[Info-vax] GFCI operation (was: VAXStation 3100)

[John E. Malmberg]

On 1/16/2019 9:38 PM, dthittner at gmail.com wrote:
> 
> Regarding GFCI, the UPS is designed to level out voltage, boosting
> low voltage (brownouts) and "throwing away" high voltage (surges) to
> protect your equipment. These will throw a GFCI when the GFCI detects
> a difference between incoming current and outgoing current. You
> should not plug a UPS into a GFCI if you want power stability, and it
> violates  fire code in some areas.

Do you have a reference to where it violates the fire code?

Any thrown away current eventually would go back into the neutral, not 
the safety ground.  It is the phase shift in when that happens that will 
cause a false positive by a GFCI.

This link is from a UPS vendor:
https://www.apc.com/us/en/faqs/FA158850/

Almost all my networking and computer equipment are plugged into UPSes 
and several of them are plugged into outlets with GFCIs on them.
I have never had a GFCI trip that I did not know the cause of on any of 
them, and it has never been from a UPS.

> In addition, many expansion enclosure power supplies of that era had 
> power supplies that would dump excess voltage on to the ground wire 
> to help protect the equipment from surges, which would also trip the
> GFCI.

Dumping any voltage directly to the safety ground seems to me to be a 
safety hazard.

Has a more reasonable explanation is found a post by "U George" where he 
quotes a response from an unnamed UPS vendor:

http://forums.mikeholt.com/showthread.php?t=69071

"EMI filters needed to control emissions usually include small 
capacitors to ground. These pass small AC currents to ground similar to 
a ground fault, but since the action of the EMI circuit is intentional 
it doesn't represent a hazardous fault. Any single data-processing 
device is likely to leak 0.5 - 2mA during normal operation."

A switching power supply is a place that I would expect to find an EMI 
filter.

In this case it could be the charging of the capacitor in the filter is 
enough to trip the GFCI.  But from all reports I have seen, this is rare.

I agree that Bill should find out exactly why the GFCI is tripping.  He 
may be right that it is a bad GFCI.  It also could be a bad component in 
one of the power supplies.

A replacement GFCI is less than $10.00 last time I looked, and the 
better AFCI+GFCI replacement is about $20.00.

Regards,
-John