Victron Orion XS DC-DC charger dedicated negative wire to starter battery vs chassis return (and possible ground loop issues)

[bennetimo]

I'm upgrading a Victron Orion 12/12-30 (isolated) charger to the new Orion XS 50A (non-isolated) in a campervan. My house system in the van centers around a Victron Lynx distributor as the main hub, and that is grounded to the chassis. As such I was using the isolated charger in a non-isolated way.

So for the original setup I had positive and negative wires from the starter battery to the input of the orion, and positive and negative out to the lynx and ultimately to the batteries. This has worked great with no issues at all but I am now upgrading to the much better XS. There is no isolated model available to do a direct swapout so I'm wiring it very similar to how explorist.life does it here: https://shop.explorist.life/shop/all-products/victron-orion-xs-12-12-50-dc-to-dc-charger-retrofit-wiring-kit/ . i.e. using the original cables, but with the two negatives combined at a junction stud and then a third new cable added to connect to the single GND terminal on the Orion. All works and I'm getting close to the full 50A rated output from the new XS. Happy days!

However despite having dedicated wires for both the positive and negative cable from starter battery to the orion, there is very little current using the dedicated return negative path. I have an amp clamp and measure only ~2A passing back along the negative wire run (16mm² over perhaps 5-6m), with the remaining ~48A instead using the chassis return path from the lynx. Therefore as it stands the dedicated negative wire to the starter battery is doing virtually nothing. The chassis ground from the lynx is connected via a large 70mm² cable to a bolted point on the metal stripped of paint so is a decent ground, but I was surprised how little current is using the dedicated wire. Most things I've read (and in sterling/renogy manuals) suggest the chassis is often a poor return path and recommend a direct path. I have double checked the ring terminal crimps etc on the negative wire and can't see any issues.

Am I missing something here, or when wiring a non-isolated DC charger would you always expect the majority of current to flow through the chassis despite having dedicated wires, making them somewhat redundant? The chassis is steel, much less conductive than copper, but with a much bigger mass so... . With the old setup using the isolated charger, there was never any normal current flow in the ground wire from the lynx. I would prefer to keep it that way.

Also the excellent 'Grounding Made Simpler' by Filter Guy (thanks, great resource) says of non-isolated DC-DC chargers:

Since the tie between House Battery DC Negative and ground should only be done in one place, the DC-DC charger becomes the bonding point to chassis and the Jumper between the DC negative Bus and the ground bus is removed. Otherwise, a ground loop would be created and possibly create RFI issues.

Since switching to the non-isolated charger presumably there is also now a ground loop, as current can flow back both through the chassis as well as via the dedicated negative wire. Are there issues with this setup other than potential RFI noise? Lots of wiring diagrams I've seen share a similar setup to mine (dedicated negative wire between batteries and orion, as well as a chassis ground), e.g. https://shop.explorist.life/wp-cont...fe-EXPLORE-50A-MOTORHOME-WIRING-DIAGRAM-1.png , so presumably would all exhibit a similar thing.

Would welcome any thoughts!

 

[mikefitz]

Am I missing something here, or when wiring a non-isolated DC charger would you always expect the majority of current to flow through the chassis

Yes, it's a low impedance path, provided the connections to the metal are correctly made.
Many recient vehicles have 'smart altetnator' charging where its critical to use chassis/vehicle metal as the negative connection to the vehicle charging system when Installing DC to DC chargers.
Ground loops are an issue with vehicle audio and radio systems, a doubt if a DC to DC charger will be compromised. If this is a worry, delete the cable to the starter and rely on the cable to the house battery and its lynx connection to chassis for the negative return.

 

[bennetimo]

Thanks for the response, it makes sense if the connections to chassis are made well like you say. But then I'm wondering why lots of other diagrams/installs out there seem to also add the dedicated wire then, if it's not really adding anything?

Many recient vehicles have 'smart altetnator' charging where its critical to use chassis/vehicle metal as the negative connection to the vehicle charging system when Installing DC to DC chargers.

I have a 2017 Ford Transit which does have a smart alternator but it has an auxiliary ground stud on the negative battery post which according to the BEMM can be used, and I haven't had any issues running the old 30A orion with it for the last few years so I think I'm ok on that front.

Ground loops are an issue with vehicle audio and radio systems, a doubt if a DC to DC charger will be compromised. If this is a worry, delete the cable to the starter and rely on the cable to the house battery and its lynx connection to chassis for the negative return.

