Open neutral while on inverter

[stevemitchell]

I have been slowly fixing small electrical items since I bought my boat, but this one has been confounding me for quite a while.

I have two 110VAC outlets that show an open neutral while away from shore power and on inverter. In any other condition, they test out fine. While showing open neutral, of course, the outlets will not work with anything plugged into them, as you would expect.

I have a Magnum 2800W inverter wired into my main AC panel. AC shore side and generator power is 110VAC.

The inverter output goes to a separate set of circuits including all of the 110VAC outlets on the boat, ice maker, and microwave.

What doesn’t make sense to me is why they show open neutral only while on inverter when no shore or generator power is present. From what I have been able to trace, the neutrals look fine, and are connected the correct way. Nothing else on board looks funky or behaves strangely. While connected to shore power, everything looks fine.

I have used both cheap plug-in testers with lights and my more expensive meters to validate that it is definitely an open neutral.

I have worked on many a Magnum inverter with all of the boat consulting I do, but I have never had one myself for any length of time. I’m not sure if there is something intrinsic to the Magnum that might be at play here, but I would guess not. This seems like something fundamental with the neutral side of things, but I have not yet been able to suss it out….

 

[twistedtree]

When the neutrals are open, are they also isolated from the ground wire?


And if you measure from the hot wire on the affected outlets over to a good neutral, is the hot wire actually hot?


My guess would be that those outlets have their neutrals wired to the inverter's input side neutral bus, not the output side. The inverter should have two separate neutral buses, just as it has input and output buses for the hot wires. The neutrals are connected when the inverter in in passthru mode, i.e. when there is shore power, and isolated when inverting. This matches your symptoms exactly.


So you will have to find where those outlet's neutrals end, and move them to the inverter output neutral bus. Hopefully it's not super hard to do, but might be, and might be why it was never done, or done incorrectly.

 

[STB]

These inverters, like most (I think), have relays that switch neutral. When on shore power, they connect neutral to shore neutral, which eventually ties it to ground. When inverting, they disconnect neutral from shore neutral and, instead, connect it to the inverter's chassis ground. So, when inverting, neutral is bonded to ground at the inverter, and when on shore power, it is bonded to ground at shore.

Regardless, if neutral is open while inverting, but not otherwise, the first thing I might start to think about is this relay and what the ground screw on the inverter is, or is not, connected to. Also, what happens when you back all the way up to the inverter and test at the screw terminals? Another possibility is just that something is disconnected along the way.

Specific configuration settings aside, I think inverters most often have the relay in "inverter mode" normally (at start up), until AC input power is detected and switches it to "shore power" mode. This keeps it open until the inverter knows it is safe to close it.

One other consequence of this is that the neutral bus for stuff running off of the inverter needs to be separate than the neutral bus for things that aren't running off of the inverter. Otherwise, the neutrals for things that aren't on the inverter and things that are on the inverter are tied together at a common neutral bus, defeating this relay.
In the past, many inverters were wired this way, and there were no "apparent" ill effect.

But, these days, ELCI breakers, whether aboard or on shore power, are becoming ore common and will trip (and properly so) if wired this way. So, although in the "olden days" configurations without these relays or that tied the connections together after these relays might have worked -- they are more likely to trip breakers as ELCI becomes more common.

When I bought my boat, the inverter had been removed and replaced with a simple charger. When I replaced the inverter, wiring it exactly into the terminals provided and labelled for the original in the main wiring locker -- it immediately popped the ELCI I had installed a few days earlier. When I checked the rest of the wiring, I saw the shared neutral bus, corrected it, and everything was good.

The prior prior owner who had set everything up was a very knowledgeable marine professional and had actually put in an external relay to switch the neutrals and everything else, even in the days when those weren't built into the inverters and not common. And, when the boat was rewired, the next owner followed this same pattern, just removing the relay and using the internal one in the inverter he had installed (and, I think had been replaced with a cheap charger for sale). But, ELCI's weren't around back when this was originally done.

