Getting a small shock in a cabinet with UPS

[Stonent]

I'm not thinking very well today, I didn't get much sleep. I was removing a PC from a cabinet today and had one hand on the metal case of the PC and the other was removing cables from the back and as soon as I touched a metal part of the cable I got a small shock. (I think it was the DVI cable to the monitor) Not a full outlet voltage shock, but enough to get my attention.

The PC I think is on a UPS, so I think there may be something to do with that that caused me to get shocked.  This cabinet also has power supplies in it to power other things that are attached to it, but nothing over 12V (slightly possible 24V)

I'm trying to wrap my head around where the voltage potential came from and I think it is the UPS doing it. I remember in the past dealing with an issue where people kept getting shocked when using a computer and they removed the UPS and the problem went away.

I remember a while back Dave's UPS video talked about the different ways to design them, but I can't watch the video at work.

Could just touching an isolated ground and an earthed ground give the shock?

EDIT: Ok I think maybe I have the scenario figured out. The monitor cable is earth grounded because it is not on the UPS. The computer chassis is isolated because the UPS is generating its own AC. The monitor cable was essentially providing the earth ground for the UPS and removing it left my hand as being the path of least resistance?

Should it work that way or do you think there may be something else going on?

[Rerouter]

even with online UPS's they tend to link the earth grounds, so there should be no reason for any voltage presented on the computers chassis ground,

[TheBay]

Hot and neutral reversed anywhere?

[Stonent]

Ok turns out this was the same cabinet that had the UPS that was shocking people, but when it was removed, we didn't get any reports of shocks.

I went to the maintenance area and told them I think we had an open ground and was told to e-mail the supervisor. They called me a few minutes ago and said the was no continuity between the cabinet (which has several bare metal power supplies mounted to it) and the ground pin on the outlet strip that the computer was plugged into.

So apparently when they were asked last night to check the cabinet out to make sure it was safe for me to stick my hands into, they saw the computer, and didn't check it.

They seem to have a fear of computers. We have another area with a device that has a pneumatic release mechanism. The mechanism is just to make sure a particular board is locked into place during firmware programming. They hold the button and insert the board, and when they release the button, air pressure causes the claws to retract pulling the board all the way in to the programming header. When programming is done, they press the button, the claws release, and a spring loaded pin pushes the board back out.

I'm about 90% sure that button has no electrical connection, it's just a mechanical valve. Yet every time they have a problem with the boards not releasing, they call for IT to look at it and then we have to have maintenance look at it because we hear a loud hissing from behind the unit.  And a few hours will pass and we'll get another call saying it still isn't working. When we ask them if maintenance looked at they'll say "They told me it was a computer problem"

 

[SeanB]

Your guys need training a little, I have stuff with microprocessors in it, called inverters. Tell them that a computer is just a very fancy inverter with no motor attached, or a PLC without a case and IO and they will be fine with them.

[Rerouter]

or more realistically, tell them its like a toaster, if the casings not grounded people die when they try to remove the bit of toast(ed electronics)

if they have half a clue about electricity they should already know all mains powered devices with a metal casing need to be mains earth referenced....

[Stonent]

More info:  I was right... I had a theory that in the absence of a chassis ground, it would try to find another path to ground and it did. The ground pin on both ends of one of the USB cables is completely vaporized.

That USB device also happens to connect to the large assembly being tested. It also made me think of a scenario where when connecting the device if positive connection is made before the negative one, the USB port will be the path to ground and the entire assembly will try to power itself from the USB ground pin.

Cue Dave's video on how to not blow up your scope!

[Bored@Work]

They seem to have a fear of computers.

Now, who might have instilled that fear in them? Wouldn't have been the computer guys, would they? 

[Stonent]

They seem to have a fear of computers.

Now, who might have instilled that fear in them? Wouldn't have been the computer guys, would they? 

It goes both ways. They fear anything that doesn't say Allen Bradley or Rockwell Automation on it. And I fear anything that DOES.  But since they have to plug their tablet PCs into the PLCs or PanelViews we have to make sure that works. 

[Stonent]

or more realistically, tell them its like a toaster, if the casings not grounded people die when they try to remove the bit of toast(ed electronics)

if they have half a clue about electricity they should already know all mains powered devices with a metal casing need to be mains earth referenced....

Well not really our local people's fault for that. I had no idea that this cabinet wasn't tied to ground and neither did they until I got shocked. The guy that installed the ground connection today was about as irritated as I was.

[SeanB]

Blame whoever either designed or built it that forgot that little very important wire. I always add them to panels if I open it and there is not one, even if the hinges are steel and are pretty good at conducting. You find a lot of equipment where those are either broken or missing.

[madires]

Most 19" cabinets come with ground straps for frame, walls and doors. Of course you're expected to connect the cabinet's ground terminal to the building's ground bar.

[Stonent]

So yeah since the USB ground isn't fused, the cable will start getting hot.

[SeanB]

Think that cable was operated a little above ratings, like at room light ratings.