LVX-6048 Ground Fault

[eabyrd]

Yeah, you’re

That is strange
Without AC in connected it starts.
What did MPP say about it ?

Let me guess (again lazy to read back ) ... asked for your wiring diagram ?
I remember there was an issue with the common neutral VS separated neutral. Where I said separated neutral is the good solution.
But in current case there is nothing connected on the AC out when the inverter do not start.
So there can not be a neutral backflow problem.
So ... inverter is faulty.

You're pretty much spot on about MPP, send us a diagram of your AC input side, take some readings across the input terminals and wait while we go dark. The seller has been good, and offered to trade me out, but I'm curious about this beast. Right now I'm asking them if the internal transfer switch also switches neutral, which it should if it's designed to form an internal neutral ground bond. The old Sigineer that I am coming from got around the neutral loop issue by only bringing in L1, L2 & G, and only outputting L1, L2, & N, but this beast chokes if you try that combination.

What's interesting to me is that the chassis to ground potential formed when the unit is ungrounded is a DC voltage of about 18 volts. I haven't historically grounded my battery bank, but I'm going to and see what that changes (if anything).

Right now I've got the unit wired without input or output ground, but I've made sure every component up to, and after the unit is tied back to the main panel with no loops. I think, I am OK (assuming I can stop licking my finger & touching the inverter chassis just to see if it still tickles) from a fault clearing perspective since every circuit has a single fault path back around a circuit breaker, an every metal other component also has an EGC path.

 

[mrzed001]

Yeah, you’re

You're pretty much spot on about MPP, send us a diagram of your AC input side, take some readings across the input terminals and wait while we go dark.

I think it is the logical way As you see I came to the same conclusion, wanted to do almost the same tests to get a picture about how it works currently. Finding errors is like an investigation.

The seller has been good, and offered to trade me out, but I'm curious about this beast. Right now I'm asking them if the internal transfer switch also switches neutral, which it should if it's designed to form an internal neutral ground bond. The old Sigineer that I am coming from got around the neutral loop issue by only bringing in L1, L2 & G, and only outputting L1, L2, & N, but this beast chokes if you try that combination.

Your problem is not the internal relay but that the unit's AC in part is broken or malfunctions.
With no AC out part it should start with or without the relay.
But only starts in battery mode with AC out connected and no AC in.

What's interesting to me is that the chassis to ground potential formed when the unit is ungrounded is a DC voltage of about 18 volts. I haven't historically grounded my battery bank, but I'm going to and see what that changes (if anything).

The batteries does not need to be grounded. But the unit does. That gives it the 0V potential.

Right now I've got the unit wired without input or output ground, but I've made sure every component up to, and after the unit is tied back to the main panel with no loops. I think, I am OK (assuming I can stop licking my finger & touching the inverter chassis just to see if it still tickles) from a fault clearing perspective since every circuit has a single fault path back around a circuit breaker, an every metal other component also has an EGC path.

Do not leave out the ground. Not a good idea
But with AC in connected inverter does not start ... so how do you use this unit?
No AC in, the PV MPPT still works and charges battery ... and AC out works too ?

 

[eabyrd]

I think it is the logical way As you see I came to the same conclusion, wanted to do almost the same tests to get a picture about how it works currently. Finding errors is like an investigation.



Your problem is not the internal relay but that the unit's AC in part is broken or malfunctions.
With no AC out part it should start with or without the relay.
But only starts in battery mode with AC out connected and no AC in.


The batteries does not need to be grounded. But the unit does. That gives it the 0V potential.


Do not leave out the ground. Not a good idea
But with AC in connected inverter does not start ... so how do you use this unit?
No AC in, the PV MPPT still works and charges battery ... and AC out works too ?

I Will need to test a fully connected (ie grounded) AC output with no AC input connection. Don’t recall trying that combination. NEC says ground batteries over 48 volts so I should go ahead and dot that i anyway

 

[mrzed001]

I Will need to test a fully connected (ie grounded) AC output with no AC input connection. Don’t recall trying that combination. NEC says ground batteries over 48 volts so I should go ahead and dot that i anyway

Your second test was this I think.
AC-IN = Ground Only
DC Battery = YES
DC PV = NO
AC-Out = Disconnected
Inverter ON.
And you get the Volts on the AC out
So the DC/AC inverter seems to work.

You can test if the MPPT is working by adding solar panels.

I think your AC in part is broken.
Can not even think of a scenario to test the AC charger only. Or the grid-tie production only.

