspect the backplane before I even touch anything else. It was a $2000 lesson in checking the physical layer first. Anyone else got a story where a dumb hardware flaw hid behind a software-looking issue?

0 comments

0 Comments

No comments yet

Be the first to share your thoughts on this discussion.