diy solar

diy solar

Signature Solar's new EG4 6000 120/240V inverter

If you never hook the grid to the input that is true. However if you hook the grid to the input, then in pass-through mode there will be two N-G bonds..... that is a no-no.

IMO a lot of this is pragmatic.
If you are energizing a legacy main panel it very likely has a legacy n-g bond.
Don't change it unless you have to.
Its usually the path of least resistance unless the electrician who originally wired the panel happened to be fastidious about keeping the neutrals and grounds separate.

Probably a good idea to put a label on the panel to make it clear where the n-g bond is.
 
Yeah no grid power at all!
I'm not an electrician, but what I've been told and read is that the ground occurs at the source of the power.

If that was grid, then it's the first breaker panel after the electric company. Usually this can be considered as the panel with the 100A or greater main breaker.

In your case, the pair of inverters is the source of power, so neutral and ground should be bonded there, and never anything after that.

I've seen a ton of posts on where to bond, but not why, other than mentioning ground loops. Well the reason you don't want ground loops is because us soft fleshy things can act as the path of okay resistance, and that often hurts.

In all seriousness though, ground loops can affect the ability of safety devices, namely GFCI and ACFI receptacles/breakers, from doing their job. Fleshy bits in the current flow may not be detected by the safety device (GFCI), plus even if we are not involved, higher risk of fire from arcs not being detected (AFCI).
 
I am NOT an expert, just my speculation.

I think it is wrong to put power in on breaker panel breakers that normally are where you attach power out, in the same breaker box where you have breakers that send power out to loads.

In my opinion, I would think it would be safer to dedicate a small breaker distribution panel for all incoming breakers to bring in all paralleled breakers, then the main breaker on that box, becomes an outgoing Main Breaker - which is then connected to a second breaker panel’s Main input breaker, to be distributed to all outgoing breaker loads.

So all paralleled inverter power sources are combined in a breaker box dedicated to only combining incoming power.

Then the Main from the AC Combiner Breaker Panel is then wired to the Distribution Breaker Panel’s Main breaker (just as if it was coming from the grid).

Any feedback on this idea?

If you like this concept, the next question is where to apply the ground neutral bond?

I speculate it should be in the Distribution Breaker Panel????
I'm in total agreement with this from a purest standpoint, but it's not always a physically clean situation when you are doing a retrofit and adding inverter power to an existing installation. Unless you are willing and have the resources to completely rebuild all of the panel distribution circuits, the ideal way to inject power into an existing system is by bringing it inbound via a breaker.

There are numerous systems that support a totally safe way of doing this, primarily being a mechanical lockout between the main breaker and the incoming alternative power source. There is a minor issue in that mechanical lockouts typically only affect the top one or two breaker slots on one side of a panel. So if you have multiple inverter breakers into that main panel, there is a possibility that the mechanical lockout only affects one of those.

In that case, I would do the combiner box between the multiple inverters and the main panel, and bring the power into the main panel from the combiner box via a much heavier breaker, such as a 60 amp double pole all the way up through a 125 amp double pole.

Transfer switches are also capable of handling this as well whether they be mechanical or electric. They are in essence A/B switches and they would switch between the incoming ground power and the incoming inverter/generator power, thus only allowing one at a time. The installation of transfer switches is only clean though if the alternative power source is running the entire house, but often the alternative power will be brought into a critical loads panel panel via a breaker as well.

IMO, neutral and ground bonding when using a combiner box is fine all the way up to and including the combiner box itself, which includes the inverters.
 
I'm not an electrician, but what I've been told and read is that the ground occurs at the source of the power.

If that was grid, then it's the first breaker panel after the electric company. Usually this can be considered as the panel with the 100A or greater main breaker.

In your case, the pair of inverters is the source of power, so neutral and ground should be bonded there, and never anything after that.

I've seen a ton of posts on where to bond, but not why, other than mentioning ground loops. Well the reason you don't want ground loops is because us soft fleshy things can act as the path of okay resistance, and that often hurts.

