Fluorscent grow light causing GFCI recptacle to not trip.

Status
Not open for further replies.
O

Ohms law

Senior Member
Location
Sioux Falls,SD
Any one have an idea why fluorescent fixture with grow lamps would not allow a GFCI receptacle to trip when pushing the test button. The light fixture is the only load on the load side of the GFCI. The second you unplug the light and test the GCFI it will trip.
 
If the light generates enough RF noise conducted down the cord, it might interfere with the proper operation of the GFCI including the test button.
In other words rather than the noise causing a false trip it might prevent the circuitry from recognizing the actual current imbalance generated by the test button. I would consider this a design problem with the GFCI as well as a potential problem with the light itself.
I cannot come up with any way that a wiring fault in the light (such as a neutral to ground connection), could interfere with the operation of the test button.
I take that back: if the light somehow introduces an out of phase leakage current which is at, say, half the trip level, then it could cancel enough of the deliberate leakage to prevent the test current from reaching the trip level.
What happens if you lift the ground in the cord and then test?

Tapatalk!
 
GoldDigger said:
If the light generates enough RF noise conducted down the cord, it might interfere with the proper operation of the GFCI including the test button.
In other words rather than the noise causing a false trip it might prevent the circuitry from recognizing the actual current imbalance generated by the test button. I would consider this a design problem with the GFCI as well as a potential problem with the light itself.
I cannot come up with any way that a wiring fault in the light (such as a neutral to ground connection), could interfere with the operation of the test button.
I take that back: if the light somehow introduces an out of phase leakage current which is at, say, half the trip level, then it could cancel enough of the deliberate leakage to prevent the test current from reaching the trip level.
What happens if you lift the ground in the cord and then test?

Tapatalk!
Click to expand...
My co-worker is the one that has asked me this question, I don't believe he went that detailed into it. He had a final on a basement and the receptacle was in a unfinished location. So by default it has to be GFCI protected. The homeowner picked this light up at Menards for her plants. I was just curious if anyone had a reason why this was happening and what I could do to fix it other than have the lady get rid of her light.

The inspector won't pass until the GFCI will trip with the light plugged in. We tried a different GFCI and the same thing happened. The brand of the GFCI is Pass & Seymor.

What would it prove by removing the ground wire?
 
It would eliminate the effect of capacitive current from the tube to the grounded sheet metal of the lamp.
If you (he) could put in a different brand GFCI it might help testing too.


Tapatalk!
 
I don't know why the GFCI won't trip but the inspector is wrong. There is only one test for a GFCI and that is to press the test button; pressing the test button with something plugged in is not required.

If this gets to be a big problem take the light and put in the trunk of the HO's car and call the inspector back. If it is still a problem call the inspector's boss.
 
I had an instructor back in '72 that would tell us that the NEC was to make things safe, not necessarily that they would work.

While using the test button with no load may allow you to pass inspection, it sure doesn't work right.

Try a different brand GFCI. Then try a different fixture.
 
ActionDave said:
I don't know why the GFCI won't trip but the inspector is wrong. There is only one test for a GFCI and that is to press the test button; pressing the test button with something plugged in is not required.

If this gets to be a big problem take the light and put in the trunk of the HO's car and call the inspector back. If it is still a problem call the inspector's boss.
Click to expand...

The inspector insisted the GFCI worked with the light because he believes that when he leaves the lady would just plug it back in. He said he doesn't want to pass the inspection knowing that someone could die because a light fixture causing this issue.

It sounds like are only options is to replace with different brand or take the receptacle out completely. I understand what the inspector is saying but at the same time we can't control what a homeowner plugs in or if they run a extension cord and leave it for permanent use. You just don't know this.
 
GoldDigger said:
It would eliminate the effect of capacitive current from the tube to the grounded sheet metal of the lamp.
If you (he) could put in a different brand GFCI it might help testing too.

Would it help to have some type of resistive load on the load side of the GFCI with the florescent fixture? Just curious.
Click to expand...
 