So, I guess another possibility is that you do have the two neutral buses, but those outlets are wired to the wrong one, or have their neutrals wired back to the ground bus instead of the neutral bus, which absent an ELCI could seem to work -- but is still messing up the point of the safety ground.

 

[stevemitchell]

Ah I can’t believe I hadn’t thought of this. Between your two replies I think I may have an idea of what’s wrong.

I am betting these two outlets have their neutral wired into the shore side and not the inverter side.

It is very hard to see the distribution end of these but that’s going to be my next step to see if that is the case.

After installing and working on many inverters in my life I can’t believe I forgot about this simple setup piece. Guess getting older had some cons.

I also removed a dangerous open relay a few days ago that was connected into this mess. It shocked me once and that was one too many times. No cover or anything. It allowed you to bypass the inverter in the event it completely failed and provide power to the isolated items. However it had no cut out protection, so you could engage it while the inverter was on! Which is bad.

Cleanup will continue.

 

[sunchaser]

Steve
Did the PO have that plug reserved for something that he didn't want tied into an older MSW inverter but genset or shore power only?

The neutral/ground switching relay in the Magnum 2800 can affect the new dock GFCI configurations due to the millisecond relay switching delay. This delay can, but not always, trip out the dock breaker when hooking up your shore power. The solution for me has been to insure inverter is off when hooking up to newer shore power setups.

 

[STB]

Hey Sunchaser,

I don't know the Magnum 2800 series, in particular, but, if things are set up properly, I really don't think it should trip GLCI/ELCI breakers. I suspect what is going on in your case is that you have one neutral bus for both inverted and non-inverted loads. This set-up was common before ELCIs/GFCIs became common, so a lot of boats are set up that way. But, these days, as you've discovered, it doesn't always work out so well. Adding a second neutral bus exclusively for the loads that can be run off of the inverter is the quick, easy, and correct fix.

I have had two different models of Xantrex Inverters (ProSine 2.0 and Freedom Xi 2000 Pure Sine) and neither tripped my ELCI, they both worked fine with ELCI, and I don't think they were special. I've also worked on boats, with a variety of inverters that were having this problem with shore power, e.g. while cruising to more modern marinas than the home marina, and it was always the same situation -- a neutral bus shared by inverted and non-inverted loads.

When I take a quick look a the Magnum 2812 manual online, it seems to work the same way as all of the rest. Take a look at Figure 4, page 6 of the Magnum MS manual (page 12 as the pdf viewer counts it) and the discussion above it.
-- https://www.magnum-dimensions.com/sites/default/files/MagArchive/64-0007-12-04-MS-Manual.pdf

Notice that the inverter's relay bonds the /output/ neutral to the ground. It does not bond the input neutral to anything but output neutral, ever.

And, that is the trick. On the /input/ shore power side, neutral and ground are /always/ independent. So, from the perspective of an ELCI/GFCI all of the current leaving via the hot is coming back via the neutral, none of it is going through the safety ground around the ELCI/GFCI -- even at start-up.

Where things usually go wrong is with boats that have had one inverter or another for many years. In these cases the set-up dates back to a time before ELCI/GFCI was common. In these cases, the inverted and non-inverted loads often share a common neutral bus. A bit of current returning through the ground for the short period before the relay flipped was, I guess, viewed as just no big deal back in "the day".

But, because a shared inverted/non-inverted neutral bus ties the output side neutral to the input side neutral, when the inverter is bonding the output side neutral to the output ground -- the shared neutral bus ties the output side neutral, which is bonded to ground by the relay, to the input side neutral, thereby bonding the input side neutral to input shore-side ground. Then the ELCI/GFCI pops because some current is returned via the ground, which is other-than-through neutral it is monitoring via a toroid coil.