 

[eabyrd]

Your second test was this I think.
AC-IN = Ground Only
DC Battery = YES
DC PV = NO
AC-Out = Disconnected
Inverter ON.
And you get the Volts on the AC out
So the DC/AC inverter seems to work.

You can test if the MPPT is working by adding solar panels.

I think your AC in part is broken.
Can not even think of a scenario to test the AC charger only. Or the grid-tie production only.

I did that yesterday as test #2 Yes I get 240 / 120 as anticipated L to L and L to N. I got cycling voltage from 0 to 110 on the N - G circuit
While conducting this test I confirmed the AC Charger worked as it kicked in to top up the battery bank while I was taking my measurements.
I've no interest in Grid-tie.

My real issue the bonded connection (which I was told it didn't form by their sales team) disqualifies it from use with a reliance controls manual transfer switch since that switch only switches the Lines and not the neutral as well. See my crude drawing attached. I would need a system that also switches the neutrals instead of relying on the neutral bar in the main box, or I need an inverter that has a common neutral and common ground like the Sol-Arks

 

[mrzed001]

I did that yesterday as test #2 Yes I get 240 / 120 as anticipated L to L and L to N. I got cycling voltage from 0 to 110 on the N - G circuit
While conducting this test I confirmed the AC Charger worked as it kicked in to top up the battery bank while I was taking my measurements.

So you started without AC in, then inverter started, AC out gets power, AC in connected again, AC in to battery charger worked (charged battery) ?

But when started with AC in then inverter failed in test #4 and #6.
Strange but it still seems to be an AC in problem.

I've no interest in Grid-tie.

My real issue the bonded connection (which I was told it didn't form by their sales team) disqualifies it from use with a reliance controls manual transfer switch since that switch only switches the Lines and not the neutral as well. See my crude drawing attached. I would need a system that also switches the neutrals instead of relying on the neutral bar in the main box, or I need an inverter that has a common neutral and common ground like the Sol-Arks

I use L+N transfer switches only. As you checked there is no continuity between incoming N to outgoing N.
It is not like a (not so sophisticated) Victron where line, line, neutral ... come in, get out ... relay attach or detach inverter to it and ready.
So you need to separate the unprotected load from the protected load. Not only on line side but neutral side too.
The separate neutral version I talked about earlier.

 

[eabyrd]

It is not like a (not so sophisticated) Victron where line, line, neutral ... come in, get out ... relay attach or detach inverter to it and ready.

Huh? If the inverter will always have attachment to main panel N and G it has no need to form it’s own bond

 

[eabyrd]

you need to separate the unprotected load from the protected load. Not only on line side but neutral side too.
The separate neutral version I talked about earlier

So it appears, or a Sol-ark that treats neutral and ground as common to the system and doesn’t form an internal bond at all (which was the architecture i thought I was buying)

 

[mrzed001]

So it appears, or a Sol-ark that treats neutral and ground as common to the system and doesn’t form an internal bond at all (which was the architecture i thought I was buying)

View attachment 62382

I am not exactly sure of anything but from the sol-ark inner diagram it seems to work like the Victron ... so has direct connection between N in and N out.

The Sol-Ark is a very good professional unit.
Its price is ... professional too. (except theirs webpage that strikes for a while)

It seems MPP does not direct connects incoming N to outgoing N.
And this is a very good thing, called galvanic separation.
If an overvoltage comes from the AC in side (lightning, LEMP, ...) it can protect your equipments in the AC out side.
Interesting.
I was wondering why the MPP/Voltronic is the only allowed one unit hybrid in my country.
Even the AC coupled and outer separator inverters are in a different "we only allow them with the following ..." part.

 

[eabyrd]

So I disconnected AC In (again) and connected AC out (L1, L2, N, G) and it starts to cycle as soon as you turn it on just like with AC In. It doesn't like the ground loop created between the main panel (fed through the reliant transfer switch) and itself.

So given that I cannot earth the inverter given that every path to earth I have is bonded to my neutral in my main panel. if I choose to run it in its current state for emergency power then I see two choices
1. Run dedicated circuits to an isolated subpanel (L, N, and G) and power it with the inverter (in this event none of the circuits or components will have be earthed, but everything will be bonded)

2. Run circuits from a properly configured subpanel (N and G tied to their counterparts in the main), but don't run ground from the inverter to the sub-panel (in this scenario all components except the inverter are properly bonded and earthed, but the inverter is neither bonded or earthed.

Question is which is the less hazardous configuration?

 

[mrzed001]

So I disconnected AC In (again) and connected AC out (L1, L2, N, G) and it starts to cycle as soon as you turn it on just like with AC In. It doesn't like the ground loop created between the main panel (fed through the reliant transfer switch) and itself.

