Inverter/charger and neutral bus question

[mermike]

I’m installing a new Victron Multiplus 12/2000-80 120v inverter/charger and a Blue Sea custom AC panel. The system is still on the bench, so I am confident it is wired per the diagrams below.

When running on shore power (i.e. plugged into the wall), everything is normal. When I unplug from shore power and the system is delivering AC from the inverter, I was getting a reverse polarity signal on the shore power ELCI, I investigated and found about 60 volts between ground and neutral.

The trouble wasn't apparent until I hooked up the first inverter-supported outlet through the new AC panel. I previously ran test loads directly off the inverter. I've double and triple checked everything from the shore power grounding and polarity to the outlets themselves. I contacted Blue Sea, thinking the ELCI was faulty, but I've since taken the entire panel out of the circuit. I disabled AES on the Multiplus and have been testing the output of the inverter directly.

I tried changing the ground relay setting on the Multiplus. It should be enabled, so that the inverter connects neutral to ground in the absence of shore power. I get 60 volts between neutral and ground with it enabled, and 40 v with it disabled. This is concerning - it should be zero if the ground relay is functioning. Shouldn’t it?

The AC wiring diagram is here:

To resolve the issue, I connected all neutrals together. In the diagram, pins 1 and 2 have been jumpered. However, this appears to be incorrect according to ABYC E-31 Figure 4 (Typical inverter/Charger with Internal Transfer to a Split Bus System).

To clarify the wiring diagram:

The 20A GCFI in the AC panel is only connected to shore power. I'm having no issues with this outlet.

Cable ID 23 runs from the ELCI output to the inverter AC input, via the AC Line Terminal. Cable ID 24 supplies AC power to the inverter/charger.

Ground from shore power runs to the galvanic isolator. All grounds are downstream of the GI.

The neutral in Cable ID 35 feeds all inverter neutrals from the inverter/charger. The ground in Cable ID 35 feeds all AC grounds from ship’s ground.

Cable ID 35 connects the inverter/charger to the AC Line Terminals. ID 69 feeds the 2 inverter-supported AC circuits in the AC Panel from the AC Line Terminals.

I’m missing something here, but can’t figure out what is. The neutral should be split, but it doesn’t work that way.

 

[IStream]

I'm trying to understand your troubleshooting sequence. If you disconnect everything from the inverter and measure the inverter's outputs directly, do I understand correctly that you see a voltage between neutral and ground? If so, the inverter is the problem. If not, I'd start adding components back one at a time from the inverter outwards until you see the voltage re-appear.

 

[mermike]

Sorry, I was trying to be brief. Yes, I disconnected everything and tested from the inverter. It's operating properly.

 

[mermike]

I borrowed a circuit analyzer and the output from the inverter checked out. Voltage, frequency, harmonic distortion, and resistance were all acceptable.

 

[mermike]

Back to the question I asked, is my design assumption, that the neutral should be isolated, correct?

 

[Stu Jackson]

Mike, this is a recurring question. Many of us see RP when the inverter runs. I wish to heck I could recall why, and also why it is not a problem. I'll see if I saved any files about this. In any event, I hope others will chime in, too.

 

[mermike]

Stu, that's encouraging. My first assumption was that I missed something obvious. Hope you find something to shed light on it.

 

[mermike]

OK here's a little more information (Thank you to Justin @ Bay Marine in San Diego), but no solution. Perhaps @Maine Sail can weigh in. Here's the relevant ABYC schematic from A-31 Battery chargers and Inverters. This is pretty much how the system is wired.

When the internal transfer switch in the inverter/charger changes from shore power to inverter mode, the incoming neutral becomes free-floating. Now the question is, if the supply neutral is open and has no current, why would the polarity indicator light up to show that it does have current, reverse or otherwise? It's literally a wire without a connection, except at the polarity indicator itself.

 

[SycloneDriver]

To the best of my knowledge the reverse polarity lamp will come on when the neutral and the ground are not bonded and there is a voltage difference between the two.

The inverter output is essentially +60v on the line and -60v on the neutral, it alternates of course. If you bond the neutral and ground the light should go out. It sounds like you inverter isn't bonding the neutral and ground completely. This doesn't sound normal to me.

If you search you-tube for neutral ground bond and Honda generator you may get some of your answers. I remember seeing this issue discussed in several videos. I still need to make a neutral/ground bonding plug for my generator.

 

[mermike]

The neutral and ground are bonded on shore when running from shore power. The neutral input to the inverter is disconnected when the transfer switch connects the inverter. The output neutral and ground are bonded at the inverter. Connecting the input neutral and output neutrals solves the problem, but is this a safe and effective solution? Is there a downside to it? How can there be a voltage difference if the input neutral has no connection?

 

[SycloneDriver]

Connecting the input and output neutrals is probably not the proper solution. The downside is that I doubt it meets ABYC standards.
I don't understand your last question.

 

[mermike]

If the input neutral is disconnected (there is no shore power connection, and there is no inverter connection), there should be no current on the input neutral, and thus no differential for the polarity indicator to measure between the input neutral and ground. Correct?

Connecting the input and output neutrals is probably not the proper solution. The downside is that I doubt it meets ABYC standards.
I don't understand your last question.

 

[SycloneDriver]

If the input neutral is disconnected (there is no shore power connection, and there is no inverter connection), there should be no current on the input neutral, and thus no differential for the polarity indicator to measure between the input neutral and ground. Correct?

Yes

 

[mermike]

Finally found the problem. It was what I like to call the ID10T error. Here's the panel schematic from Blue Sea:

The circled neutral connection was wired to the input neutral, not the inverter output neutral. This put current on a wire that should have been disconnected when the inverter/charger was in inverter mode, lighting the polarity indicator. Bonding all the neutrals was one solution, but not the correct one. Glad I kept digging.