The measurements look good and the GFCI shouldn't trip at 240 microamps which is what the 120V with a 500K load represents. From what I've read they should trip around 2 ma or less depending on the class of GFCI. Is the green ground wire tied to the light housing like it should be? I'm also wondering if you're running into a situation where the resistance shown by the VOM, which measures resistance with a low voltage, is not what the leakage resistance is with 120 volts on the circuit. That's 170 volts peak. Think of an extreme case where you apply a high enough voltage and any insulation will eventually break down. So, the trick is to somehow measure the leakage current in real life with 120 VAC applied. If your VOM measures AC current you have it licked. Power the light without the GFCI and open the green ground wire, insert the VOM and measure the AC current flowing in it. Should be ZERO. Another way if your VOM doesn't measure AC current is to open that green again and insert a 1K resistor, or anything reasonable. Measure the AC voltage across the resistor. 1 ma would equate to 1 volt across a 1K resistor. E=IR. I think I would prefer measuring this way as all VOMs don't measure AC current and although they should be protected against over-current, less of a risk in damaging the VOM in the event of a short.
Sounds like you know what you're doing but I have to say it anyhow, use common electrical sense when doing this stuff so you don't light yourself up and it's always good advice to have someone with you.
Hope this helps.
Al
Edit: Thinking a little more, that 500K you measured should be more like infinity or at least in the several megohm range. I think you do have a leakage problem that shows up with full voltage applied like I stated above.
Al
Bookmarks