GFCI Leakage Current

[bwat]

Had a strange thing happen the other day. Troubleshooting a 2-pole (240V, class A) GFCI breaker issue and found that when all wires are disconnected from breaker it was measuring roughly 26K-ohm from L1 load terminal to the white neutral pigtail, and about 60K-ohm from L2 load terminal to same pigtail. After measuring this, I thought “that can’t be right” so I had the breaker replaced.

A measurable resistance such as this would mean that there’s automatically some amount of leakage current that would go towards tripping the breaker’s GFCI coil. Granted, they would cancel each other out a bit, but this would mean there’s already several mA of current going toward tripping the GFCI circuit without even a load on it. Must be some type of partial short in the breaker I though. Maybe the test button is partially stuck...

So the new breaker arrives... I open it up and do the same test. Get out of here. Same exact readings. So is this normal? Unless I’m being foolish with something here, this means 4.6mA leakage from L1 and 2 mA leakage from L2... on the breaker itself. Surprisingly high. Or at least not what I personally expected.

 

[Dennis Alwon]

The neutral is connected to electronics inside the breaker but I am not sure why you would get a reading to the load with the breaker off.

 

[synchro]

Had a strange thing happen the other day. Troubleshooting a 2-pole (240V, class A) GFCI breaker issue and found that when all wires are disconnected from breaker it was measuring roughly 26K-ohm from L1 load terminal to the white neutral pigtail, and about 60K-ohm from L2 load terminal to same pigtail.

Was this measurement made when the breaker was removed and switched ON? If so, then the resistances you measured could actually be physically located between the line sides of L1, L2 and the neutral pigtail (and not on the load side of the toroid CT).

If the resistance was present with the breaker OFF then something more complex might be happening with internal sensing circuitry, and perhaps the leakage goes away once some initialization has been completed? Also the measured resistance might not be caused by an actual resistor but from leakage through semiconductor devices that goes away after the breaker is powered up. This is just some speculation on my part.

 

[bwat]

To add detail, the breaker was definitely in the OFF position when these measurements were taken. 100% certain of that. I believe the breaker being ON may have shown the same values when I tried that at a later time, but I’m not absolutely certain of that one.

 

[Jerramundi]

Could be a bad batch. I had a similar problem once and just kept returning them until I got one with zero reading. Maybe an entire pallet got banged around on a truck?

From the basic component diagram provided by Siemens, there should be zero continuity between either L1/L2 and N... at least as far as (1) the coil and (2) the relay are concerned. Any current induced on the coil should then flow to the relay and, I assume, out via the neutral, activating it and tripping the breaker... but the current to the relay is provided via induction on the coil and there is NO direct connection to L1 and L2 as far as I'm aware.

Then again Dennis is right about there being more internal circuity than that basic diagram lets on.

The only way to know for certain would be pull up an exact diagram of ALL of the internal components of your particular model breaker and go from there. Most spec sheets just show a coil and a relay.

I will say that that 4.6mA you got is suspiciously close to the tolerance level... drawing my eye to what I think is a solenoid?

 

[GoldDigger]

It seems possible that the design of the breaker actually does power the electronics from the line side of the breaker. Otherwise there would be a self-test delay in getting power to the load each time the breaker is turned on. Even with this, there should be no leakage from line side to load side, but with circuitry connected to the neutral on both sides of the breaker contacts it could still cause an apparent impedance between Line L1 and Load L1.
The assymmetry between L1 and L2 is likely because the elecronics is likely to be powered primarily or entirely between L1 and neutral.

 

[Jerramundi]

Well, whose gonna be the hero of the story and waste $50.00 on dissecting a GFCI breaker in order to discern every component on the PCB? lol

 

[ptonsparky]

Well, whose gonna be the hero of the story and waste $50.00 on dissecting a GFCI breaker in order to discern every component on the PCB? lol

I suspect we have at least one individual that has done that.

 

[Jerramundi]

Just out of curiosity and decided to see if I could find more information than what the general wiring diagram on most spec sheets features (i.e. the coil and the relay)... and it seems really difficult. Probably because it's proprietary information.

Here is the standard diagram featured on most spec sheets for GFCI breakers (this one is Siemens 1-pole and 2-pole):



And here is an internal shot of looks to be a CH Type GFCI Breaker, 1-Pole

 

[Jerramundi]

I suspect we have at least one individual that has done that.

I'm seriously thinking about it for a winter project, lol. I understand how the coil and relay function. But I would be curious to see what exactly is on the PCB. Obviously if one were to discern what is exactly on the PCB it would have to be shared discreetly because spreading that information around is most likely somehow illegal, lol.

I would like to think as electricians we would be purview to that information, but I doubt Siemens or any other company is going to give that level specificity without some sort of NDA or something.

