Having issue with GFCI tripping. Looking to see if Victron will comment or if there is a solution that doesn’t involve bypassing the ground/safety. I do understand that GFCI are set to 4-6ma in north America. There are past threads about this issue but have no real solution and I haven’t seen any comment from Victron on the issue.
This system is installed in a camper trailer with Multiplus II 12/3000/120 120v, 3x 330ah Smart Lithium, 440w Solar threw Smart solar MPPT 100/30. The rotary switch is used to supply the main trailer panel from shore 1(original) OR shore 2(new/Inverter).
Problem is that when the trailer is connected through MultiPlus to shore power which is provided from a 15amp GFCI on a house the GFCI trips. The receptacle trips and not the overcurrent Breaker in the house panel.
The GFCI doesn’t trip when using Shore 1. Trailer and all appliances work normally.
GFCI doesn’t trip until Multiplus is connected to the main trailer panel. Multiplus does work for charging the batteries without passing power on to the main panel, but this leaves the trailer without 120v while charging.
We have isolated every part of the system. Removed rotary switch and installed new line to main panel. GFCI trips immediately when MultiPlus connects to shore power and connected to trailer panel. If the line to trailer panel is connected to multiplus and removed from trailer panel GFCI holds.
Replaced inverter with new one and same results as above. This would indicate both multiplus are operating the same. Both correctly or both malfunctioning (very unlikely but not inconceivable)
Placed a short line and receptacle on the inverter output and was able to run 1500 watt heater without tripping GFCI
There is no connection between ground and neutral in the trailer. There are no shared neutral points between shore 2 and panel (ac in & ac out are isolated)
I had another electrician with 40+ years come out and confirm/check things with me and he also doesn’t know what is going on. There is talk on other threads that it might be EMI filter or capacitive leakage current, which my college also mentioned. Hopping for some sort of clarification on this issue, even if there isn’t a solution yet.
I did disable and still trips GFCI (2 second delay). The “non” connection I was referring to was to eliminate that from the potential cause of tripping GFCI. I have come across this both in residential and mobile installations.
Happy to answer any questions if you would like to discuss further. This is a reasonably common problem in the marine world and the solution is simple enough in concept. A plausible explanation is available for GFCI tripping in many cases.
In many cases intermittent tripping occurs with no fault. It is just plain old fashioned engineering that can be resolved if the problem is understood and once understood it will be crystal clear.
I did see your previous comments and found it very helpful. The more I dig into this issue the more I am finding that the culprit is a combination of the extremely low trip threshold of 7ma(recorded test) and the way the inverter transfers/switches between pass threw and inverting. I’m definitely curious about the b-stop device but being a Italian make i don’t know if it would work in north America (120/240v 60Hz vs 230v 50Hz) or would even be certified to be used here. The only other issue is cost 200euro = 300$ Canadian not including shipping and customs resulting in 400-600$+ device. As of right now client can charge lithium and run trailer, all though it does require switching where he plugs in the shore power cord to the trailer (inconvenience but not end of the world). Going to keep an eye on this issue and will update if any new info or solution comes in.
Unfortunately i don’t have a meter sensitive enough to measure 1mA increments. Did borrow a colleges ideal supply SureTest Circuit Analyzer, and found that GFCI will Trip at 7mA no mater where i test (direct on plug, in trailer, output of inverter). If the SureTest would trip lower at any point i would belive there is another leakage of current some where. The problem of GFCI tripping when multiplus is transferring is still not fully diagnosed. My current conclusion is that the combination of extremely low trip threshold of 7mA and the way the Multiplus transfers is causing the issue. No solution yet but still troubleshooting.
@LAD_Systems …Ok, firstly, a trip threshold of 7mA is very low for circuits that are aggregated together. It may be ok for one device but for a distribution point it is way too low. In some cases we have measured leakage of 4.5mA for one device.
You can also see the calculations for the typical leakage current for a single IEC socket with an EMI filter included. Ok, its not much current for one but if you have many of them it all adds up.
