Checking ground current

Question

Background:
I had a contractor finishing my basement, and part of that was installing underfloor heating using several heating mats in parallel.  It worked initially but soon after it just started tripping on a ground fault (built in to thermostat).  I think my contractor is a little out of his element so I took some time to try diagnose the issue.

There are no obvious issues in terms of element resistance or total current draw, so I started to check for ground leakage.  I don't have a megger and so I used my DMM to check resistance to ground for each element, nothing measurable.

Here is my question:
As a more detailed check for leakage to ground I used my DMM to measure the current.  For each mat I DISCONNECTED the neutral side, connected one lead of my DMM to the hot side of an element and then touched the other lead to a hot wire (not from thermostat).  So in theory this should just measure current leaking from the element to ground, right?

I ended up measuring between 4-8 ma for each element.  Seems a little high if this was correct.

ArchonOSX · Answer

GFCI's are set to trip between 4-6ma of leakage current. This is a pretty sensitive setting to measure.

It is possible the GFCI in the thermostat has gone bad. Like all electronics, it is prone to failure at some point.

I would try a new thermostat to see if it holds. If not, you need to investigate the wiring further.

Good luck!

batsplatsterson · Answer

Your test is good as long as your meter is accurate in that low range.  Most clamp meters are not, although they make clamp meters for instrumentation that will measure in that range.

With the leads in series as you did your test, many decent quality multimeters will be accurate in that range.

xeeka · Answer

As long as there are hot, ground and neutral wires to the heating elements that are running  close to each other - e.g. in a long Romex cable or conduit -, the measured leakage current could result from long cables to the heating elements since there is some capacitance or series capacitances between wires - even if neutral is disconnected at one location. If possible, a double disconnection of the neutral wire(s) - and step by step of the ground wire(s) - (close to the GFCI and close to the heating elements) could yield more information. 
Different locations of the DMM - close to the GFCI/ panel or close to the heating element - would also yield more Information concerning the location of the resistances/capacitances.

A DMM would always  measure infinite resistance with pure capacitors since it works with DC ("I used my DMM to check resistance to ground for each element, nothing measurable.")

The reason why it worked for some time and then tripped more often could be a change of the insulation of either the cable to the heating elements or of the heating elements themselves, f.e. rodents, higher moisture levels, dirt on connections, higher temperatures etc.

Again a DMM could fail since this additional (parallel to the capacitance) resistances could be very non-linear, i.e. the low DMM test voltage (<10V=) could yield an infinite resistance, although it is finite at higher voltage. This is why insulation tests are normally done with high voltage, >50V= or 200V=.

Checking ground current

3 Answers

Add your own answers!

Ask a Question