By waiting to turn the inverter on until the relay has already switched to shore power mode, what i think is happening is that you are letting the inverter detect shore power and go into shore power mode before coming on, so it never ties the output neutral to the output ground for the shared neutral bus to tie the input neutral to the input ground.

I'm actually a little surprised this works. On my current inverter, for example, the relay's "normal" position is to bond the neutral and ground, and it doesn't unbond them until it detects shore power. And, this is true on my inverter, even when the inverter is turned off. I think the inverter on my old boat worked the same way -- except there was a jumper one could move to change the behavior (I think moving that jumper probably violated NEC and other codes/guidelines, at least in situations where they apply).

At any rate, I think you've got a neutral bus that is shared by your inverted and non-inverted loads and if you add another bus and separate the two, you'll have wiring that'll make things work effortlessly for you and meet more modern "standards".

 

[twistedtree]

Just a couple of cautions re inverters...


- On some, the input and output neutrals are common and always conencted. Outbacks are like this


- On others, the input and output neutrals are separate and switched together only when in passthrough mode. Magnum appears to be like this, as is Victron.


- Some models have internal relays to break/create the neutral/ground bond.


- Some models have no neutral/ground bonding and it needs to be dealt with externally.


Need to look closely at the specific model.

 

[Maerin]

Ah I can’t believe I hadn’t thought of this. Between your two replies I think I may have an idea of what’s wrong.

I am betting these two outlets have their neutral wired into the shore side and not the inverter side.

It is very hard to see the distribution end of these but that’s going to be my next step to see if that is the case.

After installing and working on many inverters in my life I can’t believe I forgot about this simple setup piece. Guess getting older had some cons.

I also removed a dangerous open relay a few days ago that was connected into this mess. It shocked me once and that was one too many times. No cover or anything. It allowed you to bypass the inverter in the event it completely failed and provide power to the isolated items. However it had no cut out protection, so you could engage it while the inverter was on! Which is bad.

Cleanup will continue.

And, if the neutrals are mixed up, it can also pop the ELCI on updated marina wiring (I know this from 1st hand experience ) Whoever installed the inverter on my boat didn't separate the neutrals, it was a tedious mess to get resolved. When separating those neutrals, don't forget the LED pilot lights on the panel. If they're on an inverter powered circuit, the tiny neutral wire for the lamp must also be connected to the inverter neutral buss. Yes, they're only a couple of milliamps.... how many are there? A 4 or 5 mils apiece, you can be up to enough to trip a GFCI in short order!

 

[Pcpete]

I added a second buss for the inverter circuit neutrals. Once I had the inverter on what’s basically a sub panel, the timing backfeed and switching were no longer an issue. No problem at Poulsbo, Roche or Rosario.

 

[sunchaser]

Hey Sunchaser,

I don't know the Magnum 2800 series, in particular, but, if things are set up properly, I really don't think it should trip GLCI/ELCI breakers. I suspect what is going on in your case is that you have one neutral bus for both inverted and non-inverted loads.

GK
Two years ago the neutral loads on our vessel were "split up" so that a separate neutral buss was installed for inverter loads only. You're a lot smarter on this stuff than I am, but my experiences on many docks in BC, WA and AK has shown that not all behave the same - even when presumably built to the same specs and code. So shutting down everything on the boat and then plugging in seems to do the trick at the stubborn places we encounter.

Obviously vessels with isolation transformers are a different story, or better said, non issue. I chose the "clean up the grounds" route and with a bit of powering down tom foolery as I mentioned, all seems to work - without an IT.

 

[STB]

Hey Sunchaser,

One thing is for sure -- I sure as heck ain't going to argue with experience. "Proof by example" always beats "proof by thought experiment." There are jokes about "scientists" who continue to believe their own thought experiments over real world empirical ones. As the old saying goes -- never argue with what works (or doesn't)!

I can only guess there is some capacitive or inductive reactance that is somehow stacking up in a way that the GFCI/ELCI sees in a way it doesn't if you energize things one at a time. But, if just the inverter does it -- I'm even out of guesses.

