diy solar

diy solar

Off-grid ground (earthing)

Callinger

New Member
Joined
Oct 2, 2021
Messages
34
I have spent hours researching this subject and I think I am no closer to answers. One post says: "ground everything metal to earth", another says: "keep everything floating". I just want to keep my equipment and people safe and healthy. I appreciate how complex the topic is:
- off-grid grounding (earthing)
- grid-tie grounding
- mobile installation grounding
- ground-neutral bonding
- A/C side earthing
- PV grounding, bonding, earthing
- lightning mitigation
Many posts and vids give much wisdom and advice but use industry terminology that is over the head of us DIY'ers such as "The only connection made between the ground and the neutral conductor is made on the service side at the meter base."
"English, Spock!"
Perhaps if I can ask one simple question at a time, someone can help us with a few key guidelines / advice.

1. we have an off-grid shipping container based system with 2 LV6548s in parallel to create 240v split phase into to a typical Square-D main panel. to resolve the ground-neutral bond issue should we: a. remove the bar in the main panel between the ground and neutral busses? or b. remove the bonding screw in both LV6548 controllers? c. both d. neither
 
I have spent hours researching this subject and I think I am no closer to answers. One post says: "ground everything metal to earth", another says: "keep everything floating". I just want to keep my equipment and people safe and healthy. I appreciate how complex the topic is:
- off-grid grounding (earthing)
- grid-tie grounding
- mobile installation grounding
- ground-neutral bonding
- A/C side earthing
- PV grounding, bonding, earthing
- lightning mitigation
Many posts and vids give much wisdom and advice but use industry terminology that is over the head of us DIY'ers such as "The only connection made between the ground and the neutral conductor is made on the service side at the meter base."
"English, Spock!"
Perhaps if I can ask one simple question at a time, someone can help us with a few key guidelines / advice.

1. we have an off-grid shipping container based system with 2 LV6548s in parallel to create 240v split phase into to a typical Square-D main panel. to resolve the ground-neutral bond issue should we: a. remove the bar in the main panel between the ground and neutral busses? or b. remove the bonding screw in both LV6548 controllers? c. both d. neither
If you have no grid connection, you have a couple of options:

1654570725772.png
OR

1654570895531.png

If you have the grid feeding the AC input, it would need to be like this:

1654571164161.png

The unconnected grounds at the critical load box are shown to meet the NEC requirement of running the Equipment Grounding Conductor with the Ungrounded (Hot) conductor. However, if they are connected at the critical load box they would form ground loops that could generate significant RFI issues. However, if the inverters are mounted on the metal walls of the shipping container, the loops become unavoidable.

As you point out, there are people that disagree with this, but from my studies and investigations, this is what I believe is the best approach.

The team at EG4 is doing a review of the whole grounding and bonding question with these inverters..... hopefully they will get back to us soon.
 
This is extremely helpful. THANKYOU!
I assume I need to make sure there is no neutral-ground bond inside the generator?
 
I assume I need to make sure there is no neutral-ground bond inside the generator?
Please review this resource. Particularly pages 12 - 14.

There are many ways to hook things up and have it work correctly. The resource shows a few 240V split phase set-ups with a generator.
 
I opened up one of the LV6548s to remove the N+G bonding screw. Could not find it. Is it allowable to follow this suggestion I found in another thread?:

Simply run grounds to the input side to each inverter so each is properly earthed. Run a single ground from the AC output of 1 (and only 1) of your split-phase units to the panel they service. Just 1. Do not bond at the panel

Simple complete grounding, with no loops and no screws removed
 

Attachments

  • 1654719768829.png
    1654719768829.png
    442.4 KB · Views: 11
I opened up one of the LV6548s to remove the N+G bonding screw. Could not find it. Is it allowable to follow this suggestion I found in another thread?:
Unfortunately, the screw is under another board. Look at the document I linked to. Toward the end of the document are pictures showing where to find it.

