GFCI trips - inductive loads, parallel wired, with relays!

[PatM]

I'm kind of stumped by a circuit that's just recently started tripping. It's been in service for five years and hadn't seen a GFCI trip until a few months ago. The gap between first and second trip was about two months but the time between gaps has been reducing exponentially.

Originally the curcuit was a 15 AMP GFCI breaker feeding four recepticals which power four stenner perastaltic pumps (about 1.9 amps each). The power to each receptacle is switched by a small automotive relay (rated for 2A inductive).

I checked continuity from hot to ground and got nothing, neutral to ground got nothing (after disconnecting the GFCI netural at the breaker), hot to netural nothing (with breaker off and everything unplugged of course). Resistance through each motor was the same as a brand new one I have on the shelf.

Electrician replaced the GFCI breaker but that made zero difference.

We then switched to a regular breaker and I installed four GFCI receptacles figuring that if it's a motor then only once receptacle would trip. It didn't even make it a day before all four receptacles were tripped.

Attached is a pic of the wiring (ground not shown due to laziness) the switches shown are actually relays.


Thinking I had a brain fault I checked and indeed all four are wired to LINE, not LOAD. I did notice a low voltage at one receptacle and when I checked the relay it appeared the contact was pretty burnt. I checked the others and they are pretty bad as well. I replaced two and on the other two I just switched to the second set of contacts which hadn't been used.

I'm thinking maybe the burnt contacts were causing the trips as these pumps aren't tripping GFCI in any other location (have six others in operation in other places) and hadn't been tripping here for several years prior to now.

Could that be the cause? Is so, what would actually cause the trips? Low voltage? (brownout) Perhaps arcing or rapid on/off from poor electrical contact? Or am I barking up the wrong electrical pole entirely?

Thanks!

 

[jim dungar]

Or am I barking up the wrong electrical pole entirely?

Have you tried blaming the motors/loads instead of the GFCI/source?

 

[PatM]

Yep, that was the first thought. I have a brand new motor that I swapped with each existing one to see if it would still trip - and it did. We have identical pumps on other GFCI circuits and they never trip.

I guess I should also try swapping across different circuits, see if a motor from the problem circuit will trip the non-problem circuit. However, the winding resistances are identical across all pumps and none have continuity to ground.

I should get the electrician to megger the conductors. Perhaps we have an arcing issue that's invisible during continuity checks. It's 120VAC so maybe try 240VAC on the megger. Didn't see any signs of arcing at any connection points though.

 

[jim dungar]

Rarely, do line side 'problems' cause tripping of current generation GFCI devices.

 

[GoldDigger]

On the other hand we do not normally see installations where the line feed to the GFCI receptacle is switched through arcing contacts while a motor load is connected to the receptacle.
that is not AFAIK a use condition that the receptacles are tested for.
On the other hand that by itself would not explain why the GFCI breaker tripped.
Hypothesis:
1. The motor starting load or turn off transient with relay contact problems, causes enough high frequency transients that the capacitive current between hot and ground is above the GFCI threshold.
2. The "automotive style" relays are not actually rated for this kind of situation. Solid state relays with zero-current off and zero voltage on switching might be better suited. Or snubbers could be fitted across the relay contacts.
Question:
Does the OP have any idea whether the trip is happening when the lumps turn on or when they turn off?

Tapatalk!

 

[PatM]

Not sure if they trip on make or break. I did catch one trip when the GFCI breaker was in place but I was using my meter for max current at the time trying to determine if it was overload or ground fault causing the trips (max showed 5.9 amps with all four pumps running). Couldn't tell if something was turning on or off.

Zero crossing SSRs would be great. I did run into a problem with those once where the installer shared a single signal common at the relays which somehow made them never want to open once they closed. Don't know why they didn't just pull four conductors as each DDC output has a common right beside it. Would something similar happen if using a shared neutral? The branch would be right near the SSRs and about 100 feet from the panel.

I've swapped around two pumps (between a system that doesn't trip and this one) so if the other three trip again and the "swapped" pump doesn't then I've probably got some leakage through the motors to ground.

