Earlier this week we had a power outage and when everything came back online, everything came back online except…the wireless access point (WAP). Ours was a Cisco WAP321. Now it’s a paperweight. After examining the setup in our server room, it didn’t seem plausible that the power outage caused the WAP’s demise since it’s plugged into an uninterruptible power supply (UPS) that’s well-grounded and has built-in protection against transient voltages (spikes and surges). Then I remembered the shocks…
Every time I unplug my laptop from the docking station, I get a small electric shock. I’ve always assumed it was static from shuffling my feet on the carpet too much. But then I started thinking, what if there’s a voltage difference between some of the equipment in the server room? That would cause the shocks I’m always getting, and would also explain how the WAP got fried despite being protected from the outside world by the UPS.
So I got out my trusty multimeter and tested voltage between each device’s case and a known common ground point (the casing of the UPS). Only one device appeared to be carrying a significant voltage difference – the Sonicwall appliance was the only device not mounted to the rack. So I dug up some old rackmount brackets, filed off the powdercoating near the mounting screws, and mounted the Sonicwall appliance to the server rack. When I re-tested, the voltages were all about the same and the resistance between cases and ground was near zero. So far so good – hopefully this keeps the next WAP from getting fried in an electrical storm.