Simply run grounds to the input side to each inverter so each is properly earthed. Run a single ground from the AC output of 1 (and only 1) of your split-phase units to the panel they service. Just 1. Do not bond at the panel

Simple complete grounding, with no loops and no screws removed
That sounds ok for grounding (No ground loops). However, it will end up with two NG-Bonds when on battery.
 
ok. so this upper circuit board must be removed? or can you get to the screw without removing it? I can't work on this till next week.
 
We will make connections as per this diagram:
1654812643975.png


Question: The cases of the 2 controllers are "earth-grounded" (as indicated in the diagram). Is the case of the main distribution panel (breaker-box) also earth-grounded to the same wire? (yes, the N-G bond link has been removed in the breaker box)
 
We will make connections as per this diagram:
View attachment 97950


Question: The cases of the 2 controllers are "earth-grounded" (as indicated in the diagram). Is the case of the main distribution panel (breaker-box) also earth-grounded to the same wire? (yes, the N-G bond link has been removed in the breaker box)
Yes
 
I updated the document to show the breaker box and xfer switch cases are grounded (assuming they are metal)
 
I've been reading this thread and many of the others on the same topic and it's driving me a little nuts (more than usual). I have two MPP LV6548 inverters and four (two more ordered to make it six) EG4 5KWH batteries. Everything is connected and configured exactly as in Will's videos, right down to the same 150A breakers. My breaker panel is NG bonded. I'd rather leave it bonded at the panel. This system is totally off grid with no plans to connect ever. I'd like to have 240V capability, but I can't determine which opinion is correct.

1. Remove a screw from one panel?
2. Remove screws from both panels?
3. Leave the screws in?

I've read opinions that say for off grid it doesn't matter and other opinions that say I have to remove one screw ONLY, and at least one stating to remove both screws. Another opinion stated that if you're not doing split phase 240V it doesn't matter. Currently I'm just running a few 120V circuits.

I would greatly appreciate some guidance.

I really appreciate this forum and all the great advice and info found here!

Thanks
 
I've been reading this thread and many of the others on the same topic and it's driving me a little nuts (more than usual). I have two MPP LV6548 inverters and four (two more ordered to make it six) EG4 5KWH batteries. Everything is connected and configured exactly as in Will's videos, right down to the same 150A breakers. My breaker panel is NG bonded. I'd rather leave it bonded at the panel. This system is totally off grid with no plans to connect ever. I'd like to have 240V capability, but I can't determine which opinion is correct.

1. Remove a screw from one panel?
2. Remove screws from both panels?
3. Leave the screws in?

I've read opinions that say for off grid it doesn't matter and other opinions that say I have to remove one screw ONLY, and at least one stating to remove both screws. Another opinion stated that if you're not doing split phase 240V it doesn't matter. Currently I'm just running a few 120V circuits.

I would greatly appreciate some guidance.

I really appreciate this forum and all the great advice and info found here!

Thanks
As you have observed there is no consensus on this.

I go with this guiding principle: At any given time there should one and only one Neutral-Ground bond in the circuit.

With multiple LV6548s, you must remove bonding screws in order to achieve this rule.

My breaker panel is NG bonded. I'd rather leave it bonded at the panel. This system is totally off grid with no plans to connect ever.
I assume the breaker panel is on the output of the LV6548s... Given that there will already be an N-G bond on the output of all the inverters, in order to have only one N-G bond, the bonding screws must be removed from all the inverters.

BTW: If you ever decide to hook up a generator, It should not have an N-G bond either.

Counter arguments:
The primary counterargument that I have heard is that removing the screws voids the warranty.....but I have heard mixed messages as to whether the warranty is voided or not. If it is true that it voids the warranty, that leaves you with a no-win choice.

Others have argued that it is OK to have multiple N-G bonds, but frankly, I have never understood their reasoning on why it is OK to have multiple N-G bonds.