Regardless, since you know that your buses are separated and nothing is cheating that -- I'm out of real ideas!

Sorry for doubting you!

Cheers!

-Greg

 

[bglad]

I think you are on the money if as you describe there is a separate inverter buss. The open neutral defect detected on inverter circuits while it is the power source must be because the circuits are not attached to the inverter neutral buss. But, I think one would verify the inverter output prior to diving into rewiring.

IMO having an inverter bypass switch is a good idea. If the inverter decides to not allow power through you can get around it. Or, if you are away from your boat for some time bypassing it might avoid inadvertently discharging your batteries.

Ah I can’t believe I hadn’t thought of this. Between your two replies I think I may have an idea of what’s wrong.

I am betting these two outlets have their neutral wired into the shore side and not the inverter side.

It is very hard to see the distribution end of these but that’s going to be my next step to see if that is the case.

After installing and working on many inverters in my life I can’t believe I forgot about this simple setup piece. Guess getting older had some cons.

I also removed a dangerous open relay a few days ago that was connected into this mess. It shocked me once and that was one too many times. No cover or anything. It allowed you to bypass the inverter in the event it completely failed and provide power to the isolated items. However it had no cut out protection, so you could engage it while the inverter was on! Which is bad.

Cleanup will continue.

 

[sunchaser]

Great thread with some smart posters, I'm not one.

We berth our vessel in BC. The dock wiring and codes there are not (currently) following the US as far as GFCI breakers set to 30 ma. In talking with the BC Hydro guys they say why do it, their codes work fine with no electrocutions or marina violations apparent. Canadian marine techs are very supportive of inverter bypass switches and beat me up a few years ago for not having one.

Our marina checks leakage routinely for offending vessels and sends electricians on board to immediately rectify issues if incipient issues detected. Divers cannot go in the water without an out of the water "watcher." A proactive approach that addresses the issue, if there is one, head on.

 

[stevemitchell]

Steve
Did the PO have that plug reserved for something that he didn't want tied into an older MSW inverter but genset or shore power only?

The neutral/ground switching relay in the Magnum 2800 can affect the new dock GFCI configurations due to the millisecond relay switching delay. This delay can, but not always, trip out the dock breaker when hooking up your shore power. The solution for me has been to insure inverter is off when hooking up to newer shore power setups.

No I don't think he wired it for that purpose. These are pretty random places, and he even said that they hadn't figured out why they didn't work away from the dock.

Part of the reason I'm delving into this is to add a galvanic isolator and ELCI breaker. I am 100% sure that if I added in the ELCI, it would trip pretty frequently because of this and other potential wiring issues. I have not had the inverter cause the shore side to trip yet, but I haven't visited many remote marinas who have the updated pedestals.

 

[stevemitchell]

And, if the neutrals are mixed up, it can also pop the ELCI on updated marina wiring (I know this from 1st hand experience ) Whoever installed the inverter on my boat didn't separate the neutrals, it was a tedious mess to get resolved. When separating those neutrals, don't forget the LED pilot lights on the panel. If they're on an inverter powered circuit, the tiny neutral wire for the lamp must also be connected to the inverter neutral buss. Yes, they're only a couple of milliamps.... how many are there? A 4 or 5 mils apiece, you can be up to enough to trip a GFCI in short order!

That's a good point on the tiny lights used in many places in the panel!

IMO having an inverter bypass switch is a good idea. If the inverter decides to not allow power through you can get around it. Or, if you are away from your boat for some time bypassing it might avoid inadvertently discharging your batteries.

I completely agree, but this implementation is dangerous in several ways. First, the relay itself is on the inside of my electrical cabinet, hidden on the wall facing you, so if you reach in, you get shocked. I know, I've had it happen twice. It needs a cover, but it is also way too large and should be replaced with something more current and smaller (and safer). Second, there is no lock out on the breaker, which is mixed in with other ones on the AC panel. You can easily turn it on, only to have created a short circuit that makes all sorts of things go bad - loud zapping noises, circuit breakers popping, and more.

