EMF issues with Victron Multiplus II. No case grounding atm, confused where to ground it...

[vas]

OK, title is trying to describe the issue, obvs not enough, so bear with me.

mobo is twin screw on shafts, hull is plywood on iroco frames, built in 74-75, 24V.
Existing el.panel restored, lots of useless wiring removed, working fine (he says!)
All through-hulls are GRP (Tru-design).
Boat is NEVER on shore power as I need to run 100m+ cables from nearest pillar and I don't like that, and I don't need that with 600W solar in the med.

There's a new (well new when rebuilt was done in 2015) reasonably thick tinned ground cable doing a loop from:
battery negative (4X6V Trojan T-105RE bats are just aft of port engine)
port engine block
port shaft seal (bronze)
port p-bracket (bronze)
port rudder (bronze)
port transom 1kg anode
stbrd transom 1kg anode
stbrd rudder (bronze)
stbrd p-bracket (bronze)
stbrd shaft seal (bronze)
stbrd engine block

There's obvs an even thicker (thumb thickness!) neg connection between blocks and neg battery which also goes to generator block (aft of stbrd engine)

Multiplus II 3000/24 is just aft of port engine and batteries wired with short massive cables to the batteries through a Victron BMV700.

There's also a raspberry-pi running VenusOS v3.0 and a Victron 100/30 MPPT controller from two serial wired 1X2m 300W panels with OV of circa 48V ea)
Finally since last season I've got a custom built 8S EVE 304Ah lifepo4 bank, with dedicated T-class fuse, it's own current monitor and bms (diyBMS - opensource) and a BlueSea contactor.
All the Victron stuff and the diyBMS are wired and communicating nicely together with bms instructing charge/discharge profile to Multi II and MPPT.

Amazing as it may seem, all works fine and I've got ample power to do whatever I want onboard. OK, took a couple of lockdowns to finalise the lifepo4 conversion, but the result is what counts

Now, following this long intro, into the Q.
When Multi is working charging mobiles, running laptops and doing silly low amp stuff, all's fine, however when I run largish tools or even worse when I run the 1800W watermaker (via an inverter through a 3phase motor to enable soft start and save the membranes) I get some serious EMF.
i can see that on my various arduino like boards (teensy to be exact) which are monitoring various engine/tanks/weather/etc states and converts them into the NMEA2000 network so that I can view data from anywhere onboard.
So a 20% fuel may start fluctuating form 15 to 25 for the duration of the high amp draw operation.

N2K bus takes it's power from the stock el.panel from the lower helm, same as all other devices.
If I run the generator (8kva) which tbh have no reason anymore to do so, and Multi is in passthrough mode, things are much better.

From my understanding this is a case of grounding issue for the multiplus and here is my Q:
what/how do I do it?
Option 1 (which I'm doing right now) is discard measurements for whichever duration of inverter is loaded... OK, works, but doesn't feel right tbh.

Option 2 is to most likely ground the Multi II case. Went through some Victron docs it's a bit confusing and looks like the only realistic solution would be to wire the case to the cable looping all metalwork at sea. I have no extra isolating "plate" in the water as some newer craft have and wont be able to do it before next season lift.

Option 3 is to ground the Multi II to the bat negative (which is more or less the same as Option2 tbh).
Not sure if #2 or #3 is going to be worse for ground loops...

suggestions please before I start testing next week that I'll be onboard!

and please don't need anyone to tell me my system is too complicated and I shouldn't have it like that, cause it's really pointless and not what I'm asking...

cheers

V.

 

[PaulRainbow]

OK, title is trying to describe the issue, obvs not enough, so bear with me.

mobo is twin screw on shafts, hull is plywood on iroco frames, built in 74-75, 24V.
Existing el.panel restored, lots of useless wiring removed, working fine (he says!)
All through-hulls are GRP (Tru-design).
Boat is NEVER on shore power as I need to run 100m+ cables from nearest pillar and I don't like that, and I don't need that with 600W solar in the med.

There's a new (well new when rebuilt was done in 2015) reasonably thick tinned ground cable doing a loop from:
battery negative (4X6V Trojan T-105RE bats are just aft of port engine)
port engine block
port shaft seal (bronze)
port p-bracket (bronze)
port rudder (bronze)
port transom 1kg anode
stbrd transom 1kg anode
stbrd rudder (bronze)
stbrd p-bracket (bronze)
stbrd shaft seal (bronze)
stbrd engine block

There's obvs an even thicker (thumb thickness!) neg connection between blocks and neg battery which also goes to generator block (aft of stbrd engine) Multiplus II 3000/24 is just aft of port engine and batteries wired with short massive cables to the batteries through a Victron BMV700.

There's also a raspberry-pi running VenusOS v3.0 and a Victron 100/30 MPPT controller from two serial wired 1X2m 300W panels with OV of circa 48V ea)
Finally since last season I've got a custom built 8S EVE 304Ah lifepo4 bank, with dedicated T-class fuse, it's own current monitor and bms (diyBMS - opensource) and a BlueSea contactor.
All the Victron stuff and the diyBMS are wired and communicating nicely together with bms instructing charge/discharge profile to Multi II and MPPT.

Amazing as it may seem, all works fine and I've got ample power to do whatever I want onboard. OK, took a couple of lockdowns to finalise the lifepo4 conversion, but the result is what counts

Now, following this long intro, into the Q.
When Multi is working charging mobiles, running laptops and doing silly low amp stuff, all's fine, however when I run largish tools or even worse when I run the 1800W watermaker (via an inverter through a 3phase motor to enable soft start and save the membranes) I get some serious EMF.
i can see that on my various arduino like boards (teensy to be exact) which are monitoring various engine/tanks/weather/etc states and converts them into the NMEA2000 network so that I can view data from anywhere onboard.
So a 20% fuel may start fluctuating form 15 to 25 for the duration of the high amp draw operation.