In all seriousness though, ground loops can affect the ability of safety devices, namely GFCI and ACFI receptacles/breakers, from doing their job. Fleshy bits in the current flow may not be detected by the safety device (GFCI), plus even if we are not involved, higher risk of fire from arcs not being detected (AFCI).
Actually, Ground loops are not a safety issue. In fact, there is nothing in the NEC that forbids ground loops. The problem with ground loops is that they create Radio Frequency Interference issues.

The problem with no N-G bonds is that there is no low impedance path to clear a fault.
The problem with multiple N-G bonds is that it puts current on the Ground wire.... That by itself is not a big deal, but it can create unexpected voltages on an Equipment grounding conductor, particularly during servicing. In addition, multiple N-G bonds creates a loop involving a current carrying conductor (Neutral) and that can *really* create RFI issues.

For a more detailed explanation of grounding, you may want to review this series of papers:
 
Actually, Ground loops are not a safety issue. In fact, there is nothing in the NEC that forbids ground loops. The problem with ground loops is that they create Radio Frequency Interference issues.

The problem with no N-G bonds is that there is no low impedance path to clear a fault.
The problem with multiple N-G bonds is that it puts current on the Ground wire.... That by itself is not a big deal, but it can create unexpected voltages on an Equipment grounding conductor, particularly during servicing. In addition, multiple N-G bonds creates a loop involving a current carrying conductor (Neutral) and that can *really* create RFI issues.

For a more detailed explanation of grounding, you may want to review this series of papers:
Good info @FilterGuy, I'll check them out!
 
I'm not an electrician, but what I've been told and read is that the ground occurs at the source of the power.

If that was grid, then it's the first breaker panel after the electric company. Usually this can be considered as the panel with the 100A or greater main breaker.

In your case, the pair of inverters is the source of power, so neutral and ground should be bonded there, and never anything after that.

I've seen a ton of posts on where to bond, but not why, other than mentioning ground loops. Well the reason you don't want ground loops is because us soft fleshy things can act as the path of okay resistance, and that often hurts.

In all seriousness though, ground loops can affect the ability of safety devices, namely GFCI and ACFI receptacles/breakers, from doing their job. Fleshy bits in the current flow may not be detected by the safety device (GFCI), plus even if we are not involved, higher risk of fire from arcs not being detected (AFCI).
Actually, Ground loops are not a safety issue. In fact, there is nothing in the NEC that forbids ground loops. The problem with ground loops is that they create Radio Frequency Interference issues.

The problem with no N-G bonds is that there is no low impedance path to clear a fault.
The problem with multiple N-G bonds is that it puts current on the Ground wire.... That by itself is not a big deal, but it can create unexpected voltages on an Equipment grounding conductor, particularly during servicing. In addition, multiple N-G bonds creates a loop involving a current carrying conductor (Neutral) and that can *really* create RFI issues.

For a more detailed explanation of grounding, you may want to review this series of papers:
So is the general consensus to leave the screw in the inverters and keep the grounds and neutrals separate in the panel?
 
I'm in total agreement with this from a purest standpoint, but it's not always a physically clean situation when you are doing a retrofit and adding inverter power to an existing installation. Unless you are willing and have the resources to completely rebuild all of the panel distribution circuits, the ideal way to inject power into an existing system is by bringing it inbound via a breaker.

There are numerous systems that support a totally safe way of doing this, primarily being a mechanical lockout between the main breaker and the incoming alternative power source. There is a minor issue in that mechanical lockouts typically only affect the top one or two breaker slots on one side of a panel. So if you have multiple inverter breakers into that main panel, there is a possibility that the mechanical lockout only affects one of those.

In that case, I would do the combiner box between the multiple inverters and the main panel, and bring the power into the main panel from the combiner box via a much heavier breaker, such as a 60 amp double pole all the way up through a 125 amp double pole.

Transfer switches are also capable of handling this as well whether they be mechanical or electric. They are in essence A/B switches and they would switch between the incoming ground power and the incoming inverter/generator power, thus only allowing one at a time. The installation of transfer switches is only clean though if the alternative power source is running the entire house, but often the alternative power will be brought into a critical loads panel panel via a breaker as well.