The system works with multiple N-G bonds so what is the problem?
There are two problems with multiple N-G bonds
1) Safety. With Multiple N-G bonds, current *will* flow on the ground wire between the two N-G bonds. This is not a direct functional problem, but it can be a safety issue....particularly when the system is being serviced.

2) Radio Frequency Interference. Multiple N-G bonds create a loop with the neutral and ground. This is like a ground loop only worse. Any loop in the system can start resonating and become an antenna for any noise that gets impressed on the loop. A loop that includes the neutral is directly carrying current from the inverter and inverters are well known for being big noise generators.... so the likely hood of having a large amount of RFI noise broadcast by the loop goes way up. This will typically not impact the function of the inverter but it can impact any radio system nearby. On Analog signals like AM, FM, Ham Radio, the RFI will show up as static and/or weak signal. On digital signals like Bluetooth or Wifi, there is error correcting protocols so the system may still work, but at a reduced performance. Sometimes a greatly reduced performance.


BTW: Even one of the proponents on the forum of leaving the multiple NG bonds admitted that on one large and spread out system he worked on the RFI noise problem was so bad with multiple N-G bonds that it made the inverters shut down.
 
Last edited:
As you have observed there is no consensus on this.

I go with this guiding principle: At any given time there should one and only one Neutral-Ground bond in the circuit.

With multiple LV6548s, you must remove bonding screws in order to achieve this rule.


I assume the breaker panel is on the output of the LV6548s... Given that there will already be an N-G bond on the output of all the inverters, in order to have only one N-G bond, the bonding screws must be removed from all the inverters.

BTW: If you ever decide to hook up a generator, It should not have an N-G bond either.

Counter arguments:
The primary counterargument that I have heard is that removing the screws voids the warranty.....but I have heard mixed messages as to whether the warranty is voided or not. If it is true that it voids the warranty, that leaves you with a no-win choice.

Others have argued that it is OK to have multiple N-G bonds, but frankly, I have never understood their reasoning on why it is OK to have multiple N-G bonds.

The system works with multiple N-G bonds so what is the problem?
There are two problems with multiple N-G bonds
1) Safety. With Multiple N-G bonds, current *will* flow on the ground wire between the two N-G bonds. This is not a direct functional problem, but it can be a safety issue....particularly when the system is being serviced.

2) Radio Frequency Interference. Multiple N-G bonds create a loop with the neutral and ground. This is like a ground loop only worse. Any loop in the system can start resonating and become an antenna for any noise that gets impressed on the loop. A loop that includes the neutral is directly carrying current from the inverter and inverters are well known for being big noise generators.... so the likely hood of having a large amount of RFI noise broadcast by the loop goes way up. This will typically not impact the function of the inverter but it can impact any radio system nearby. On Analog signals like AM, FM, Ham Radio, the RFI will show up as static and/or weak signal. On digital signals like Bluetooth or Wifi, there is error correcting protocols so the system may still work, but at a reduced performance. Sometimes a greatly reduced performance.


BTW: Even one of the proponents on the forum of leaving the multiple NG bonds admitted that on one large and spread out system he worked on the RFI noise problem was so bad with multiple N-G bonds that it made the inverters shut down.
Thank you! At least I'm not feeling as nuts as before. So it sounds like there is no down side with removing both screws, other than the potential voiding of the warranty. I think that's the route I'm going to take. I would argue with them on the warranty what with all the varying information, especially some of the posts including responses from MPP advising to remove the screws. I'm probably going to end up with a couple more of these inverters so I'd like to get it all figured out if I can.

I really appreciate your help!
 
We have a system similar to yours. We elected to remove the N-G bond screw from only one of our LV6548's and from the A-C electrical panel. This is working for us and we are using 240 volts. We use this system to power an FM radio station and we do not notice any RFI interference in our equipment. I have not checked to see if we have current on our neutrals and grounds. I will do that next time I am on-site.
 
As you have observed there is no consensus on this.