N2K bus takes it's power from the stock el.panel from the lower helm, same as all other devices.
If I run the generator (8kva) which tbh have no reason anymore to do so, and Multi is in passthrough mode, things are much better.

From my understanding this is a case of grounding issue for the multiplus and here is my Q:
what/how do I do it?
Option 1 (which I'm doing right now) is discard measurements for whichever duration of inverter is loaded... OK, works, but doesn't feel right tbh.

Option 2 is to most likely ground the Multi II case. Went through some Victron docs it's a bit confusing and looks like the only realistic solution would be to wire the case to the cable looping all metalwork at sea. I have no extra isolating "plate" in the water as some newer craft have and wont be able to do it before next season lift.

Option 3 is to ground the Multi II to the bat negative (which is more or less the same as Option2 tbh).
Not sure if #2 or #3 is going to be worse for ground loops...

suggestions please before I start testing next week that I'll be onboard!

and please don't need anyone to tell me my system is too complicated and I shouldn't have it like that, cause it's really pointless and not what I'm asking...

cheers

V.

You didn't mention the AC system. All outlets and appliances should obviously have an Earth connection, this would typically be a busbar in the consumer unit, but elsewhere is fine, as long as they are all connected together. AC out on the Multiplus should also be connected to this busbar. Because you're fitting it on a boat you must Earth the case. Connect the case to the common busbar, then run a cable from the busbar to one of the anodes.

You must connect the common Earth to the water and it should only be a single connection.

 

[vas]

Paul thanks, was hoping you'd reply ,

I didn't mention much other than it was the stock el.panel. Got pics somewhere.
Yes all 230V sockets (schuko and 3pin 10A inline Italian style I think) are earthed on the panel.
AC from generator goes to Multi, AC out of multi goes straight to the el.panel and is also earthed there.

I'll check the layout again tomorrow if I have time and come back on that.

can you explain what that "single connection" mean?
iirc the anodes all link to the bat negative.
Should I isolate one of them and get that common earth from the el.panel to it?
Seems to me (have to check though!) that bat negative and 230VAC earth are "joined" ie. there's no separate AC earth going to water. Is that how they built things back then and should I upgrade/alter my layout?

cheers

V.

 

[PaulRainbow]

Paul thanks, was hoping you'd reply ,

I didn't mention much other than it was the stock el.panel. Got pics somewhere.
Yes all 230V sockets (schuko and 3pin 10A inline Italian style I think) are earthed on the panel.
AC from generator goes to Multi, AC out of multi goes straight to the el.panel and is also earthed there.

I'll check the layout again tomorrow if I have time and come back on that.

can you explain what that "single connection" mean?
iirc the anodes all link to the bat negative.
Should I isolate one of them and get that common earth from the el.panel to it?
Seems to me (have to check though!) that bat negative and 230VAC earth are "joined" ie. there's no separate AC earth going to water. Is that how they built things back then and should I upgrade/alter my layout?

cheers

V.

Single connection means one connection from AC Earth to the water, hence one of the anodes.

I don't think it will be an issue with both anodes connected together and separating them will be difficult (maybe impossible). Easy to separate each anode and have it connected to it's respective rudder, p-bracket etc, but you then have to connect the engine negatives, which joins it all back up again, unless the engine blocks are all isolated returns.

 

[kashurst]

I am with Paul on what he stated. The other thing to have a look at is the AC power run from the Multi unit' Does that cable(s) run close to the cables for the various tank sensors? If so try moving them apart as much as possible if the case grounding doesnt do the trick. Failing that, you may need to look at how the 12v negative (0v) connections are made. Ideally you want all the low power sensitive stuff (electronics) to have a star type connection back to the 12 V DC fuse /isolator panel, or possibly even a separate new cable just for the electronics back to the 12v Battery common negative point. Also check the state of teh connections from the battery negatve to the anode(s) too. Obviously try the easy stuff first.

 

[vas]

thanks K.,

I've checked that and I've made sure the NMEA2000 cable bringing 12VDC to my custom boxes runs completely separately to the duct that has the 230VAC stuff - actually runs 1.5m away and parallel to AC for 3m in the e/r.
Tank sensors and other low current kit have short clear (mostly!) runs to the custom boxes.
Topology isn't star as it's N2K so it's a backbone with lots of tees and drop cables. All that end on the main el.panel on a 2A fuse (iirc) Negative also goes to the panel, have to check if I can move it closer and far away from others - that's a good point (and easy to do!)

I'll do some multimeter Ohm testing for sure during the w/e after I earth the multi II case!

cheers

V.

 

[kashurst]

Just another thought have you checked the NMEA bus only has one 0v (negative) connection back to the 12VDC isolator panel?
You may inadvertenty have more than one which wouldn't stop it working but could create a loop for currents to circulate. Try disconnecting the known Ov connection at NMEA bus, where you know you connected it and check it is actually isolated from the battery 0v

 

[vas]

Just another thought have you checked the NMEA bus only has one 0v (negative) connection back to the 12VDC isolator panel?
You may inadvertenty have more than one which wouldn't stop it working but could create a loop for currents to circulate. Try disconnecting the known Ov connection at NMEA bus, where you know you connected it and check it is actually isolated from the battery 0v

good point!
I should only have one garmin yellow power tee on the system.
hm I have to check as at some point was thinking of adding a second one as backbone was close to 20m with all the going around... Not sure I fitted the second or not