I just did some snooping around and found that the test button on some GFCI receptacles will not work if the voltage is too low. P&S specs are for 120v + 10, - 15 percent. So a voltage of 102 is at the bottom of the specs.

Other, cheaper brands may have a higher minimum voltage range.

So, I would first check the voltage at the receptacle under a load. And I would do so with different loads. If the test button won't work with say, a portable heater plugged in it, that would tell me it's the amount of load affecting the receptacle, not the type of load.

If there is a substantial voltage drop, or the voltage is just too low to start with, there is bigger fish to fry than trying to make the receptacle pass inspection.
 
How long did you hold the test button in? As I recall the resistor that the test button is connected to is designed to flow 8mA at 120 volts. The UL standard permits a 3.7 second time to trip for an 8mA fault. If the voltage is low the test current would also be lower. The standard would permit a 5.59 second time to trip at 6mA. The standard does not require the GFCI to trip if the fault current is less than 6mA.

It is required not to trip with a fault current of less than 4mA, permitted, but not required to trip for fault currents of 4 to 6mA, and must trip for fault currents above 6mA.

Yes, I am aware that most GFCIs trip instantly when the button is pushed, but the standard does not require them to do so.

Per the UL standard, the maximum permitted time to trip in seconds is equal to the quantity (20/fault current in milliamps) raised to the 1.43 power.
 
Ohms law said:
That is very possible. I hope it wouldn't be something so simple. That would be funny:lol:
Click to expand...

I was looking at some GFCIs and the reset button is REAL close to the hole for the grounding conductor. A round cord plug could easily cover it.


2423589_600.jpg
brown_gfi_15a_outlet_receptacles_1595__83563.1350940393.1280.1280.jpg
NEMA_5-20RA_GFCI_Tamper_Resistant_Receptacle.jpg
 
don_resqcapt19 said:
How long did you hold the test button in? As I recall the resistor that the test button is connected to is designed to flow 8mA at 120 volts. The UL standard permits a 3.7 second time to trip for an 8mA fault. If the voltage is low the test current would also be lower. The standard would permit a 5.59 second time to trip at 6mA. The standard does not require the GFCI to trip if the fault current is less than 6mA.

It is required not to trip with a fault current of less than 4mA, permitted, but not required to trip for fault currents of 4 to 6mA, and must trip for fault currents above 6mA.

Yes, I am aware that most GFCIs trip instantly when the button is pushed, but the standard does not require them to do so.

Per the UL standard, the maximum permitted time to trip in seconds is equal to the quantity (20/fault current in milliamps) raised to the 1.43 power.
Click to expand...

Would you have a link to the UL standard?

Unfortunately I was not at the home were my co-worker was at. He is a pretty smart guy I wouldn't think he would not over see something so simple as a plug covering the reset button, although it is very possible.
 
Ohms law said:
Would you have a link to the UL standard?

Unfortunately I was not at the home were my co-worker was at. He is a pretty smart guy I wouldn't think he would not over see something so simple as a plug covering the reset button, although it is very possible.
Click to expand...
There are no legal copies of UL standards on the net. There are some documents that you can find that quote the standard. The standard number is 943.
You can read some details here.
 
K8MHZ said:
Another possibility...

Could the plug from the light be holding the reset button in so it's can't pop out when the test button is pressed?
Click to expand...
I have never had one fail to trip because something was preventing the reset button from popping out. I'd almost be willing to bet they would not pass UL either if they did that.

K8MHZ said:
I just tested the GFCI receptacle in my bathroom. If I hold the reset button in about 1/16" of an inch, the test button won't trip the receptacle. The buttons on my receptacles are flush with the face.
Click to expand...
Just checked the one in my bathroom - same thing, but if the reset button is pushed in (may even have been less than 1/16") there is no voltage at the receptacle either. It is a P&S, not the most recent 1595 model but the 1594 that preceded the ones available now. I think the later model was a new UL requirement that they not reset when the GFCI circuitry no longer is functioning, and/or maybe will not allow it to reset if line and load terminals are reverse wired.
 
Status
Not open for further replies.
Top