On Tuesday, February 01, 2011, Pete Turnbull wrote:
On 01/02/2011 18:53, Tony Duell wrote:
Yes, you're right. I am sure there;s a limit
to the mazximum
permitted leackage current. It's not zero, but ti's probably not
that high either. Mins filters often have cpaciaotrs from the
power-carrying wires to ground, and they will pass some AC
current.
The RCDs on the supplies into our data centre are settable; they can
be set at least as low as 30mA but normally they're set to over an
amp and can be set as high as 6.3A residual current. What's allowed
depends on the context (and that includes the time they take as well
as the current).
The ground fault breaker (RCD by you) on the 480V supply the work's main
datacenter (which is rated at 1MVA, or 1200A currently - an upgrade to
5MW is coming this year) was tripping on us when it was set to 200-300A,
and is now set higher. IIRC, the tripping was caused by current spikes
to ground from "leaky" motors, and probably some capacitors or other
power filter equipment.
Pat
--
Purdue University Research Computing ---
http://www.rcac.purdue.edu/
The Computer Refuge ---
http://computer-refuge.org