EG4 6000EX - Double Ground/Neutral

[Zwy]

I have an EG4 3kW and I measured the voltage as you requested and I got 62Volts. You stated if the voltage is more than a volt or so, it would indicate there is no bond while in inverter mode.

62V sounds like a parallel path.

Signature solar had me test my unit for continuity a few days ago and told me if I had continuity, the inverter had an internal bond. I did what they did and it had continuity when the inverter was powered on and the on/off switch was on. (Note: i had disconnected my ground and AC outs to do the test.

I also have my ground and neutral (bonded) in the little panel i have that is fed by my inverter.

This is important. The AC input was disconnected but the small panel powered by the inverter was still connected?

With N-G bonded in both the inverter and small panel, you have created a parallel path. I'd remove the bonding screw in the panel and retest.

Just like the creator of the thread HLD. So is this an issue for me too?

Yes.

 

[SoFar]

62V sounds like a parallel path.


This is important. The AC input was disconnected but the small panel powered by the inverter was still connected?

With N-G bonded in both the inverter and small panel, you have created a parallel path. I'd remove the bonding screw in the panel and retest.



Yes.

I'm confused, Filterguy said "That is a sure sign the neutral is not bonded to the ground."

You are saying it is bonded in the inverter?

 

[Zwy]

I'm confused, Filterguy said "That is a sure sign the neutral is not bonded to the ground."

You are saying it is bonded in the inverter?

Possibly, why would voltage between N and G be 1/2 of Line output? I would expect that with a parallel path or loop or floating ground.

But if there was bonding in the inverter, the voltage between N and G should be less than 1 volt. There isn't a ground electrode on the panel, correct? And you have the EGC from service panel disconnected? If you don't have a ground electrode on the small panel and you don't have the EGC from the service panel connected, you will not have earth reference.

It appears to be a parallel path or loop somewhere or even lack of earth reference. I'd remove the N-G bond in the small panel and retest voltage between N and G. I'd also test voltage between N and G with an EGC to the input of the inverter to have earth reference. I'd add a load in the small panel so there is current on N and test voltage between N and G. If you have an amp clamp, then test for current on G with a load present and see if the current is roughly the same on N.

 

[FilterGuy]

I have an EG4 3kW and I measured the voltage as you requested and I got 62Volts. You stated if the voltage is more than a volt or so, it would indicate there is no bond while in inverter mode.

i disconnected the feeds to the panel (including the ground) when I checked the voltage.

@Zwy When he measured the voltage, the panel on the output was disconnected so the bond in the panel did not impact the measurement.
To me, it is obvious the inverter is not bonding while in battery/inverter mode.

With the panel (and bond) hooked up to the output, there will be one bond when the unit is in battery mode (Good), but when in passthrough there will be two bonds (Bad).

Assuming there is only one inverter, the three ways to fix this are.

1) Install the bonding screw if EG4/Sig Solar allows it.
2) Install a jumper from output neutral back to input neutral even though EG4/Sig Solar refuses to say whether this is supported.
3) Throw out the EG4 and buy an inverter that supports and documents proper grounding and bonding.

 

[Zwy]

@Zwy When he measured the voltage, the panel on the output was disconnected so the bond in the panel did not impact the measurement.

My understanding is he disconnected the AC input from the main panel. The small panel supplied by the inverter is still connected.

Maybe I'm reading wrong but that was my take on the situation.

To me, it is obvious the inverter is not bonding while in battery/inverter mode.

With the panel (and bond) hooked up to the output, there will be one bond when the unit is in battery mode (Good), but when in passthrough there will be two bonds (Bad).

That is what my thinking is, if the bonded small panel is connected to AC output which I understand to be the case. If the small panel is connected and bonded and the inverter is bonding N-G, then he has 2 bonds.

Again, my understanding is the small panel is connected.

Assuming there is only one inverter, the three ways to fix this are.

1) Install the bonding screw if EG4/Sig Solar allows it.
2) Install a jumper from output neutral back to input neutral even though EG4/Sig Solar refuses to say whether this is supported.

I'd rather determine what is connected and disconnected before moving on to a solution. I'd prefer an earth reference no matter what with an EGC from the service panel to the inverter connected.

3) Throw out the EG4 and buy an inverter that supports and documents proper grounding and bonding.

You have a good point there.

 

[SoFar]

My understanding is he disconnected the AC input from the main panel. The small panel supplied by the inverter is still connected.

Maybe I'm reading wrong but that was my take on the situation.



That is what my thinking is, if the bonded small panel is connected to AC output which I understand to be the case. If the small panel is connected and bonded and the inverter is bonding N-G, then he has 2 bonds.

Again, my understanding is the small panel is connected.


I'd rather determine what is connected and disconnected before moving on to a solution. I'd prefer an earth reference no matter what with an EGC from the service panel to the inverter connected.

You have a good point there.

I have no AC input. Just "for now" only have 8 36v panels (280w), EG4 3kW 120V inverter and two 5kW lithium batteries.

According to tech at SigSolar he told me after I did continuity test over phone with him the neutral & bond were connected in AIO.

