GFCI Testing

[Alarm Slave]

aside from the customary "test" button that is provided on GFCI breakers and GFCI outlets does anybody know of a real world method that won't physically get you into zapped?
Seems to me that the "test" button is merely a switch a like life safety devices the "test" button being pressed only lets you know that the "test" button makes the circuit. Any ideas?

 

[iwire]

The test button does not just trip the GFCI it really does test it and it is the only UL backed way to field test a GFCI.

All the low dollar gizmos that claim to be testers are just 'indicators' and none of them can trip a GFCI that is used as a replacement on an ungrounded circuit as specifically allowed by 406.3(D)(3)

 

[Alarm Slave]

GFCI Testing

Does that mean the little tester your friendly inspector uses to test your wiring at the receptacles is not UL approved?

 

[480sparky]

Does that mean the little tester your friendly inspector uses to test your wiring at the receptacles is not UL approved?

Correctamundo!

 

[iwire]

Does that mean the little tester your friendly inspector uses to test your wiring at the receptacles is not UL approved?

UL does not approve anything.

Those things, to the best of my recollection are UL listed as indicators.

They can certainly check if a normal looking duplex is in fact connected to an operational GFCI. But once your at the GFCI itself the test button should be used.

The test button but actually creates a calibrated current imbalance to test the electronics and relay of the GFCI.

 

[480sparky]

Part of UL 1436 reads:

1.1 These requirements apply to outlet circuit testers, including screwdriver and pen-style voltage presence indicators, for use on 15-, 20-, and 30-A, 3-wire, 125-, 250-, 277-, 480-, or 600-V receptacles, ground-fault circuit-interrupter testers and arc-fault circuit-interrupter indicators for use on 15- and 20-A, 3-wire, 125-V receptacles, and similar indicating devices that are:

a) Intended to be connected to the receptacle for a period of time only as long as is necessary to note the indicated pattern of lights or other similar indicating means, and
b) Not intended to be a comprehensive instrument or to determine the quality of the grounding circuit.
​

 

[gar]

090712-1329 EST

The Leviton 7899 has a very interesting test function.

Pressing the TEST button mechanically unlatches the mechanical latch. This is all that is tested at this point.

When the RESET button is pressed the electronics and functionality of the trip coil are tested. If this fails the unit will not latch.

To me this seems better than doing the test of the electronics when the TEST button is pressed. In the Leviton design you are testing the electronics any time the device is RESET at the time of RESET and not just when the TEST button is pressed.

Internally some, if not all, GFCIs use a 15,000 ohm resistor from hot on the output side of the current transformer to neutral on the input side of the current transformer to create an unbalanced test current of 8 MA at 120 V.

You could do your own external test with a 15000 ohm resistor connected from a neutral or EGC that precedes the device under test to a hot slot in the device under test.

.

 

[LarryFine]

You could do your own external test with a 15000 ohm resistor connected from a neutral or EGC that precedes the device under test to a hot slot in the device under test.

I just use my wiggy (K-60) as an external tester.

 

[iwire]

I just use my wiggy (K-60) as an external tester.

I have done that many times myself but I would call that an indication not a test.:smile:

I think it is safe to assume the running load of wiggy at 120 volt is far above 8 ma and the inrush well above that. :smile:

 

[don_resqcapt19]

Gar,
Is that the newer design GFCI with the "lockout function"? It sounds like it is, and the description would prevent the GFCI from providing power to the outlet if the electronics have failed, providing a safer installation. In the older design GFCIs, you could reset them even when the electronics could no longer provide the required protection.
As a side question, did your testing indicate that the device would kill the output power if the electronics have failed. My understanding is that they do not and the only provide the "lockout protection" when you try to reset them after a test or trip.

 

[gar]

090712-1718 EST

don:

This is clearly a newer design and does have the lockout function.

Certain types of electronic failure in this device will cause a trip.

I will suggest that there is no way to design any product such that every possible failure mode will cause a failure to a safe condition.

In fact you may not be able to define a safe condition. Suppose you have a machine in which a person is caught. Should an emergency stop button cut all power, or is it better to leave the machine power on and automatically shift to manual mode. My suggestion --- do not get caught in the machine. That is the safest approach.

On some machines with light curtains, where I had some input on the control design, and where the machine cycle was such that manual entry during the cycle was part of the cycle it was possible to check the light curtain at least once per machine cycle. This was done by knowing that at certain points in the cycle that manual entry was required and thus a change of state should occur. If this change did not occur the cycle was inhibited. Normally light curtains only do a self test at power on. This was some years ago. It may be that some new designs have created ways to do self tests at times other than power up.

.

 

[cpal]

Back in the 70's we had a meter that dialed in the mA to Gd. to prove the GFCI breakers (prior to rec) good. Are they still around???

 

[LarryFine]

I think it is safe to assume the running load of wiggy at 120 volt is far above 8 ma and the inrush well above that. :smile:

I believe the K-60 has a 4K impedance, which would have about 4x the current of the resistor.

What about the typical 3-prong plug-in receptacle indicator/GFCI tester?

 

[hurk27]

I have done that many times myself but I would call that an indication not a test.:smile:

I think it is safe to assume the running load of wiggy at 120 volt is far above 8 ma and the inrush well above that. :smile:

I thought we have already determined that any external tester only served this very function? if so why by a GFCI tester when a wiggy will serve the same purpose, I mean its not really testing the function of the GFCI but just indicating that a said receptacle is in fact protected by a GFCI device:grin:

 

[hurk27]

Back in the 70's we had a meter that dialed in the mA to Gd. to prove the GFCI breakers (prior to rec) good. Are they still around???

GreenLee made one, I had it till it stopped working about a year ago. I think the settings was 3, 5, 7 and 30ma <<for class B devices

 

[electricalperson]

GreenLee made one, I had it till it stopped working about a year ago. I think the settings was 3, 5, 7 and 30ma <<for class B devices

i have the slater gfci tester. is that the one your talking about? it dials in the trip current

 

[hurk27]

i have the slater gfci tester. is that the one your talking about? it dials in the trip current

It was a greenlee, yellow, and black case with a short pigtail to plug into an outlet.
It would send a pulse every 5 sec. at the rated set setting so you could reset the breaker and make sure it trips twice, why? I have no idea they design it that way.

 

[ELA]

090712-1718 EST

On some machines with light curtains, where I had some input on the control design, and where the machine cycle was such that manual entry during the cycle was part of the cycle it was possible to check the light curtain at least once per machine cycle. This was done by knowing that at certain points in the cycle that manual entry was required and thus a change of state should occur. If this change did not occur the cycle was inhibited. Normally light curtains only do a self test at power on. This was some years ago. It may be that some new designs have created ways to do self tests at times other than power up.

.

Light curtains have progressed Gar,
http://www.triadcontrols.com/superlight_vi_description.html

 

[gar]

090713-0918 EST

ELA:

Thanks for the reference.

Presented is a good design philosophy and it is described in a very concise fashion.

.

 

[hornetd]

I tested my Wiggy.

I tested my Wiggy.

I have done that many times myself but I would call that an indication not a test.:smile:

I think it is safe to assume the running load of wiggy at 120 volt is far above 8 ma and the inrush well above that. :smile:

Well, just for drill I ran my Wiggy through a current multiplier loop and threw a peak holding Fluke meter on it. Mine actually drew nine milliamps and the meter did not detect any inrush.