Data Center Battery Ground Bus

[Dumbo2929]

So a contractor says he measured about 80V AC on the DC battery ground bus (I believe it is 48VDC battery system) for a good size college data center. I'm not sure where the reference for the measurement was made. Anyone ever heard of something similar??

 

[itsmoked]

Sure! If the DC ground bus is being floated you could easily have ~half the local line voltage appearing on it. You could test this by hooking an AC trouble light from that ground to an earth ground. The bulb may partially light if things are really dangerous or it would just drop that measured voltage to zero if it's a low leakage.

NOTE: When 'Data Center' and 'Batterie' are uttered in the same sentence you better be very careful with even tests like that above, as some ground fault system could trip killing your data center.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[pwrengrds]

The last time I saw AC riding on DC like that I found the battery was open circuited and it was comming through the charger. When I turned off the charger the DC went away and we found a battery connection was corroded open.

 

[ruggedscot]

Really dont like the idea of linking that bulb to ground mate - Most UPS units would pick this up as a ground fault and you could be looking at the unit going to bypass and running the facility at the mercy of the utility....

Hmmmm

First what type of UPS is it ? 48v DC seems a little light for a good sized data centre.

Need to know what UPS you are running, then take it back from there, if you are running a motor gen set type then the thristor driver for the motor may be imposing an ac waveform on the dc link - this would be a duff connection but a good ups would alarm due to this - why was it being checked do you have a problem with it?

 

[swn1]

Last time I saw AC on the DC bus it was in a lineup of medium voltage switchgear with 48VDC floating control power.

Somebody had modified one of the units to be a motor starter. In the process they re-wired about half of the control panel to run from the local control power transformer.

Typical sloppy practice, they didn't clearly differentiate between the "return" side of the DC ladder logic diagram and the "neutral" side of the AC ladder logic diagram. They were tied together somewhere in the rat's nest and that made the control power bus for the entire lineup hot.

The "neutral" AC leg was actually one side of a 240V secondary with a grounded center tap.

 

[ruggedscot]

Shows just why its better to have one side of your control circuit grounded, also the dc should always be grounded aswell. Hate floating systems especially in computer rooms.

Rugged