 

[synchro]

It seems possible that the design of the breaker actually does power the electronics from the line side of the breaker. Otherwise there would be a self-test delay in getting power to the load each time the breaker is turned on. Even with this, there should be no leakage from line side to load side ...

That sounds like a good argument for the electronics being powered from the line side of the breaker switch. But there still could be other circuitry connected after the switch but before the toroid transformer (for example to determine if the switch is ON/OFF or perhaps other more sophisticated functions). If that was the case then you might see the impedance from those connections between the line outputs and the neutral pigtail, but it would not effect ground fault sensing because the circuits are connected before the toroid.

 

[Jerramundi]

...before the toroid transformer...

Is that the official terminology for the coil that gets energized via induction from the difference in magnetic flux? I understand that's the geometric object the coil is wrapped around.

I'm honestly just curious because I just refer to it as a coil. Or is there another component in the GFCI breaker called such?

 

[synchro]

Is that the official terminology for the coil that gets energized via induction from the difference in magnetic flux? I understand that's the geometric object the coil is wrapped around.

I'm honestly just curious because I just refer to it as a coil. Or is there another component in the GFCI breaker called such?

I've seen it called a coil or sense coil, but it really is a transformer. And more properly a current transformer or CT.
You are correct that "toroid" or "toroidal" describes the physical geometry of the transformer, and therefore such a transformer is often described that way.
There is another component if a GFCI called a coil, and that is the relay coil or trip coil which trips the breaker if a ground fault is detected.

 

[Jerramundi]

I've seen it called a coil or sense coil, but it really is a transformer. And more properly a current transformer or CT.
You are correct that "toroid" or "toroidal" describes the physical geometry of the transformer, and therefore such a transformer is often described that way.
There is another component if a GFCI called a coil, and that is the relay coil or trip coil which trips the breaker if a ground fault is detected.

Roger, Roger. "Coil" and "Sensing Coil" is what I'm used to. I was just wondering if you were referring to the same component when you called it a "toroid transformer" because, if you see Post #12, there is a blue, cylindrical object which I'm struggling to identify and wasn't sure if that's what you were referring to.

I suppose use the term "Coil" is a bit general and could misattributed to various components and perhaps your terminology would help to better distinguish what I'm used to calling the "Coil" or "Sensing Coil" from other components.

I'm aware of (1) the relay as a tripping mechanism, (2) the "Coil" or "Sensing Coil," (3) what I believe is the solenoid as part of OCP, and (4) how they all function... but anything beyond that in regards to the innards of a GFCI breaker is still new to me.

 

[xformer]

is this a True GFCI? or are we talking a combination ARC Fault/GFCI??

 

[bwat]

Thanks for the great answers everyone. Sounds like this is an interesting topic and not as straight forward as I always envisioned.

is this a True GFCI? or are we talking a combination ARC Fault/GFCI??

Just GFCI as far as I know. Original was a Murray MP-GT, purchased maybe 15 years ago. New one is Siemens QPF equivalent.

 

[rnatalie]

Nope, this is to be expected. The GFCI plucks a small amount of current off the line side of the breaker to power itself (including the Neutral-Ground signal injector). Unless tripped, the Load side wires (hot and neutral) are connected to the line side. The power for the GFCI circuitry doesn't trip things because it comes off before the toroid transformer that measures the imbalance.

 

[ELA]

Did you try reversing the (+) and (-) of your meter? Same or different reading?
The tear down I did years ago of a GFCI showed the "Green Status LED" connected to the load side. From load (hot): -> LED->diode -> 26K ohms Resistor : to Neu

 

[Jerramundi]

Nope, this is to be expected. The GFCI plucks a small amount of current off the line side of the breaker to power itself (including the Neutral-Ground signal injector). Unless tripped, the Load side wires (hot and neutral) are connected to the line side. The power for the GFCI circuitry doesn't trip things because it comes off before the toroid transformer that measures the imbalance.

Humor me, because I'm trying to understand this too....

If power is "plucked" from Line-Side L1/L2 for internal components.. and dissipates said power through Line-Side N (i.e. the CB neutral pigtail), that still doesn't explain any type of continuity and therefore resistance between Load-Side L1/L2 and Line-Side N (i.e. the CB neutral pigtail), with the breaker in the OFF position... (see below)

However, if the breaker is in the ON position, I can understand the potential for continuity and resistance between those points IF traveling through some other sort of internal components... (see below)

 

[Jerramundi]

Based on my limited understanding, the only you're getting a reading between Line-Side N (i.e. the CB neutral pigtail) and Load-Side L1/L2, with the breaker handle in the OFF position, is if the breaker is taking power for some other internal components from the LOAD SIDE of L1/L2 as opposed to the line side.