No trips since I changed the relays (about six hours ago) which is a hopeful sign. If it makes it more than a day with no trips then I'll start thinking maybe the relays were the problem. If it makes it two days then I'll be sure - haven't gone more than a day without tripping in around a week.

 

[GoldDigger]

On the other hand if you replaced the original failed relays with identical spec units they may not last either.

Tapatalk!

 

[pjholguin]

Do these loads need to on a GFCI circuit? If so are they required to be class A GFCI protected? Can you change to a GFPE circuit?

I know this doesn't answer your question, but it may give you time to do a more involved troubleshooting plan.

Best of luck,
PJHolguin

I'm kind of stumped by a circuit that's just recently started tripping. It's been in service for five years and hadn't seen a GFCI trip until a few months ago. The gap between first and second trip was about two months but the time between gaps has been reducing exponentially.

Originally the curcuit was a 15 AMP GFCI breaker feeding four recepticals which power four stenner perastaltic pumps (about 1.9 amps each). The power to each receptacle is switched by a small automotive relay (rated for 2A inductive).

I checked continuity from hot to ground and got nothing, neutral to ground got nothing (after disconnecting the GFCI netural at the breaker), hot to netural nothing (with breaker off and everything unplugged of course). Resistance through each motor was the same as a brand new one I have on the shelf.

Electrician replaced the GFCI breaker but that made zero difference.

We then switched to a regular breaker and I installed four GFCI receptacles figuring that if it's a motor then only once receptacle would trip. It didn't even make it a day before all four receptacles were tripped.

Attached is a pic of the wiring (ground not shown due to laziness) the switches shown are actually relays.


Thinking I had a brain fault I checked and indeed all four are wired to LINE, not LOAD. I did notice a low voltage at one receptacle and when I checked the relay it appeared the contact was pretty burnt. I checked the others and they are pretty bad as well. I replaced two and on the other two I just switched to the second set of contacts which hadn't been used.

I'm thinking maybe the burnt contacts were causing the trips as these pumps aren't tripping GFCI in any other location (have six others in operation in other places) and hadn't been tripping here for several years prior to now.

Could that be the cause? Is so, what would actually cause the trips? Low voltage? (brownout) Perhaps arcing or rapid on/off from poor electrical contact? Or am I barking up the wrong electrical pole entirely?

Thanks!

 

[GUNNING]

On the other hand if you replaced the original failed relays with identical spec units they may not last either.

Tapatalk!

I would suspect the Automotive relays were not rated for 120 volts but for 12 volts and that is where the ground faults are occurring. If they all are tripping anyway I might think the feeds are damages in the wiring someplace. Its not a dead short it is a GF which tells me to look at the insulation values where wires are near a ground or points where current can go to ground. Moisture in the relays?

 

[PatM]

I just got back to work (3 day weekend) and I've not received any complaints about trips. The trend logs don't show anything that resembles a trip... Just waiting for the boss to get in and find out if they had any that I'm just not seeing.

If it is fixed then I know it was the worn out relays arcing so I'll switch to zero crossing SSRs.

To tell the truth, I'm not sure if GFCI is required or not but the pumps feed chemicals into swimming pools so it would seem logical to have GFCI despite the absence of conductive materials between the motor and wet end.

Thanks for the ideas!

 

[PatM]

As I mentioned above, there was something about SSRs that was twitching my memory. I went through some SSR docs and it turns out you can't use a zero crossing with inductive loads because the current lag messes up the zero crossing detector. EIther it won't turn on or it won't turn off.

So then it's back to random on SSRs but the inductive load means snubbing is a must or the SSR can break or explode. Most often recommended is a MOV (metal oxide resistor) but those deteriorate and are sometimes very hard to match to the SSR. Mismatching or leaving in too long can cause things to break or explode as well.

So since it has to be random on and an SSR is such high maintenance with an inductive load, we figure we'll just upgrade the relays. The current ones are rated for 2A inductive at 120VAC but the contacts are toast after only five years. If I go too big on the relays then the lack of arcing can let the contacts glaze over. I guess I'll look for ones rated around 5A inductive at 120VAC. Might be fun finding those with a 12VDC coil.