welding vs GFI

GFI circuits hate welders.

I trip one of the GFIs in my box when I use my TIG, the MIG doesn't seem to do it.

The same happens to me. Of the four separate GFI circuits in my shop, one trips only when I TIG. The other three almost never trip with use of any tool or the similar lamp that I have mounted to my welding bench.

GFI (and security lights) are susceptible to the HF used with TIG.
It helps to eliminate any ground loops. In particular, have the workpiece grounded to the welder -- and no other ground connection. In the case of a metal welding table, if that is grounded to the welder then try NOT to have any other ground connection, as thru the frame of the lamp clamped to the table to the gnd of the GFI. OK to ground the lamp thru it's 3-wire plug, but insulate it from the metal welding table. Connect the welder's ground as near as possible to the welding site.
I have no problem with GFI's or security lights, even with TIG. (Miller Dialarc HF)

Ditto for the VFD (Variable Frequency Drive; synthesizes variable frequency 3-phase from 230v single phase) on my Delta wood-turning lathe. Pops GFCI every time.
Delta sells a ferrite donut with the leads wrapped around it that is supposed to fix this, but it didn't help with my situation. I plan to (but haven't yet) try increasing the number of wraps to see if it will work. My guess is that the power supply draws some kind of spike when starting up that trips the GFCI. Or maybe there is some kind of capacitive coupling between one side of the hot leads and ground. The ferrite donut should reduce any differential-mode spike between the hot leads (which is what a GFCI is supposed to detect).
-- --Pete "Peter W. Meek"

That GFCI receptacle has a balanced current transformer inside that is VERY sensitive - If it thinks that there's more than a 3 Milli-amp leakage to ground (power going out on the hot wire that is not coming back on the white wire) it trips.
Problem is, when you are welding and the light is close, or shares a common ground through the inverter-fed equipment chassis or the welding table, you get induced currents in the light cord that can fool the GFCI into thinking there's a problem when there really isn't.
If it likes to trip out on you, either cheat and run a cord from a non-GFCI outlet, or put one in dedicated for that purpose.
You could also build a transformer-isolated work light - get a Malibu Light transformer to make some 12 VAC, and then use any landscaping light fixture to make your work light. You can easily go from 10W up to 75W with an MR-16 prefocused reflector lamp, and you can have your choice of beam spreads with a half-dozen steps between Very Narrow Spot to Very Wide Flood.
Oh, and if you have a refrigerator or freezer in the garage, make sure it is on a dedicated appliance circuit that you have installed for the task - NOT on the GFCI-protected garage receptacle circuit like the current codes call for. You do NOT want to come home from a 2-week vacation to find that GFCI tripped and the freezer thawed out, and a biology experiment is going on inside...
-->--

Peter - Do you have the donut next to the VFD or next to the power plug ? The antenna is at the VFD and the donut shortens the effective length.
Might just move it around and see what happens.
Martin
Peter W. Meek wrote:

Good point. I put the donut in a box with a power cord and a receptacle. I plugged the full length of the lathe's power cord into the receptacle (so as to not modify the lathe). I will try moving the donut closer to the VFD.
--Pete "Peter W. Meek" Building a new house with a shop at:

GFI? I will no doubt regret asking, but it is the weekend, and I've had a real sh1t week. My head is below the parapet in anticipation :-) Regards GeoffH Norfolk UK not VA

Ground Fault Interrupter. They function by measuring the current on the hot and neutral wires. If the currents differ by greater than a set amount (typically 3 mA), they trip open the circuit. The assumption being that the current difference is due to a fault (short) to ground.
Gary