Ok, the multiplus switching can be explained as we have experienced exactly the same thing. When the multiplus is inverting, the neutral ground link is established (if that option is checked during commissioning). This is so any RCD downstream of the inverter will work even when the multiplus is disconnected from shore power and that ground neutral link snuffs out any downstream leakage. In other words any leakage on the output of the multiplus is never seen on the input when the multiplus is in inverter mode. When the multiplus is in passthrough the ground neutral link on the output of the multiplus is open, allowing the small leakage downstream of the multiplus to be seen by the RCD supplying the multiplus. This adds to the accumulation and trips the RCD.
The Italian company who makes the device is working on the lower voltage version for 120V countries. It already handles 60 or 50 hz. You are right about certification and that may be a problem. It is so simple however, and it is important to note that no power is supplied through the B-Stop and it merely sits across the active, neutral and ground. In fact you could simply plug it into the outlet to solve this problem. That’s what makes it look like snake oil, and that is why I wrote the paper describing it. It seems too good to be true. My only regret is I didnt invent it.
The frustration of nuisance tripping in the marine industry is rife with people going to all sorts of (sometimes dangerous) methods to prevent it. This device resolve the problem and very importantly, does not in ANY way reduce the effectiveness or sensitivity of the original RCD installed. That is very important as RCDs save lives.
Please read the paper and let me know if it doesn’t make sense. Always happy to hear any views on the contents. Nothing better than peer review.
It has been 12 days since I installed the B-Stop and a full week since the cement plant should have returned to normal operations. As stated in this thread, I was only encountering trips on week days at very specific times.
I’ve measured the mA leakage from my 3 inverters and they added up to very close to the tripping threshold. See this discussion for easy reference: RCD tripping on shore power
I had one trip 2 weeks ago. The ELCI tripped at 5:14 am which was one of the regular trip times. We installed the jumper and increased the mA correction from 5mA to 10mA.
So far, no more RCB trips in the past two weeks. (touch wood!)
Is there a solution to this that doesn’t involve purchasing more hardware?
I have a brand new Multiplus II 120vx2 installed and it is being powered by a 15a extension cord from a GFCI outlet outdoors. This setup worked for years with no issues with the same camper, sans MP2.
The GFCI trips after about one minute connected to the trailer. The MP2 chassis is connected to PE at the exact same point as the previous inverter using the same 8 AWG wire. The only thing between the outlet and the MP2 is a 3P 50a MCB per Victron’s wiring schematic - it has not tripped at all. The AC wiring then runs directly to the same AC breaker/branch panel that it did prior to installing the MP2.
The GFCI trips when the MP2 is in charging mode or inverter mode. Even if I set current limits. I am able to avoid tripping at least for a period of time if I disconnect the AC loads including the refrigerator but that is obviously not going to work.
I have to assume the MP2 is causing this and I’m growing increasingly frustrated with it. I don’t want to disable the ground relay unless I’m supposed to but Victrons instructions are very vague here. It seems to me that if the MP2 is behaving as it should, there should be no reason to turn off the ground relay in my setup.
I’m still curious as to why a device like the B-stop is necessary after installing a Multiplus.
My prior setup (with same camper/trailer):
15a GFCI Outlet >
20a Extension Cord >
15-50a SmartPlug Adapter >
50a SmartPlug Trailer Input >
Trailer’s 50a House AC/DC panel
This setup had no issues with GFCI tripping in the last four years.
New setup, same camper:
15a GFCI Outlet >
20a Extension Cord >
15-50a SmartPlug Adapter >
50a SmartPlug Trailer Input >
3-pole MCB (fed with 6/4 SOOW) >
Multiplus II 120v x 2
Trailer’s 50a House AC/DC Panel (fed with 6/4 SOOW)
I cannot get the GFCI to stop tripping with this setup whether the MP2 is in charger or inverter mode. As soon as I start flipping AC loads on, or even just the charger, the GFCI will trip after a minute or two. The refrigerator seems to be the major factor - when it is being powered by the inverter, the GFCI will trip. Currently, the refrigerator is plugged in via a second extension cord to the second outlet on the same GFCI and has been fine for the last several hours. When I connect the rest of the trailer to use the MP2 to charge the batteries, the GFCI trips after about a minute. FYI, the 50a MCB feeding the MP2 has not tripped at all and I’m getting any other signs of a wiring issue.
