0

Using GFCIs with transformer or other high-inrush current loads is problematic, because in some designs the differential transformer core can locally saturate and produce an invalid output in the sense winding, tripping the GFCI - without any ground leakage.

What are the standards/requirements applying to GFCIs (in various markets) in terms of maximum peak current they must handle without spurious tripping? I'd expect this peak rating to be higher than continuous rating.

Most GFCIs on the market do not specify such current in their spec sheets/datasheets, thus I presume that the actual requirement would be buried in a standard somewhere...

This question is about applications where a load may present, say, 10Arms continuous current, but have 10x+ higher inrush current (e.g. 100A or more). In practice, I've seen many GFCIs rated at 2x the continuous current (20A in this example) trip spuriously at 100A half-wave inrush. I would be surprised if this requirement wasn't addressed in the standardese, thus the question.

Kuba hasn't forgotten Monica
  • 32,734
  • 1
  • 38
  • 103
  • This question might be better to move to Home Improvement where they have more people that are electricians. – Voltage Spike Jul 06 '21 at 16:49
  • I already asked the electricians. As a rule, they are standardese-incompatible. They hate the stuff (and reasonably so - it's impenetrable unless you have lots of time to study the legalese). – Kuba hasn't forgotten Monica Jul 06 '21 at 16:59
  • I agree there should be a specification available. But I also know that GFCIs don't just trip on ground-fault leakage, but also on some overload value, too. (So, I've read before.) I believe an experienced electrician will know something about this. I'm not one, though. The fact that you see it as "spurious" does worry me that it's just a practical "known from experience" and not a specification, though. – jonk Jul 06 '21 at 16:59
  • Spurious here means unintended. I'm pretty damn sure GFCIs must be tested for pulse/overload conditions, and up to a certain limit they must not trip. – Kuba hasn't forgotten Monica Jul 06 '21 at 17:01
  • @jonk *GFCI's trip [...] also on some overload value, too* - that's always the case when using typical compact differential transformers. They have unequal coupling constants from primary windings to the single secondary when the currents are high enough. **At high currents the sense transformer has poor common mode rejection - either at the fundamental, or in low harmonics**. A GFCI sense transformer is basically a balun :) – Kuba hasn't forgotten Monica Jul 06 '21 at 17:05
  • @Kuba Are you talking about trip current to ground (from what I remember it's in the uA range)? Or Max current that can be handled by the device? Keep in mind that most circuits can only handle 20A max before the breaker trips, I think this question could only be answered by knowing the code, as many GFCI designs will probably rely on where they fit in the overall design of the installation and code. – Voltage Spike Jul 06 '21 at 17:07
  • @Kubahasn'tforgottenMonica What you write makes sense. Both in that they must be tested for over-current (I believe that must be true) and that their sense transformer isn't wound and tested against NIST-traceable standards for perfection, so there must be a magnitude that is sufficient to trigger them, anyway. Yup. Makes sense. I suspect the over-current is tested and that the residual imbalance isn't, unless it affects the tests they do try for other reasons. – jonk Jul 06 '21 at 17:13

0 Answers0