Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

RCD upstream of UPS tripping

  1. Oct 11, 2007 #1
    Gents, Ladies,

    I have a design where I have:

    AC -> RCD -> UPS (dual conversion apparently) -> PowerStrip -> Equipment

    If I induce an earth leakage fault downstream of the UPS (i.e. at the powerstrip), the RCD upstream of the UPS trips.

    The UPS is in the "normal" not bypass state, and the UPS is supposedly dual conversion.

    I cannot understand how this can happen unless the UPS is not true dual conversion !

    Can someone confirm or debunk my understanding here please ?

  2. jcsd
  3. Oct 11, 2007 #2


    User Avatar

    Staff: Mentor

    Another name for the RCD is GFCI, BTW. Does the GFCI trip if the UPS is in bypass mode too? I agree that if it really has switching power supply transformer isolation between the AC Mains with the GFCI, and the load where you introduce the ground fault, that the GFCI should not see anything different, other than a slight change in its AC load (no common-mode load change).

    On second thought, it might see a common-mode load change because of the winding-to-winding capacitance of the switching transformer. Does it only trip the GFCI if you introduce a hard fault to ground (like with a wire)? What if you use a 10kOhm resistor instead? Try some resistor faults right at the GFCI to see where its sensitivity threshold is, and then do the same downstream of the UPS. It may just be the capacitive common-mode transient from a hard wire fault downstream of the UPS that is tripping the GFCI.
    Last edited: Oct 11, 2007
  4. Oct 11, 2007 #3

    Thanks for your reply !!! I new that the RCD was called something else - but couldn't remember ...

    Being an Aussie - RCD = Aussie; GFCI = American ??? :-)

    In bypass mode, the RCD/GFCI trips as expected.

    In normal mode, the RCD/GFCI trips (not really as expected - but noting your second though!)

    I am using a tool to induce the earth leakage fault - it allows me to change the mA-erage to see how quickly an RCD responds. I came across this situation when I was testing to see if the UPS had anything in it to cope with a earth leakage fault (i.e an experiment) ...
    Anyway, I'll look at the tool and also try as you suggest ...

    Thanks !
  5. Oct 12, 2007 #4


    User Avatar

    Staff: Mentor

    Glad to help, qbnchopper, and welcome to the PF. It's a great place.

    I'd really be interested in the results of your experiments. Your initial test results are a bit non-intuitive for me as well.
  6. Oct 12, 2007 #5


    User Avatar
    Science Advisor
    Homework Helper

    The UPS loops the ground circuit thru to the UPS outlet.
    I'm not entirely certain, but I think they also loop the neutral line thru.
    If so, then at some point in a normal mains circuit the neutral and ground lines have a common bond.
    Then a leakage current from the UPS hot to ground will loop back thru the neutral to complete the circuit. The unbalanced current in the RCD (GFCI) neutral will cause the breaker to trip.
  7. Oct 15, 2007 #6

    Thanks and Interesting .... in my case the earth and neutral are not bonded .... however, on another problem that I am working that is related, earth/neutral are bonded upstream of the ups .... have to have a think about this ....

Share this great discussion with others via Reddit, Google+, Twitter, or Facebook