When I first checked inverter with battery on, but AIO in off position there was no continuity. After turning the AIO on, I got continuity.


There is a note in manual that says "When the inverter is working in PV mode, battery mode or standby mode, the output neutral is connected to the ground of AC output (neutral/ground bonded).

When the inverter is working in AC mode, neutral of output is disconnected to grounding of AC output and connected to neutral of AC input."

See attached pic.

My understanding from phone call the bonding "somehow discoonects" when connected to AC input

 

[FilterGuy]

According to tech at SigSolar he told me after I did continuity test over phone with him the neutral & bond were connected in AIO.

When I first checked inverter with battery on, but AIO in off position there was no continuity. After turning the AIO on, I got continuity.

The fact that you had no continuity when the inverter is off tells me the screw is not there. The fact that you saw 62V between neutral and ground on the output tells me the screw is not there. What is the manufacture date of the unit?

Measuring continuity with the unit on is a bogus test. The voltage from the inverter will confuse the ohmmeter. (It is *really* sad the sig solar techs are so poorly trained that they don't know this)


An ohm meter or a continuity tester works by putting a voltage accross the two points and measuring the current.

This is a simple schematic of an ohmmeter





Now lets put 120V AC on it.




The 120V is going to make any measurment by the ohmmeter completely bogus.

 

[Bud Martin]

And hopefully did not damage the meter!
BTW, welcome to DIYSolarforum-SSsupport.com

 

[SoFar]

The fact that you had no continuity when the inverter is off tells me the screw is not there. The fact that you saw 62V between neutral and ground on the output tells me the screw is not there. What is the manufacture date of the unit?

Measuring continuity with the unit on is a bogus test. The voltage from the inverter will confuse the ohmmeter. (It is *really* sad the sig solar techs are so poorly trained that they don't know this)


An ohm meter or a continuity tester works by putting a voltage accross the two points and measuring the current.

This is a simple schematic of an ohmmeter

View attachment 132351



Now lets put 120V AC on it.


View attachment 132352

The 120V is going to make any measurment by the ohmmeter completely bogus.

It's a fairly new inverter. It's the newer model (second version) I purchased it in late November.

I thought I read posts where they were not putting in ground screw in them anymore.


Here is a link on YouTube where guy says you need to add bonding screw for his camper

 

[FilterGuy]

I thought I read posts where they were not putting in ground screw in them anymore.

Yes, they made that change without telling anyone sometime in Q4 of last year.

 

[SoFar]

Yes, they made that change without telling anyone sometime in Q4 of last year.

So user ZWY says my unit is N-G bonded, but you say it isn't. And Sig Solar tech told me it was after I did the continuity test.

You can see how easy it is to get confused.

EDIT
After re-read I g user ZWY posts, I think he concurs with you, that it is not bonded in battery/ inverter mode.

 

[FilterGuy]

So user ZWY says my unit is N-G bonded, but you say it isn't. And Sig Solar tech told me it was after I did the continuity test.

You can see how easy it is to get confused.

Yes, I see how you could be confused and I don't know how to advise you on who to believe. All I can do is give my reasoning for why I think the other opinions are incorrect and why I think the screw is not there.

* Sig solar had you do a bogus test. (See explanation above)
* @Zwy believes you measured the 62V when the inverter output was hooked to the small distribution box with the N-G bond. (How you would see 62V across an NG bond is beyond me)

Meanwhile, you reported three things that, in my opinion, clearly indicate there is no bonding screw:
1) 62V across the output neutral and the ground is a clear indication that there is no bond and therefore no bonding screw. If there was a bond between Neutral and ground, the voltage would have to be close to zero.
2) Every inverter that does dynamic bonding that I have worked with uses the Normally-Closed position of the bonding relay for the bond. This means that if the inverter is off, the relay is in the bonding-on state. You measured continuity with the inverter off and saw no continuity. This indicates the screw has been removed.
3) You have a new unit and EG4 has said they are now shipping all models without the bonding screw.

 

[SoFar]

Yes, I see how you could be confused and I don't know how to advise you on who to believe. All I can do is give my reasoning for why I think the other opinions are incorrect and why I think the screw is not there.

* Sig solar had you do a bogus test. (See explanation above)
* @Zwy believes you measured the 62V when the inverter output was hooked to the small distribution box with the N-G bond. (How you would see 62V across an NG bond is beyond me)

Meanwhile, you reported three things that, in my opinion, clearly indicate there is no bonding screw:
1) 62V across the output neutral and the ground is a clear indication that there is no bond and therefore no bonding screw. If there was a bond between Neutral and ground, the voltage would have to be close to zero.
2) Every inverter that does dynamic bonding that I have worked with uses the Normally-Closed position of the bonding relay for the bond. This means that if the inverter is off, the relay is in the bonding-on state. You measured continuity with the inverter off and saw no continuity. This indicates the screw has been removed.
3) You have a new unit and EG4 has said they are now shipping all models without the bonding screw.

Thanks.