I was doing some more playing around with it and noticed something interesting... Because of the now extra ground loop, when the key is turned in the ignition there are a couple of amps that flow through the ground wire to the lynx and back via the dedicated negative wire run. And when starting the engine, there's a peak of about 40A from the starter motor current that flows back in that same loop. So presumably now some of the current is taking a long winded route back from the starter motor, via the chassis to the back of the van, up the ground wire to the lynx, and back to the negative starter battery via the dedicated wire. It seems wrong to me to have current from the van side using the leisure side's wiring, but I haven't seen this mentioned or talked about anywhere, only in relation to a bit of RFI noise.

I can remove one of the wires like you say to avoid the loop, but I'm curious how others are wiring these. The explorist.life diagram I linked to in my previous post presumably has the additional dedicated wire with the assumption that it's going to carry most of the current back to the battery (and avoiding using the chassis), otherwise it seems a bit redundant. But from what I'm seeing it won't carry much current and instead will just add a ground loop that makes it harder to reason about the system.

 

[mikefitz]

there's a peak of about 40A from the starter motor current that flows back in that same loop.

For this to happen the path from battery negative to van metal is suspect . This path has a current sensor. By making a connection to the battery post direct you are offering an alternative path.

 

[bennetimo]

For this to happen the path from battery negative to van metal is suspect .

Hmm what do you mean by suspect? The current must be flowing through the chassis, up the ground wire to the lynx, then via the dedicated wire run straight back to the negative post of the starter battery. I can measure the current with an amp clamp on those wires and see it.

If I remove the dedicated wire back from the orion to the starter battery negative and just rely on the ground to chassis via the lynx, then the ground loop is removed and the starter motor and other van currents don't flow through the leisure system wiring as expected. But renogy/sterling manuals etc recommend adding a direct link so presumably a lot of installs out there would see this kind of thing happening

 

[outboard1196]

Hmm what do you mean by suspect? The current must be flowing through the chassis, up the ground wire to the lynx, then via the dedicated wire run straight back to the negative post of the starter battery. I can measure the current with an amp clamp on those wires and see it.

If I remove the dedicated wire back from the orion to the starter battery negative and just rely on the ground to chassis via the lynx, then the ground loop is removed and the starter motor and other van currents don't flow through the leisure system wiring as expected. But renogy/sterling manuals etc recommend adding a direct link so presumably a lot of installs out there would see this kind of thing happening

Maybe an insufficient ground connection from battery to chassis.

 

[bennetimo]

Maybe an insufficient ground connection from battery to chassis.

I checked out all the connections today and didn't find any issues. The starter motor was recently replaced so I did wonder if the ground cable hadn't been torqued down properly but it was fine, and the connections to the starter battery were fine as well. I did disconnect them and clean them with some IPA before bolting them back down but I don't think it changed anything.

Playing around with it some more though it feels like there isn't a bad ground issue, it's just that the 'new' ground from the wire from the back of the van is another decent current path so it's just dividing proportionally across them both. The starter has a ground wire attached that bolts onto the chassis under the van onto one of the members under the radiator. And the starter batteries are under the driver seat, so the negative wire runs from there to presumably be bolted to the chassis under or near the battery box (that's the one connection I haven't been able to trace completely as it requires physically removing both batteries to get to).

The negative cable on the starter battery looks like it's either 35mm² or 50mm², and the wire I've run is 16mm². With the chassis being a great conductor it's acting like a big bus bar and now the current from the starter motor has two parallel paths back, one just being longer going to the back of the van first then back through the 16mm². So it's not as optimum as the short path along the chassis from the starter motor but good enough to carry a portion of the current, the ~40A I'm seeing out of the full ~400A of starter current that I measure with an amp clamp. So about 10%.

The 16mm² I've run can carry that 40A of current no problem (rated for 110A), but it still feels a bit wrong to have sizeable current from the van itself flowing through a house wire that wasn't intended/specced for it.

I can disconnect the dedicated 16mm² wire and just use the van chassis for 100% of the Orion charger current and that would stop the starter current passing through. Still just curious though as a lot of diagrams and manuals I've seen seem to advocate for running an additional direct wire, but then presumably those setups would have this same quirk...?

 

[outboard1196]

I checked out all the connections today and didn't find any issues. The starter motor was recently replaced so I did wonder if the ground cable hadn't been torqued down properly but it was fine, and the connections to the starter battery were fine as well. I did disconnect them and clean them with some IPA before bolting them back down but I don't think it changed anything.

