🚀 go-pugleaf

In article <334@uncmed.med.unc.edu> uchuck@uncmed.med.unc.edu () writes:
+---------------
| ... a DPDT main relay with the PC and its peripherals on the moving
| contacts and the power line on the normally open contacts, I had the idea
| of connecting the normally closed contacts together and then to ground in
| the hopes of preventing any lightning hits getting to the PC.
| 
| Now, I begin to wonder... What is the effect of shorting the power
| cord(s) of the PC and associated equipment together and to ground, even
| under normal (read operator induced) conditions????  Can anyone think why
| this might be a problem?? | Chuck Bennett
+---------------

CAUTION! You could actually create quite a safety hazard! DPDT relays have
been known (depending on the brand) to have fault modes (or even normal
operation!) in which one pole switches before the other. It all depends on
the spacing between the contacts, how much the contacts "stick" when opening
(due to micro-welds due to arcing when closing), the stiffness of the piece
of fiber/plastic which mechanically ties them together, etc., etc. Now if
you simply short the normally-closed contacts, you're pretty safe against
an out-of-order switching, since until both contacts go to "closed" the
other end is open. But if you also ground this shunt, then an out-of-order
switching can momentarily connect the "hot" wire to "ground" (NOT "neutral"!),
thus creating a dangerous "ground fault", exactly the opposite of what a
safe system would do. In fact, this is the very situation "ground-fault
interrupters" are designed to prevent.

[Why is it dangerous? Well, all "grounds" aren't the same, and if you happen
to have one hand on a "hot" ground and another on a "cold" ground...]

And even without grounding the normally-closed strap, there's another relay
fault mode to worry about: transient bridging. Again, depending on the relay,
it's load capacity, history, etc., it can be possible for the normally-open
and normally-closed "stationary" contacts to be briefly connected together
through the "movable" contacts (some audio relays are even deliberately des-
igned this way, to avoid switching "clicks"), which in the wiring connection
you propose would briefly (and then possibly permanently!) short the "hot"
power line to "neutral". Heat, light, (hopefully) blown fuse/breaker, and
(hopefully no) fire.  For this reason, the UL (or CSA or VDE) would probably
never approve the power connection you suggest above...   ;-}   ;-}

[Hmmm... There just *may* be special-order (and exPENsive!) relays which could
safely be used in this case, but a garden-variety DPDT mains relay isn't it.]


-Rob

-----
Rob Warnock, MS-9U/510		rpw3@sgi.com		rpw3@pei.com
Silicon Graphics, Inc.		(415)335-1673		Protocol Engines, Inc.
2011 N. Shoreline Blvd.
Mountain View, CA  94039-7311