My long term plan may simplify all of this. Like several other boats I've had, I am considering using a Victron Quattro inverter/charger and setting it up where all AC power (both generator and shore side) routes through it first, after going through an ELCI breaker and galvanic isolator. One output from the inverter goes to the always-available/inverted panel, and the rest goes to the standard AC panel. This simplifies the wiring, provides a very nice automatic AC switch between shore/generator, and with the addition of Victron's BMV battery monitor and Color Control GX display, adds a wealth of info on the electrical system performance. It also allows me to balance my shore power usage a lot more - one of my biggest challenges right now is dealing with the 30 amp dance.

I'm also going to be replacing all of my batteries with LiFePO4, upgrading the alternators and adding external regulators, and a few other things to support this.

With this config, I may still need separate busses for the neutral, although I can't remember my Victron setup from the last boat, I think you might not need that. Regardless, I will still be tracing out how the current setup is wired, as I don't plan the above updates for a month or two.

 

[Pcpete]

Sunchaser, the change in the US has little or nothing to do with electrolysis. It has to do with fresh water drowning due to stray current. It’s another of those laws where they think marinas cant figure out if they are on fresh water or salt so they make it a rule everyone has to follow.
My understanding is because of the poor communication about the new code, the implementation has been slowed.
When I was trying to confirm the correct action with my system, I plugged, inserted, a 2ma contractor gfi into the shore power line and figured if I could keep the 2ma gfi from blowing, ten would be no issue.

 

[twistedtree]

With this config, I may still need separate busses for the neutral, although I can't remember my Victron setup from the last boat, I think you might not need that. Regardless, I will still be tracing out how the current setup is wired, as I don't plan the above updates for a month or two.

Separate neutrals aren't necessarily needed, and a common neutral isn't necessarily a bad thing. It all depends on the approach to handling AC power switching, adn the particular equipment used. Much of the switching is actually to deal with neutral/ground binding, and the neutral ends up getting split out as one way to deal with that.


If you have space and are so inclined, an isolation transformer can allow you to significantly simplify the on-board wiring, have a common neutral, and have a fixed (not switched) single bonding point between neutral and ground. Plus of course all the isolation benefits, and the option to take in either 120V or 240V shore power.


And speaking of bonding, with a separate inverter neutral, are you bonded on board when on inverter power or generator power? And are you not bonded on board when on shore power? I saw in the Magnum manual that the inverter does not do bonding as some other inverters do.

 

[sunchaser]

Sunchaser, the change in the US has little or nothing to do with electrolysis. .

I clearly understand the purpose.

 

[stevemitchell]

Separate neutrals aren't necessarily needed, and a common neutral isn't necessarily a bad thing. It all depends on the approach to handling AC power switching, adn the particular equipment used. Much of the switching is actually to deal with neutral/ground binding, and the neutral ends up getting split out as one way to deal with that.


If you have space and are so inclined, an isolation transformer can allow you to significantly simplify the on-board wiring, have a common neutral, and have a fixed (not switched) single bonding point between neutral and ground. Plus of course all the isolation benefits, and the option to take in either 120V or 240V shore power.


And speaking of bonding, with a separate inverter neutral, are you bonded on board when on inverter power or generator power? And are you not bonded on board when on shore power? I saw in the Magnum manual that the inverter does not do bonding as some other inverters do.

I had an isolation transformer 2 boats ago, and it was definitely a wonderful piece of equipment to have for the reasons you cite, among others.

I haven't settled on my design yet for this boat, but I have a galvanic isolator from another install that never happened and was considering using that, fully aware that it is not a replacement for an isolation transformer. I did find a nice compact isolation transformer from Charles which I might be able to fit into my design, although I am already having challenges with the space for the inverter

Well it is interesting you brought up bonding, as there is a major problem there that I have been working to fix as well. Not only is it not bonded when away from the dock and on inverter power, but when the generator is on, almost all outlets (haven't confirmed all of them yet) have open grounds.