Playing around with it some more though it feels like there isn't a bad ground issue, it's just that the 'new' ground from the wire from the back of the van is another decent current path so it's just dividing proportionally across them both. The starter has a ground wire attached that bolts onto the chassis under the van onto one of the members under the radiator. And the starter batteries are under the driver seat, so the negative wire runs from there to presumably be bolted to the chassis under or near the battery box (that's the one connection I haven't been able to trace completely as it requires physically removing both batteries to get to).

The negative cable on the starter battery looks like it's either 35mm² or 50mm², and the wire I've run is 16mm². With the chassis being a great conductor it's acting like a big bus bar and now the current from the starter motor has two parallel paths back, one just being longer going to the back of the van first then back through the 16mm². So it's not as optimum as the short path along the chassis from the starter motor but good enough to carry a portion of the current, the ~40A I'm seeing out of the full ~400A of starter current that I measure with an amp clamp. So about 10%.

The 16mm² I've run can carry that 40A of current no problem (rated for 110A), but it still feels a bit wrong to have sizeable current from the van itself flowing through a house wire that wasn't intended/specced for it.

I can disconnect the dedicated 16mm² wire and just use the van chassis for 100% of the Orion charger current and that would stop the starter current passing through. Still just curious though as a lot of diagrams and manuals I've seen seem to advocate for running an additional direct wire, but then presumably those setups would have this same quirk...?

Probably nothing to worry about if there are no issues , power will flow any path it finds. Your meter should show positive and negative current, figure out which way its flowing if its negative (flowing away from the starter battery ) It's using the new conductor as another path to ground.
Me personally, I would connect the orion / battery bank to chassis as close to the Orion as possible.

 

[bennetimo]

Probably nothing to worry about if there are no issues , power will flow any path it finds. Your meter should show positive and negative current, figure out which way its flowing if its negative (flowing away from the starter battery ) It's using the new conductor as another path to ground.
Me personally, I would connect the orion / battery bank to chassis as close to the Orion as possible.

Thanks, yeah the new conductor is definitely being used as a path to ground for the starter motor. The house/leisure battery system is all in the back of the van, and the Orion is very close to the lynx which is then grounded with a short 70mm² to chassis. The van starter batteries are under the driver seat (Ford Transit) and presumably they're grounded to the chassis very close by but haven't traced that wire.

Would you personally still run the additional negative wire between starter and leisure batteries, assuming it was a non-isolated setup where both are grounded to the chassis already?

 

[outboard1196]

Thanks, yeah the new conductor is definitely being used as a path to ground for the starter motor. The house/leisure battery system is all in the back of the van, and the Orion is very close to the lynx which is then grounded with a short 70mm² to chassis. The van starter batteries are under the driver seat (Ford Transit) and presumably they're grounded to the chassis very close by but haven't traced that wire.

Would you personally still run the additional negative wire between starter and leisure batteries, assuming it was a non-isolated setup where both are grounded to the chassis already?

I'm no expert but I don't see a problem either way ( no wire simpler and cheaper, but you're already there). I recently looked at the orion xs for myself and, assumed the negative on the orion is only for reference and does not carry any significant current ? I could be wrong

 

[bennetimo]

I'm no expert but I don't see a problem either way ( no wire simpler and cheaper, but you're already there). I recently looked at the orion xs for myself and, assumed the negative on the orion is only for reference and does not carry any significant current ?

The old orion 30A that I replaced was an isolated one, so that had positive and negative from each battery straight into it, and was effectively two separate circuits. So there'd be ~34A flowing on the positive and negative on the input side and ~30A on positive and negative on the output side (more on input because of the only 87% efficiency of the old model).

With the non-isolated one there's 50A exactly on the input positive from starter battery to the orion (the new one has a hard input limit), and a max of exactly 50A on the output positive as well. I believe the short negative wire from the GND on the orion will only carry the difference in current between the input and output, so quite small. But the full 50A of current has to flow back to the starter battery somehow, at the moment for me that's pretty much all going through the chassis. I could remove the connection from lynx to chassis and then all the current would be forced to travel back through my wire, and the house system would still share a common ground via the orion negative wiring. But I'd rather not do that as the existing ground wire from the lynx is a nice beefy purpose made connection.

I prefer the isolated way as I find it easier to reason about, I provide the wire and that's what's used, as opposed to it going via chassis and taking an indeterminate path back. I can visualise it as two separate circuits that meet at the orion from each side, whereas the non-isolated version is more a single circuit that flows from starter battery to orion to leisure batteries and back so intermingles the two.

 

[outboard1196]

The old orion 30A that I replaced was an isolated one, so that had positive and negative from each battery straight into it, and was effectively two separate circuits. So there'd be ~34A flowing on the positive and negative on the input side and ~30A on positive and negative on the output side (more on input because of the only 87% efficiency of the old model).