IMO, neutral and ground bonding when using a combiner box is fine all the way up to and including the combiner box itself, which includes the inverters.
Thank you, this is good to hear, that what I had speculated, is a plausible solution.
 
So is the general consensus to leave the screw in the inverters and keep the grounds and neutrals separate in the panel?
Based on the debates we have had on the forum I would say there is *not* a consensus. I can only give you my understanding and recommendation.

First the basic rules.
  1. There should always be one N-G bond in the circuit.
  2. There should never be more than one N-G bond in the circuit.

How to meet these rules depends on what the inverter does.

Since this thread is about the 6000EX, you are obviously asking what should be done for bonding on the 6000EX. Unfortunately, I don't yet know. Do you know if there are even bonding screws on that unit? (If you do know please let us know. It would be very helpful)

From the responses I have seen on this thread, it looks like the output has an N-G bond in both the pass-through mode and in inverter mode in it's default configuration out of the box. What I don't know is if it is a dynamic bonding system or a common neutral system.

Always having a bond on the output is great for using a single inverter, but since you are asking about a dual inverter, it can be problematic. (There will be two N-G bonds in the system). If there are bonding screws in the inverters, the screw should be removed on ONE of the inverters.)


For incompleteness, here are descriptions of what I have found on other inverters:

* Switched Neutral, Dynamic bonding AIOs or inverter chargers
A lot of the low-cost Chinees All-in-ones (MPP, EG4, etc) are built with an internal transfer switch that switches neutral and has dynamic bonding that adds an N-G bond when in inverter mode and removes it when in pass-through mode. For these inverters, it is usually best to leave the bonding screw in and not do a bond in the critical load panel. However, if you are stacking these inverters to get split-phase or single-phase with more wattage, the bonding screws should be removed in all but one of the inverters. Unfortunately, I have yet to find one of these that document the bonding scheme.

BTW: The Victron inverter-chargers I have worked with have Switched Neutral with Dynamic bonding. They document it all very well.

Common Neutral, no bonding AIOs or inverter chargers.
* Some higher-end all-in-ones like Schnider and Solark don't switch the neutral. I call these common neutral because the input and output neutral are the same conductors. These inverters count on the N-G bond in the grid connection for both inverter mode and transfer mode. The only time you would add an n-g bond with these inverters is if the installation is completely off-grid and there is no existing N-G bond.

IMHO. Common Neutral inverters are the best solution for stationary systems that have a constant tie to the N-G bond from the grid. However, in a mobile system where the grid connection comes and goes, it is very difficult to make these work well for bonding.

Switched Neutral, No bonding
*Some of the low-cost Chinese all-in-ones switch neutral but do not have dynamic bonding. (Growatt seems to have a few of these). These are a PITA. The only way I know how to deal with these is to connect the input neutral to the output neutral and treat it like a common neutral inverter. This seems to work but it is not clear whether the manufacturers support it. What is worse, on most of these there is no documentation that explains any of this to you.

Battery-powered Inverters.
The regular battery inverters (Not AIOs or Inverter Chargers) are all over the map. Most of them do not add an N-G bond and you need to have one in the system you are driving. Some (like giandle) do not add an N-G bond and the inverter does not provide an N-G bond. The output circuit is left floating.

Grid Tie inverters that feed power back to the grid.
As far as I know, all of these do not create an N-G bond and count on the circuit they are tieing to having the bond.
 
Last edited:
Based on the debates we have had on the forum I would say there is *not* a consensus. I can only give you my understanding and recommendation.

First the basic rules.
  1. There should always be one N-G bond in the circuit.
  2. There should never be more than one N-G bond in the circuit.

How to meet these rules depends on what the inverter does.

Since this thread is about the 6000EX, you are obviously asking what should be done for bonding on the 6000EX. Unfortunately, I don't yet know. Do you know if there are even bonding screws on that unit? (If you do know please let us know. It would be very helpful)

From the responses I have seen on this thread, it looks like the output has an N-G bond in both the pass-through mode and in inverter mode in it's default configuration out of the box. What I don't know is if it is a dynamic bonding system or a common neutral system.