I go with this guiding principle: At any given time there should one and only one Neutral-Ground bond in the circuit.

With multiple LV6548s, you must remove bonding screws in order to achieve this rule.


I assume the breaker panel is on the output of the LV6548s... Given that there will already be an N-G bond on the output of all the inverters, in order to have only one N-G bond, the bonding screws must be removed from all the inverters.

BTW: If you ever decide to hook up a generator, It should not have an N-G bond either.

Counter arguments:
The primary counterargument that I have heard is that removing the screws voids the warranty.....but I have heard mixed messages as to whether the warranty is voided or not. If it is true that it voids the warranty, that leaves you with a no-win choice.

Others have argued that it is OK to have multiple N-G bonds, but frankly, I have never understood their reasoning on why it is OK to have multiple N-G bonds.

The system works with multiple N-G bonds so what is the problem?
There are two problems with multiple N-G bonds
1) Safety. With Multiple N-G bonds, current *will* flow on the ground wire between the two N-G bonds. This is not a direct functional problem, but it can be a safety issue....particularly when the system is being serviced.

2) Radio Frequency Interference. Multiple N-G bonds create a loop with the neutral and ground. This is like a ground loop only worse. Any loop in the system can start resonating and become an antenna for any noise that gets impressed on the loop. A loop that includes the neutral is directly carrying current from the inverter and inverters are well known for being big noise generators.... so the likely hood of having a large amount of RFI noise broadcast by the loop goes way up. This will typically not impact the function of the inverter but it can impact any radio system nearby. On Analog signals like AM, FM, Ham Radio, the RFI will show up as static and/or weak signal. On digital signals like Bluetooth or Wifi, there is error correcting protocols so the system may still work, but at a reduced performance. Sometimes a greatly reduced performance.


BTW: Even one of the proponents on the forum of leaving the multiple NG bonds admitted that on one large and spread out system he worked on the RFI noise problem was so bad with multiple N-G bonds that it made the inverters shut down.
Now I'm really going nuts. I removed the screw from the secondary unit and went ahead and changed it to 2P2 and 180 degrees. BTW, I was able to remove the screw without removing the board. Just a long magnetic screwdriver and very carefully going in above the board and removing the screw. Then I shut it down and set the primary unit to 2P1 and shut it down. Now when I turn them on they come on then shut down in about 15 seconds. Of course I had forgotten the communication cable. Dang. I go to install the cable and I don't find the ports. In Will's video it was a serial cable and there's a serial cable that came with each unit but all I'm seeing are RJ45 ports. Am I missing something. Also, is there some way to reset the whole thing to defaults? Otherwise if the cable doesn't fix it I'm screwed.

Update: I figured out how to get it back to SIG. If I have PV power turned on the unit will stay on so you can make the changes. I had to do it about 5 times to get it to stick but it finally did. I'm still stuck on the communication cable thing but at least I'm back up and running again.
 
Last edited:
Now I'm really going nuts. I removed the screw from the secondary unit and went ahead and changed it to 2P2 and 180 degrees. BTW, I was able to remove the screw without removing the board. Just a long magnetic screwdriver and very carefully going in above the board and removing the screw. Then I shut it down and set the primary unit to 2P1 and shut it down. Now when I turn them on they come on then shut down in about 15 seconds. Of course I had forgotten the communication cable. Dang. I go to install the cable and I don't find the ports. In Will's video it was a serial cable and there's a serial cable that came with each unit but all I'm seeing are RJ45 ports. Am I missing something. Also, is there some way to reset the whole thing to defaults? Otherwise if the cable doesn't fix it I'm screwed.

Update: I figured out how to get it back to SIG. If I have PV power turned on the unit will stay on so you can make the changes. I had to do it about 5 times to get it to stick but it finally did. I'm still stuck on the communication cable thing but at least I'm back up and running again.
I am not at home so don't have access to the manual but as I recall the COMS port uses the RJ45 plugs.
 
Back
Top