Wait, the AC in N is not connected. The ground is not a problem to go around. Only the N.

So given that I cannot earth the inverter given that every path to earth I have is bonded to my neutral in my main panel.

And that is perfect this way. Ground it, and have the bond on the main. No reason to do any other way.

if I choose to run it in its current state for emergency power then I see two choices
1. Run dedicated circuits to an isolated subpanel (L, N, and G) and power it with the inverter (in this event none of the circuits or components will have be earthed, but everything will be bonded)

2. Run circuits from a properly configured subpanel (N and G tied to their counterparts in the main), but don't run ground from the inverter to the sub-panel (in this scenario all components except the inverter are properly bonded and earthed, but the inverter is neither bonded or earthed.

Question is which is the less hazardous configuration?

I think the good approach is the the simplest approach.
I copied this stupid pic from my previous post (maybe I did not explain it then):

The green circuit is the unprotected load circuit (L1,L2,N is on your inverters AC in side) (goes into the house just draw it there).
That come and goes (L1,L2,N) from your main subpanel

The purple is your protected load circuit (L1,L2,N is on your inverter AC out side ... totally separated from main panel counterparts).
That come and goes from your protected load subpanel.

No continuity between main N and protected N. The earth bonds only the unprotected N on your main panel in bypass mode.
In battery mode (grid is down) inverter connects inside the outgoing G-N (and to the incoming G)

Ground is common, BUT 1-1 RCD needs to be in the "bottleneck" of this 2 circuits. So where green starts from main. And for purple directly after the inverter AC out.

You see if there is a single wrongly connected protected N on the house that should be only going to the protected N (but comes back on the unprotected N) ... then that goes around and comes back on the inverter AC in side. (starts as purple but end as green.)

So I think there is 2 possible error:
1, your inverter is broken. This is my main opinion about this not starting and cycling without throwing an error message
2, Not separated N, N loop (but it should not make such a mess)

 

[eabyrd]

Wait, the AC in N is not connected. The ground is not a problem to go around. Only the N.

AC in was completely disconnected, but ground is the issue. The AC out isn’t feeding to a dedicated subpanel

 

[mrzed001]

AC in was completely disconnected, but ground is the issue. The AC out isn’t feeding to a dedicated subpanel

AC in ground was connected? Should be at all time.

 

[eabyrd]

By way of update MPP solar last week requested video of the unit doing it's on & off cycle when connected to AC Input only, so I sent them such the same morning. I've not heard back since

No matter what they do at this point I'm probably going to swap it out for an LV-6548 since that has the UL listing, and I only run 120 volt loads anyway. Bummer to have to run two strings of 8 panels, but I can setup 2 4-panel runs with each having it's own MPPT so there is an advantage there

 

[mrzed001]

By way of update MPP solar last week requested video of the unit doing it's on & off cycle when connected to AC Input only, so I sent them such the same morning. I've not heard back since

No matter what they do at this point I'm probably going to swap it out for an LV-6548 since that has the UL listing, and I only run 120 volt loads anyway. Bummer to have to run two strings of 8 panels, but I can setup 2 4-panel runs with each having it's own MPPT so there is an advantage there

I think they will say that the unit is broken. I am eager to wait for the response.

You know if you change to the LV-6548 then that is an off-grid inverter, and not a hybrid (can not produce to grid).
And for that inverter you even more need the neutral separation

 

[eabyrd]

I think they will say that the unit is broken. I am eager to wait for the response.

You know if you change to the LV-6548 then that is an off-grid inverter, and not a hybrid (can not produce to grid).
And for that inverter you even more need the neutral separation

Producing to the grid is nothing to me. Never was. i am trying to get a handle on neutral routing now. Do you have docs from MPP?

 

[mrzed001]

Producing to the grid is nothing to me. Never was. i am trying to get a handle on neutral routing now. Do you have docs from MPP?

The other thing that Hybrid does that your non protected load gets power also from it (if not blackout). So it lowers your power bill even there.

The LV-6548 user manual you mean?
http://www.mppsolar.com/manual/SPLIT PHASE LV/LV6548 split phase manual-20210416.pdf

 

[eabyrd]

No benefit if you have export turned Off which I must in PA. I have the manual, but thanks anyway. It is as unhelpful as the LVX manual

 

[mrzed001]

No benefit if you have export turned Off which I must in PA. I have the manual, but thanks anyway. It is as unhelpful as the LVX manual

I think it has a limiter ... so you can export as much as you use in the non protected side.