Always having a bond on the output is great for using a single inverter, but since you are asking about a dual inverter, it can be problematic. (There will be two N-G bonds in the system). If there are bonding screws in the inverters, the screw should be removed on ONE of the inverters.)


For incompleteness, here are descriptions of what I have found on other inverters:

* Switched Neutral, Dynamic bonding AIOs or inverter chargers
A lot of the low-cost Chinees All-in-ones (MPP, EG4, etc) are built with an internal transfer switch that switches neutral and has dynamic bonding that adds an N-G bond when in inverter mode and removes it when in pass-through mode. For these inverters, it is usually best to leave the bonding screw in and not do a bond in the critical load panel. However, if you are stacking these inverters to get split-phase or single-phase with more wattage, the bonding screws should be removed in all but one of the inverters. Unfortunately, I have yet to find one of these that document the bonding scheme.

BTW: The Victron inverter-chargers I have worked with have Switched Neutral with Dynamic bonding. They document it all very well.

Common Neutral, no bonding AIOs or inverter chargers.
* Some higher-end all-in-ones like Schnider and Solark don't switch the neutral. I call these common neutral because the input and output neutral are the same conductors. These inverters count on the N-G bond in the grid connection for both inverter mode and transfer mode. The only time you would add an n-g bond with these inverters is if the installation is completely off-grid and there is no existing N-G bond.

IMHO. Common Neutral inverters are the best solution for stationary systems that have a constant tie to the N-G bond from the grid. However, in a mobile system where the grid connection comes and goes, it is very difficult to make these work well for bonding.

Switched Neutral, No bonding
*Some of the low-cost Chinese all-in-ones switch neutral but do not have dynamic bonding. (Growatt seems to have a few of these). These are a PITA. The only way I know how to deal with these is to connect the input neutral to the output neutral and treat it like a common neutral inverter. This seems to work but it is not clear whether the manufacturers support it. What is worse, on most of these there is no documentation that explains any of this to you.

Battery-powered Inverters.
The regular battery inverters (Not AIOs or Inverter Chargers) are all over the map. Most of them do not add an N-G bond and you need to have one in the system you are driving. Some (like giandle) do not add an N-G bond and the inverter does not provide an N-G bond. The output circuit is left floating.

Grid Tie inverters that feed power back to the grid.
As far as I know, all of these do not create an N-G bond and count on the circuit they are tieing to having the bond.
I’m actually not sure! I didn’t see anything in the manual. Here are some photos
 

Attachments

  • E337BF94-B49C-46DC-826B-98A238155DDB.jpeg
    E337BF94-B49C-46DC-826B-98A238155DDB.jpeg
    230.5 KB · Views: 17
  • F33CE50D-42EA-4798-AB1C-46B3FF4E2AB5.jpeg
    F33CE50D-42EA-4798-AB1C-46B3FF4E2AB5.jpeg
    186.4 KB · Views: 17
  • 0293CBDA-271A-4484-8586-12A970796B51.jpeg
    0293CBDA-271A-4484-8586-12A970796B51.jpeg
    203 KB · Views: 16
  • 67BEA322-A52A-4DC4-A9ED-50A841869111.jpeg
    67BEA322-A52A-4DC4-A9ED-50A841869111.jpeg
    172.9 KB · Views: 15
  • E0CC0695-0E03-47CB-B6D1-BD3171A36435.jpeg
    E0CC0695-0E03-47CB-B6D1-BD3171A36435.jpeg
    142 KB · Views: 16
I’m actually not sure! I didn’t see anything in the manual.



<Warning: RANT MODE ON>
Why can't these $%$#!@& manufacturers give us information on the grounding of the inverter!!! Why do we have to reverse engineer their product to get this basic information? This information is fundimental to understanding how the inverter should be safely wired and yet EG4, MPP and Growatt all leave this critical information out of their manuals. It is nearly criminal negligence!!

Ok... I feel better now.
<RANT MODE OFF>

Here are some photos
Thanks for the pics!! Very helpful.

The circled screw is almost certainly a Bonding screw.