I had hoped at some point to feed my inverter when batteries got low in my off-grid setup from an outlet from house, but will refrain from doing that with the bonding issues.

Thanks again.

 

[Zwy]

So user ZWY says my unit is N-G bonded, but you say it isn't. And Sig Solar tech told me it was after I did the continuity test.

You can see how easy it is to get confused.

EDIT
After re-read I g user ZWY posts, I think he concurs with you, that it is not bonded in battery/ inverter mode.

One question, is the N-G bonded small panel connected to the inverter output?

As for whether the inverter is dynamically bonded, I agree with @FilterGuy. Why is there 62V between N and G at the inverter? If the small panel is connected and N-G bonded and the inverter is not, my assumption is 1/2 the current is traveling back to the inverter (source) on G. Which can only be possible with some type of path inside the inverter. Maybe the unit is defective, something just doesn't seem right.

If the small panel isn't connected during testing, then ignore all the above.

 

[summit]

It seems the problem is between bypass through mode (ACIN) and inverter mode. In bypass, NG is bonded back in the main panel via the EG's ACINput. Thus there would be 2x NG bonds with the NG bond in the critical load box. In inverter mode, ACOUT-N is float and thus needs the NG bond in the critical load panel. I guess you can implement a relay to dynamically NG bond in your critical load box: unbond while in bypass-mode and bond in inverter mode....but where's the trigger signal ? is there an ON/OFF switch ?

 

[PreppenWolf]

One question, is the N-G bonded small panel connected to the inverter output?

As for whether the inverter is dynamically bonded, I agree with @FilterGuy. Why is there 62V between N and G at the inverter? If the small panel is connected and N-G bonded and the inverter is not, my assumption is 1/2 the current is traveling back to the inverter (source) on G. Which can only be possible with some type of path inside the inverter. Maybe the unit is defective, something just doesn't seem right.

If the small panel isn't connected during testing, then ignore all the above.

You can test the above if the OP has a clamp meter. If half the current is returning to ground we will see it.

 

[SoFar]

One question, is the N-G bonded small panel connected to the inverter output?

As for whether the inverter is dynamically bonded, I agree with @FilterGuy. Why is there 62V between N and G at the inverter? If the small panel is connected and N-G bonded and the inverter is not, my assumption is 1/2 the current is traveling back to the inverter (source) on G. Which can only be possible with some type of path inside the inverter. Maybe the unit is defective, something just doesn't seem right.

If the small panel isn't connected during testing, then ignore all the above.

The N-G is bonded in my small panel that is fed from my inverter. When i did the test that FilterGuy recomended I disconected everything from the AC out on the inverter (ground, neutral, hot). The only thing connected to the inverter when i did my test was the battery cables. I beleive i had my solar PV disconnected when i did the test.

 

[Zwy]

The N-G is bonded in my small panel that is fed from my inverter. When i did the test that FilterGuy recomended I disconected everything from the AC out on the inverter (ground, neutral, hot). The only thing connected to the inverter when i did my test was the battery cables. I beleive i had my solar PV disconnected when i did the test.

Thank you, yes, the inverter is not bonding N-G.

 

[ig1628]

Here a new update from EG4 about NGB

 

[FilterGuy]

OK!!! Thank you for the video and the fix EG4/Sig Solar. This is a huge improvement! Up till yesterday, I was under the impression the relays in the internal transfer switch were not individually addressable. The fact that they are addressable is a game changer.

* With the new firmware, the system operates as a common neutral and that works well for most (All?) stationary systems.
* For mobile solutions, the 'mobile' firmware that uses the bonding relay will work well in most cases (See limits below).

An important question:
When will this all be documented in the manual?

A question to satisfy my curiosity:
Why not put a new 'program' in the firmware that allows the user to select common neutral vs mobile? Was the solution done to get it out quickly with a plan for a configurable firmware solution coming later?

A question about shipping with the bonding screw.
If the relays are individually addressable, the new firmware can be set up to not engage the bonding relay. Therefore the bond would not be created even if the bonding screw is in place. Why not do this and go back to shipping with the screws so the mobile user does not have to source his own screw and open the inverter to install it?

Finally, there are some limitations for the mobile Firmware that should be noted

Stacking inverters:
* With the common neutral firmware, the bonding relays are not used and therefore the relays do not impose a limit on stacking (There may be other limits though)
* With the mobile firmware, The only stacking that should be done is two inverters in split phase mode. Two in parallel mode or more than two inverters in any mode create a situation where the bonding relay is undersized.
* With the mobile firmware, The bonding screw in one of the inverters should be removed.

NOTE: Hopefully you can make this configurable in a future version of the firmware... better yet, when the mobile firmware is operating, only enable the dynamic bonding on the master inverter and not on the other inverter.​

Non-Neutral-switching transfer switches on the output.
* With the common neutral firmware, a non-neutral switching transfer switch on the inverter output will work fine. This means the Reliant multi-circuit transfer switch is OK to use.
* With the mobile firmware, a non-neutral-switching transfer switch could create two bonds when on battery power and therefore should not be used.