You've got that right. Scaring the crap out of me.

Will check that today.

Yes its grounded. Both are.

That is sooo true LOL. 1500v rocks your world too, got some scars to prove that. LOL

How old are your pumps? Id be calling Tunze. 35v leakage is NOT COOL. Sounds like the case on the pump has a defect in it, expoing the windings to the water. If the GFCI tripped plugging the Tunze in when ground probe installed.

The light shade should be grounded, and you stated that you got shocked with light out as well, so I could rule the light out. I have seen instances in HO-T12 lights, where the ballast would induce 40-50V into the reflector/shade. But that was with a Magnetic ballast installed in the light, inducing a voltage into the shade, not acctually shorting or leaking. Anything that has a magnetic field, can induce electricity into metal near it.

GFI only compares the current through the 'hot' wire against the current in the 'Neutral', if the current through them is different as in some current is leaking to ground, the GFI detects the difference and trips, so it actually doesnt need the ground connection to work its just safer to have it properly grounded as this will reduce the shock risk for a person handling the device.

120 or less is not kids play when its your forehead thats sparking. :lol:

Doug, I know this all serious and stuff, but I just can't help laughing my a$$ off !! Maybe you need rubber gloves and a rubber hat!

When checking for stray voltage with a volt meter, where do you put the probes. In the water? I've never had to do this but am still curious of the procedure.

:robot: Thats me if it gets me again. :lol!:

I started out, {because it was the grounded pendant and water contact that got me}, with a measure from the pendant, {red probe} to the water [black probe, common}. When the ground probe was installed, I dont get a reading now.

Unless you forget the ground is out and you're forehead arcs across to the pendant. :surprise:

Just to clarify that up.