1664176844802.png


The interesting thing is that I count 5 power relays.....
1664177236897.png

Based on the number of relays, I am almost positive the inverter is doing dynamic bonding. These are the type of relays used for the internal transfer switch for all of the low-cost AIO inverters. What I am scratching my head on is what all 5 are for.
  1. Hot 1
  2. Hot 2
  3. Neutral
  4. Bonding
  5. ????
Someone might suggest the fifth is to switch ground, but there is no need to switch ground, it is an unsafe idea to switch ground, and I don't know of any other inverter that switches ground


Havings said all that, based on everything so far in this thread, this is my best guess at a model for the inverter.

1664177929917.png
 
<Warning: RANT MODE ON>
Why can't these $%$#!@& manufacturers give us information on the grounding of the inverter!!! Why do we have to reverse engineer their product to get this basic information? This information is fundimental to understanding how the inverter should be safely wired and yet EG4, MPP and Growatt all leave this critical information out of their manuals. It is nearly criminal negligence!!

Ok... I feel better now.
<RANT MODE OFF>


Thanks for the pics!! Very helpful.

The circled screw is almost certainly a Bonding screw.

View attachment 113596


The interesting thing is that I count 5 power relays.....
View attachment 113598

Based on the number of relays, I am almost positive the inverter is doing dynamic bonding. These are the type of relays used for the internal transfer switch for all of the low-cost AIO inverters. What I am scratching my head on is what all 5 are for.
  1. Hot 1
  2. Hot 2
  3. Neutral
  4. Bonding
  5. ????
Someone might suggest the fifth is to switch ground, but there is no need to switch ground, it is an unsafe idea to switch ground, and I don't know of any other inverter that switches ground


Havings said all that, based on everything so far in this thread, this is my best guess at a model for the inverter.

View attachment 113599
Not to give you more work, but in the 6500 thread I think you had Bunch of different scenarios (inverter scenarios) and diagrams. That was very helpful for couple people I know. Would it be possible to do it for this or with it follow the same?
 
<Warning: RANT MODE ON>
Why can't these $%$#!@& manufacturers give us information on the grounding of the inverter!!! Why do we have to reverse engineer their product to get this basic information? This information is fundimental to understanding how the inverter should be safely wired and yet EG4, MPP and Growatt all leave this critical information out of their manuals. It is nearly criminal negligence!!

Ok... I feel better now.
<RANT MODE OFF>
+1
 
Havings said all that, based on everything so far in this thread, this is my best guess at a model for the inverter.

That looks workable at least for the common use cases.

q: As drawn the MPPT controller appears to be isolated, Is it actually isolated?

q: Is there a chassis grounding lug on this unit?

q: Do we know the specs for the relays, as in how much current and how many cycles?
 
Not to give you more work, but in the 6500 thread I think you had Bunch of different scenarios (inverter scenarios) and diagrams. That was very helpful for couple people I know. Would it be possible to do it for this or with it follow the same?
I would really like to be more positive about the model before I do that. However, it is looking like it is going to be mostly the same.
 
q: As drawn the MPPT controller appears to be isolated, Is it actually isolated?
Most of these units seem to have a separate board for the MPPT controller so yes, they are somewhat isolated. Having said that, the drawing I provided is not intended to be too literal. There are lots of connections from controlling logic/processor to all of the different components.

q: Is there a chassis grounding lug on this unit?
Since I do not have a unit I do not know. Maybe @Marqese can tell us.

q: Do we know the specs for the relays, as in how much current and how many cycles?
There are a few different versions of these relays that have the same basic package. I would need to get a clearer pic of the relays so I can get the specific model number to look up
 
@Marqese could you please test/inspect your all in one?

Check for continuity between pv negative and battery negative.
If there is continuity then the mppt controller is non-isolated.
Also let us know if the unit has a chassis grounding lug.
 
I would think, SignatureSolar could and should contact their OEM (assuming Voltronic), and request a detailed Schematic Diagram of the 6000, and provide a copy of it for us to download. I think this should be the standard across the industry.

Provide a schematic for every inverter, battery, etc.

[ A circuit diagram (wiring diagram, electrical diagram, elementary diagram, electronic schematic) is a graphical representation of an electrical circuit. ]
 
Back
Top