Dangerous leaking voltage? on Magnum MS2000 inverter/charger

[cp3sail]

Hi cruising friends,


I appreciate all the advice as i have lurked and read posts over the years.


I wanted to add this (issue?) with my Magnum MS2000 inverter/charger, as I searched for answers and found none. So as the story progresses i will post results and hopefully this can help someone in the future.


The unit has been in service for a few years. It works like a drafthorse. Big HEAVY loud and trustworthy. It powers the splendide washer and the vitamix. What more could i ask for?


Well then i dove the boat one month and POOF no zincs on the frigoboat keelplate. That was a fast degredation i thought to myself and so i went about looking at where i have DC grounding shared with AC lines. (I realize this could be pure coincidence).



I have only two shared ground points.

- The main DC bus is grounded to the main AC ground bus at the main panel.
- The Magnum Inverter/Charger has Case ground. This is Grounded to the AC ground Bus AND the DC ground bus per ABYC recs.


- ALSO, side note, when the inverter/charger is in "inverting" mode, a relay closes and connects the AC Neutral to the AC ground. So at that point my AC ground, AC neutral and DC ground bus are all connected.



So, I started taking wires aside and clamping my AC and DC ammeter around them to peek for stray current.



I didn't find anything larger than .1A which my meter seems to flash intermittently when the ammeter is over thin air anyhow. So I began taking the grounding wires off and checking voltage potential readings.



THIS IS WHERE IT GOT INTERESTING.

(read: this is where i realized i was in over my head already, and really just wanted my multi-meter to beep at me. I might as well have probed the shore transformer had there been bare contacts available for my overzealous meter probes)



So I pulled the DC ground bus to AC ground bus jumper at the main panel. Nothing showed on meter.



So I pulled the DC and AC equip ground off of the Magnum Inverter/Charger Case, and got a light shock. Hmm!
- Meter check read 37.8 VAC between the (disconnected) AC ground wire and the Case ground lug. Charger was in Standby mode, batteries at 13.5 VDC.

- So I turned off the breaker in the Black Line-In to the Magnum, Meter read 0.0 VAC between the (disconnected) AC ground wire and the Case ground lug.
- So I rotated my multi contact Shore/Inverter/Off switch to Inverter back at the main panel and set the Magnum to Inverter mode. Meter read 14.0 VAC between the (disconnected) AC ground wire and the Case ground lug.
- So I called Magnum.


(insert elevator music here. 2hrs later, talked to techs, they said loose connectors or a ground loop likely or replace the AC circuit board in the unit)

I wasn't buying that idea. Over the phone we tested for ground loops, there were none. I have very few connections in the AC system, all new 3yrs ago, all checked out and felt good.


-So I pulled the unit off the wall and pulled the cover off to isolate it and bench test it repeatably. (I am 8mos out of my 3 year warranty)







https://drive.google.com/file/d/1nMn...ew?usp=sharing





https://drive.google.com/file/d/1PKL...ew?usp=sharing



As you can (hopefully images loaded) see the internals are easy enough to probe with my multimeter of curiosity.



I began to test from the Case lug to the internal leads (now exposed by removal of cover), instead of the Case lug to the external AC ground wire.





1. (Inverter/Charger OFF, ALL wires disconnected)


- Hot Line In (black) to the Case: OL ohms, 0.0 VAC.
- Neutral Line In (white) to the Case: OL ohms, 0.0 VAC.

- Neutral Line In (white) to the boat AC Ground: OL ohms, 0.0 VAC.





2. (Inverter/Charger ON, AC wires Connected, DC disconnected)


- Hot Line In (black) to the Case: 1123 Ohms, 95 VAC. (this seems bad as case is currently un-grounded)

- Neutral Line In (white) to the Case: OL ohms, 3.7 VAC. (this seems bad)

- Neutral Line In (white) to the boat AC Ground: 0.0 VAC.

- Case to the boat AC Ground: 3.7 VAC. (this seems bad)

* (note. this has dropped from the 35 VAC seen when the batteries were

connected in charger standby mode, and the 14.0 VAC seen when
batteries were connected in Inverter Mode)
- Hot Line In (black) to the boat AC Ground: 120 VAC. (as expected)
- Hot Line In (black) to Neutral Line In (white): 120 VAC. (as expected)


So, this was the information which I had to present.

Thank you for reading this far through.



I can make many more tests and photos as requested.
Thanks All!

[cp3sail]

I should add, what i think is happening is that AC current is leaking from the circuit board to the case via the circuit board grounding.


This is based on seeing a resistance of only 1123 Ohms between the AC Line In (black wire) and the Case when the unit is powered up. (this resistance rises to OL when the unit is powered off)



The voltage leakage also Increases when DC voltage is hooked up to the unit. So i worry that replacing the 300$ "AC Circuit Board" may not solve the problem. As leakage seems to also be related to the DC side of things.

[cp3sail]

Here are the rest of the links to the photos of the meter leads hooked up with AC power:


https://drive.google.com/drive/folde...FQ?usp=sharing

[cp3sail]

Oh the MOST IMPORTANT thing I wanted to get across to other with these Magnum units in service is this:

The Charger function and Inverter Function Both appeared to still work flawlessly.



I had NO INDICATION that current was leaking to ground, and presumably to the water as well.



CHECK YOUR GROUND TO CASE VOLTAGE!?



This seems very bad for the underwater metals and probably unsafe at ~30VAC potential.
I have GFCI's on all circuits and reverse polarity indicators on the main AC panel.

[smac999]

wonder if a main shore ELCI / RCD would trip

ac leakage doesn't really hurt metal. it's DC leakage that does. AC leakage is harmfull to people in the water. though your inverter can not leak to water. as it has to return to souce. it may on shore power though as the current can leak back to land. (unless you have an isotransformer)