ECN Forum
Posted By: sparkyinak Inverter tripping GFCI down line from it - 05/25/14 07:35 AM
I have an interesting problem. I picked up a used pure sine wave inverter, bench tested it and seem to work just fine. I installed it and power it up and it's tripping an older GFCI that is on the boat about 5 seconds after throwing the power. Even when there's nothing plugged into the GFCI. I used my meter to test the inverter output. Both the voltage and and frequency is spot on. The inverter turns out was not bonded internally but that shouldn't matter. I bonded it at the panel temporary and the GFCI won't reset. I transferred back to shore power and took out the Shore power GFCI. Now I ran out of daylight.

I'm gonna pull the boat's GFCI tomorrow and remove my temp bond at the panel. It appears to be the inverter because the GFCI circuit in question works fine on shore power. How can a power source trip a GFCI, down line? My only thought the GFCI is bad but nothing happens to it on shore power. The only common denominator is the used inverter
Posted By: Tesla Re: Inverter tripping GFCI down line from it - 05/25/14 08:08 AM
Inverters are 'spikey.'

The spikes are enough to upset the electronic balance circuit in your GFCI.

If you were to run the juice through a (ferro-resonant) transformer the problem figures to drop away.

(The magnetic circuit would clip the spikes.)
I agree high frequency spikes are the most likely cause.
Could be worth trying ferrite rings on the output leads from the inverter.
Thanx!

How do I determine the size or capacity? Any particular ones recommended or stay away from?

The inverter is supposedly a true sine wave however I do not have access to an oscilloscope. Can it still be spiky and put out a true sine?

I'm looking at several online. They have impedance ratings. Will they affect power down line or will just affect the spiking?
Posted By: Tesla Re: Inverter tripping GFCI down line from it - 05/25/14 09:03 PM
"True Sine" comes from the sales department.

Solid state electronics can NEVER produce a true sine wave. It's solid state!

The sales liturature is alluding to some attempt at replicating a rock-steady 60Hz output with nominal sine characteristics, nothing more.

Since there are no absolute standards as to what constitutes a true enough sine wave, the marketing departments throw the term around pretty loosely.

Your best bet is to salvage ferrite rings from a power supply. Such devices are parked out on the curb every day of the week: abandoned TV sets -- particularly the prior generation of high power, wide screen dimensions.

You'll find that not only are these give-aways... but that each one is stuffed with ferrite rings in the bowels of the power supply.

Just be sure to bleed off the capacitors, first.

( The ferrite rings are used precisely because they clip off high voltage spikes, BTW.)

Impedance in a ferrite ring is an explicit reference to its INDUCTANCE. Capacitance and resistance are zero for such devices.

Adding inductance affects reactive power demand -- and thus 'chokes' the net power delivered. Unless you've ganged up quite a few, this effect is going to be trivial and inconsequential. You're only clipping off the high frequency spikes.

I may have some pretty big split ferrites around here somewhere if you can't find any (cubes an inch or more on the side) You can loop the wire through them a couple times to increase the damping.
Thanks a lot guys
© ECN Electrical Forums