My understanding is that ground relay shouldn’t be in play when the MP2 is simply passing AC so I don’t think disabling the ground relay will change the ability for the GFCI to stay engaged. It seems like the ground relay is otherwise unnecessary since the current is coming downstream from an RCD when I’m connected at home and the MP2 is in charge-only mode but I would want the ground relay enabled prior to allowing the unit to go into inverter mode. Changing this setting is a hassle for me since I don’t own a Windows computer or a computer new enough to run a compatible version of VictronConnect so I hope that isn’t the solution.
Hi Brad, It is a bit hard to diagnose without knowing the actual leakage currents you are observing. I am assuming the GFCI has a 30mA trip rating. That is maximum it can trip at but can trip at 50% of that rating so 15mA.
It would be good to know what the leakage current is directly on the Mulitplus. The ground relay operation is set as a default so if you havent changed it it is in the correct position.
I will take a direct reading of the leakage when it warms a bit outside. I’m in the US and the outlet is a UL listed device so acceptable leakage ranges from 4-6ma. It is a Leviton 20a self-testing GFCI.
I took direct measurements last night by putting the MP2 in charge mode and shunting the ground on the AC-Out of the MP2 at the splice that I made to tie the wiring back into the RV’s AC panel. The readings were abnormally high, in my opinion, coming in at around 100-150ma. I was able to reduce these by shutting off the main 50a panel breaker feeding the AC branch lines of the RV - this reduced it to about 70ma, still too high. I flipped the disconnect switches for all DC loads and the MPPT chargers (it was dark and they should have been off anyway) and, oddly, the leakage dropped completely to zero.
I don’t entirely trust these readings so ordered a clamp meter to take new measurements. I will also try shunting the ground on the AC-In side of the MP2 to see what the GFCI is seeing.
Other things:
Swapped out the homemade 15-50a adapter for the store-bought adapters supplied by Airstream (this made no difference)
Disconnected the refrigerator altogether from the GFCI and tried to charge batteries with MP2 (this made no difference)
Tried charging the battery with very low DC current of 15a with all loads AC and DC disconnected (this made no difference)
I have to assume that there is a fault in the AC wiring, DC wiring, or both, so I will be going through an extensive cleanup of the old wiring this weekend along with the new wiring. I purchased 6AWG wire ferrules and a crimp tool which I plan to redo the AC wiring with - the spring-clamps are nightmare with stranded wire so its possible I have a stray. I’ll also be capping or removing all the old DC battery wiring and replacing the chassis ground on the MP2 (I have one now going to the negative bus bar on a Lynx Distributor but will beef up the gauge and connector). Otherwise, I’m out of ideas.
Hi Brad. If the GFCI is genuinely 4- 6mA trip it is very low. That may be ok for one device but if that is feeding anything like a distribution board it will trip as the accumulation of leakage current across several devices will trip.
In Australia it is common to have a 30mA rated trip which can trip at 50% of that or 15mA. If we used 4-6mA trip rated RCBs we would struggle to have any large boats at a Marina ever stay connected.
That is the standard for any GFCI outlet in the US. Any outlet installed outdoors must be on a GFCI. A quick browse at my local hardware store shows that all 1-pole 20a GFCI breakers sold are “Class A” breakers which adhere to the 4-6ma US standard. No idea how other countries do it, but everyone plugging their MP2 into an outdoor outlet in the US would be plugging into a Class A GFCI, assuming the installation meets building code.
If Victron is somehow creating a current leak higher than this, then their units are incompatible with any residential circuit with a GFCI in the US. Ockham’s Razor points to the fact that I likely have a wiring fault somewhere - especially considering the high current on the ground I detected downstream of the MP2 with it off.