With the non-isolated one there's 50A exactly on the input positive from starter battery to the orion (the new one has a hard input limit), and a max of exactly 50A on the output positive as well. I believe the short negative wire from the GND on the orion will only carry the difference in current between the input and output, so quite small. But the full 50A of current has to flow back to the starter battery somehow, at the moment for me that's pretty much all going through the chassis. I could remove the connection from lynx to chassis and then all the current would be forced to travel back through my wire, and the house system would still share a common ground via the orion negative wiring. But I'd rather not do that as the existing ground wire from the lynx is a nice beefy purpose made connection.

I prefer the isolated way as I find it easier to reason about, I provide the wire and that's what's used, as opposed to it going via chassis and taking an indeterminate path back. I can visualise it as two separate circuits that meet at the orion from each side, whereas the non-isolated version is more a single circuit that flows from starter battery to orion to leisure batteries and back so intermingles the two.

Unless I missed something there shouldn't be any current flowing towards your starter battery ( I'm assuming your charging the house battery by alternator) you can leave it as is with the ground wire going to the starter battery and lynx grounded to chassis if you like. I would probably go with just the lynx to chassis and remove the ground wire to starter battery for simplicity.

 

[bennetimo]

Unless I missed something there shouldn't be any current flowing towards your starter battery ( I'm assuming your charging the house battery by alternator) you can leave it as is with the ground wire going to the starter battery and lynx grounded to chassis if you like. I would probably go with just the lynx to chassis and remove the ground wire to starter battery for simplicity.

Hmm there's 50A that flows out from the starter battery to the orion, it has to also go back otherwise there isn't a complete circuit. The alternator chargers the starter batteries and the orion takes power from the starter (so indirectly from the alternator) to charge the leisure batteries in the back.

Yeah removing the additional wire to the starter is an option I'm toying with. It would remove the dual paths for the starter battery which is good. But the reason I have it in the first place is because I didn't want any current flowing through the chassis (with the isolated charger on the same wiring that was the setup). With the isolated one as the negatives are galvanically separated the current had no choice but to flow back through my wire and not the chassis. With the new unit though the negatives are tied so it is free to take the most optimum path, which just happens to be ~95% the chassis and only ~5% my wire

 

[outboard1196]

Hmm there's 50A that flows out from the starter battery to the orion, it has to also go back otherwise there isn't a complete circuit. The alternator chargers the starter batteries and the orion takes power from the starter (so indirectly from the alternator) to charge the leisure batteries in the back.

Yeah removing the additional wire to the starter is an option I'm toying with. It would remove the dual paths for the starter battery which is good. But the reason I have it in the first place is because I didn't want any current flowing through the chassis (with the isolated charger on the same wiring that was the setup). With the isolated one as the negatives are galvanically separated the current had no choice but to flow back through my wire and not the chassis. With the new unit though the negatives are tied so it is free to take the most optimum path, which just happens to be ~95% the chassis and only ~5% my wire

Was your lynx grounded to the chassis before?
If it was then the system weren't isolated like you through, it's essentially the same as the Orion XS. I get what your saying about completing a circut but power is only flowing to ( charging) your house battery not the other way around. Cars are designed to have the chassis as a common ground point all system ground to the chassis at the closest easiest point they don't make home runs on the ground side since the chassis is a good low resistance conductor. Don't over think it I don't like the Orion XS not being isolated either but it's what we have

 

[outboard1196]

One other issue I see is is you leave the ground cable going from the starter battery to the orion.
if your factory ( Ford) battery to chassis ground fails or becomes high resistance it will start using the other cable and might over load the current capacity of it. Poor ground cause hell on automotive systems, wich might make diagnosis harder.

 

[bennetimo]

Was your lynx grounded to the chassis before?
If it was then the system weren't isolated like you through, it's essentially the same as the Orion XS

Yeah it's always been grounded to the chassis so I was using the old one in a non-isolated way. But I think wiring the XS combining the negatives is very similar with a nuanced difference. The isolated one transfers power using a transformer so the two sides are completely separate, so even though both sides were grounded to chassis and referenced the same ground, they're like two separate circuits. At least, with the old unit you could measure with an amp clamp and visually see the current on both the positive and negative leg of the input being ~34A, and on both the positive and negative output leg being ~30A. And no current ever flowed through the chassis via the lynx-chassis connection even with it connected, as there was no direct path back using it.

I get what your saying about completing a circut but power is only flowing to ( charging) your house battery not the other way around. Cars are designed to have the chassis as a common ground point all system ground to the chassis at the closest easiest point they don't make home runs on the ground side since the chassis is a good low resistance conductor.