I'm pretty sure that when this Magnum was added in the last few years, some hacks were put in to accommodate it as the previous inverter was very simple and old, and it looks like a few things were... missed, to say the least

 

[STB]

Out of curiosity, what type of switch to folks normally use for the inverter bypass? I've never found one that fully does the job designed for marine use. The ones I normally see are bulky ones designed for commercial use on shore.

 

[sunchaser]

Out of curiosity, what type of switch to folks normally use for the inverter bypass? I've never found one that fully does the job designed for marine use. The ones I normally see are bulky ones designed for commercial use on shore.

Mine is a Blue Seas rotary switch mounted in a Scepter junction box.

 

[STB]

Hey Sunchaser,

Which wires does it switch?

Thanks!

-Greg

 

[sunchaser]

Hey Sunchaser,

Which wires does it switch?

Thanks!

-Greg

Greg
The best I can tell you is the inverter 110 inputs and outputs are connected in the bypass mode. That said, the details are in the brain of my very good installer. Obviously the charging function is not functional in the bypass mode.

 

[STB]

Hi Sunchaser,

Thanks so much!

-Greg

 

[stevemitchell]

I was able to track down one of the two outlets wiring. It required my endoscope camera since much of the neutral wiring is at the bottom of my electrical closet. After some wire tracing and testing, I found that it was on the wrong side of the neutral bus, connected to shore-powered-only devices instead of the inverter neutral bus, as suspected.

Moving it solved that outlets problem.

The other one is still a mystery and appears to be wired correctly, so I will be digging a bit deeper.

Meanwhile, my new inverter/battery design, which will be ordered here in about 2 weeks, has all shore and generator power going directly into the inverter (Victron) to help with load balancing and overall power control, and will eliminate the need for separate busses here anyhow.

 

[sunchaser]

Meanwhile, my new inverter/battery design, which will be ordered here in about 2 weeks, has all shore and generator power going directly into the inverter (Victron) to help with load balancing and overall power control, and will eliminate the need for separate busses here anyhow.

Do you mean all power? Not knowing your boat, but on mine that would include AC, water heater, washer/dryer, electric stove and engine battery charger - which are on independent 110/220 circuits and intentionally do not go through inverter.

 

[stevemitchell]

Do you mean all power? Not knowing your boat, but on mine that would include AC, water heater, washer/dryer, electric stove and engine battery charger - which are on independent 110/220 circuits and intentionally do not go through inverter.

Yes many newer inverters including the last two or three I’ve had from Victron and MasterVolt allow you to route your main shore or generator directly to the inverter. Then there are two AC outputs - one for inverted stuff and one for non inverted stuff like water heater, etc.

The benefit is that the inverter can see ALL power being consumed and change how much it is using for charging to ensure you don’t trip the upstream breaker. In addition some inverters can use battery power to augment if you don’t have enough from shore which is nice.

It also gives you good visibility on overall power usage. The downside is if the inverter dies you have a single point of failure.

I’ve had this setup on the last two boats I’ve owned and love it.

 

[twistedtree]

And of course a bypass switch will address the single point of failure risk.

 

[stevemitchell]

And of course a bypass switch will address the single point of failure risk.

Yup it sure will! I plan on having a proper one rather than the crazy one installed now.

Any suggestions for one appreciated!

 

[twistedtree]

Yup it sure will! I plan on having a proper one rather than the crazy one installed now.

Any suggestions for one appreciated!

Nordhavn does it with an ABB switch. I don't know part number, or cost. You could do it with a rotary switch too, I'm sure. Seltzer I think is the name of a company with a wide range of rotary switches. I've also seen it done with breakers with a mechanical interlock, which is probably the least expensive approach.