Has anyone got a 240v GFCI to work properly when wired HHG?
Yes, I have them in my panel and have connected 220V pump motors and and arc welders (both HHG). They don't trip the GFCI.
There's a lot of confusion here and at the Mike Holt forums. It is clear that when wired for 240v you do not need a neutral on the load side (see wiring diagram on p11 from Square-D link below). The manufacturers state that you need neutral on the line side for the GFCI's "test" button to work (see both links).
The test button works by returning some current from one of the load side H terminals through a path other than the other load side H or N terminal. It can connect either load side H to either the opposite line side H or N. Apparently they have chosen to connect to N. OP could connect the breaker's white wire to the feed G in order to have the test button work. It is only connected to G when the test button is pushed and as such, I suppose, might not constitute a code violation as long as the breaker's N terminal is not connected to anything. Or he could wicker his own test button in which a resistor is connected between load side H to G through the button contacts.
Only reason it wouldn't would be because of device failure which is what the test button is there to detect.It's not clear to me if the GFCI will function without the "test" button,..
Just so.... but without it you have no way to know.
That would depend on whether you bond it on the line or load side. If you do it on the line side then it should work. Imbalance (0-sequence) current flows back through the core except in the case of a fault to ground in which some of it doesn't. OTOH if you were to bond on the load side any imbalance current would return partly through the ground wire.It's also not clear if the GFCI "test" button would work if you tied the ground wire to the neutral bus in your spa panel.
If it does you'd be using an uninsulated ground wire as the neutral. Even if it does, it's certainly not to code, and I wouldn't do it.
The main reason for not wanting to go there is that it has been known to cause serious overheating of some participants in this forum.
Should say test device I would think unless they have something in there that won't allow the thing to function unless it senses 120 between 1 or both phases and the neutral and I can't think what advantage there would be to that. More modern GFCI's now launch a 120Hz test signal and expect it to be completely returned through the load terminals. This is, I think, to detect faults between like phases of separate circuits. Maybe it has something to do with that.From Siemens[/URL]: "Note: A load neutral is not required on the circuit. However, the white line neutral (pigtail) must be connected to the panel neutral for the device to function."