ECN Forum
Posted By: Rich Thomas Surge suppressor trips GFCI - 03/19/05 11:59 PM
After installing GFCI on a branch circuit I discovered that the transient voltage surge suppressor (TVSS) I used for the television trips the GFCI. At first I thought the TV, VCR, DVD, or clock had developed a ground fault, but after unplugging all the loads, the GFCI still tripped until I removed the TVSS.

The surge suppressor is a 120V/60Hz 10A DataShield Model 85. It is the type that has an integral relay, so the relay disconnects the unit from the line until reset. Operating the TVSS's reset button momentarily energizes the relay, then the GFCI trips deenergizing the relay again.

When unplugged, continuity measurements show an open circuit for all combinations measured at the plug. Because of the relay, continuity measurements from outside the unit to the electronics is impossible because the deenergized relay keeps the electronics open circuited.

I'm guessing the unit is the type that shunts surges from HOT and/or NEUT to GND, rather than shunting surge from HOT to NEUT.

When on an branch circuit without GFCI, the relay can be energized and the unit does carry loads normally, although I have no way of knowing if it offers surge suppression anymore.

It appears that the TVSS itself has a ground fault. I tried to open it up, but would have to almost destroy the unit to access the electronics inside. The only thing exposed once the bottom cover is removed are a 10A fuse and the bottom of a tightly secured circuit board.

I'll be throwing this old, $75 unit away, but I'm curious about what is causing the ground fault. My guess is the neutral maybe grounded inside the unit.

Thanks for reading all this.

What do you think could be the problem???
Posted By: NJwirenut Re: Surge suppressor trips GFCI - 03/20/05 12:22 AM
Could be a leaky MOV or other component connected between hot and ground. A lot of these units incorporate RF noise filtering components, typically a capacitor from both sides of the line to ground. In some cases, the AC current flow through the filtering network to ground is enough to trip a GFCI.

[This message has been edited by NJwirenut (edited 03-22-2005).]
Posted By: dereckbc Re: Surge suppressor trips GFCI - 03/22/05 02:43 PM
I agree, the unit has components installed L-G, and/or N-G.
Posted By: hbiss Re: Surge suppressor trips GFCI - 03/22/05 05:43 PM
This is probably a good example of why not to install GFCI receptacles where they are not required.

-Hal
Posted By: sparky66wv Re: Surge suppressor trips GFCI - 03/22/05 08:56 PM
I have a question in a similar vein:

How many lighted switches on a circuit would it take to open a class A gfci (10ma)?

Just curious.

-Virgil
Posted By: NJwirenut Re: Surge suppressor trips GFCI - 03/22/05 09:17 PM
Do lighted switches use ground as a return path?

Every one I have installed (admittedly not all that many) either had a neutral connection for one side of the lamp (types that are lighted when switch is ON), or had a small neon lamp that relied on a path to neutral through the load being switched (lighted when switch is OFF).

[This message has been edited by NJwirenut (edited 03-22-2005).]
Posted By: sparky66wv Re: Surge suppressor trips GFCI - 03/22/05 09:50 PM
NJ,

Yup, the few lighted switches I've used have no neutral connection and look just like a regular switch. I assume a connection line to ground or perhaps something more exotic like a line-to-line-in-series-with-itself-which-sounds-impossible-to-me type thing. (Uses the disconnected hot leg through a resisted cold filament for it's return path? i.e. back feeding the neutral when "off" through the lamp itself?)

Dunno... I assumed fractions of milliamps of current for the LED; being able to run thousands, well tens, maybe hundreds... Assume 13 switches controlling individual luminaires at 180VA each on a 20A circuit. Not enough current through 13 LEDs to matter.

My Mom's garage has such a switch on a GFCI circuit... BTW.

-Virgil

Pardon the "mad editting"... Dern perfectionists... (talking about myself here... [Linked Image] )



[This message has been edited by sparky66wv (edited 03-22-2005).]
Posted By: NJwirenut Re: Surge suppressor trips GFCI - 03/23/05 12:04 AM
Sounds like a standard type lighted toggle to me. Inside, you will find a neon lamp (with a series current limiting resistor) wired ACROSS the switch contacts.

When the switch is off, the lamp is across the contacts, and the current goes from the hot side of the switch, through the neon lamp, and out on the other side of the switch, through the cold filament of the switched load to neutral. The lamp only allows a couple mA to flow, so the load sees practically no voltage across it.

When the switch is turned on, the lamp in the handle is shorted by the contacts, and goes out, while full current is allowed to flow to the downstream load.

If the lamp being switched burns out or is removed, there is no current path, and the lighted switch stops working. No path to ground is needed. You can prove this to yourself by letting the switch hang out of the box with no ground wire connected to the yoke. It will still function.

[This message has been edited by NJwirenut (edited 03-22-2005).]
Posted By: CTwireman Re: Surge suppressor trips GFCI - 03/23/05 12:08 AM
If the switch is controlling an incandescent light bulb, unscrew the bulb and watch the switch. It would fascinate me to do this as a kid.

Peter
Posted By: hurk27 Re: Surge suppressor trips GFCI - 03/24/05 05:53 AM
Of course that only works if the switch is off. [Linked Image]
© ECN Electrical Forums