Not sure I'm following you on that one, as for current to flow there has to be a closed loop. If 50A leaves the starter battery to the orion on the positive, it must ultimately return back to the starter battery negative. It just depends how it gets there, it either goes via the chassis or it goes via the dedicated wire I've added. I removed the ground wire to the lynx as a test, and doing that you can see the full 50A flows on both the positive wire to the orion and my negative wire back. If I instead keep the chassis-lynx connection and disconnect my wire, then I have the full 50A flowing through the chassis back. And if I keep both, then I have ~48A using the chassis and ~2A using my wire, presumably just because of the physics and the relative resistances of the two paths back.

One other issue I see is is you leave the ground cable going from the starter battery to the orion.
if your factory ( Ford) battery to chassis ground fails or becomes high resistance it will start using the other cable and might over load the current capacity of it. Poor ground cause hell on automotive systems, wich might make diagnosis harder.

Yeah that was sort of the basis of my concern as well in my original post, as ~40A of the starter battery current already flows through the lynx-chassis connection. So if for whatever reason there was an issue with the starter battery ground cable, then a lot more of that current could instead flow through my wire which is obviously not spec'd to carry starter motor current. And if the engine still started fine, you probably wouldn't even notice. That's why I'm still pondering it, as it seems like adding that wire introduces that potential problem but there are so many references, diagrams, manuals out there that suggest adding the additional wire. Presumably because of it being 'more efficient', 'better connection' etc, though from what I've measured it doesn't really end up carrying any of the current anyway.

 

[turbodb]

Yeah it's always been grounded to the chassis so I was using the old one in a non-isolated way. But I think wiring the XS combining the negatives is very similar with a nuanced difference. The isolated one transfers power using a transformer so the two sides are completely separate, so even though both sides were grounded to chassis and referenced the same ground, they're like two separate circuits. At least, with the old unit you could measure with an amp clamp and visually see the current on both the positive and negative leg of the input being ~34A, and on both the positive and negative output leg being ~30A. And no current ever flowed through the chassis via the lynx-chassis connection even with it connected, as there was no direct path back using it.

Not sure I'm following you on that one, as for current to flow there has to be a closed loop. If 50A leaves the starter battery to the orion on the positive, it must ultimately return back to the starter battery negative. It just depends how it gets there, it either goes via the chassis or it goes via the dedicated wire I've added. I removed the ground wire to the lynx as a test, and doing that you can see the full 50A flows on both the positive wire to the orion and my negative wire back. If I instead keep the chassis-lynx connection and disconnect my wire, then I have the full 50A flowing through the chassis back. And if I keep both, then I have ~48A using the chassis and ~2A using my wire, presumably just because of the physics and the relative resistances of the two paths back.


Yeah that was sort of the basis of my concern as well in my original post, as ~40A of the starter battery current already flows through the lynx-chassis connection. So if for whatever reason there was an issue with the starter battery ground cable, then a lot more of that current could instead flow through my wire which is obviously not spec'd to carry starter motor current. And if the engine still started fine, you probably wouldn't even notice. That's why I'm still pondering it, as it seems like adding that wire introduces that potential problem but there are so many references, diagrams, manuals out there that suggest adding the additional wire. Presumably because of it being 'more efficient', 'better connection' etc, though from what I've measured it doesn't really end up carrying any of the current anyway.

Curious where you ended up with this?

I'm about to install an entirely new house electrical system centered around a Lynx Distributor and using the Orion XS for charging the house battery. This is new to me, as previously my house battery was AGM, and was linked (or unlinked depending on conditions) to the starting battery directly.

I'm considering two options:

1. Run a chassis ground from the Lynx Distributor; connect the "GND" terminal on the Orion XS to the negative bus of the Lynx Distributor.
2. Run a chassis ground from the Lynx Distributor; connect the "GND" terminal on the Orion XS to the negative terminal of the starting battery.

Sounds like you currently have (or had) a different setup:

• Run a chassis ground from the Lynx Distributor; connect the "GND" terminal on the Orion XS to both the negative bus of the Lynx Distributor and negative terminal of the starting battery.

It also sounds like you've considered removing the GND wire from the Orion XS to the negative terminal of the starting battery (chassis ground), but haven't considered removing the wired GND wire between the Orion XS and the Lynx Distributor. Is there a reason for that? (It seems to me that doing so might take the house system out of the equation from a starter/van perspective.

Here's a diagram of what I'm planning, which I suspect is similar to your system: