Problems with Eg4 6500ex buyer beware.

[Zwy]

Depends on design. Some very cheap capacitor dropper power supply circuits may get upset but those are worst type of designs.

Did you ground your battery or solar array negative wire? Is your inverter neutral bonded to ground? Is your light switch wired to hot wire or neutral?

Battery cabinet and solar array are grounded but not the battery negative.

Inverter is EG4 6500EX and these do not bond N-G with the later firmware. EG4 decided to go to a common neutral type of system where N and G are pass thru, N is not switched. The inverters rely on the upstream N-G bond at the service disconnect.

The older and original LV6548's did switch neutral and do bond N-G at the inverter. I ran a pair for 3 weeks waiting for the 6500EX replacements. Worked perfectly on the system, no problems. Video was in that thread.

Light switch is wired to hot.

If you go thru the videos in the linked thread, you will see I saw a ground issue immediately on my scope and spent the afternoon looking for it. I finally added a N-G bond at the loads panel. It cleared up the problems I saw then. Fast forward to evening, 800W load and lights were strobing. I saw the inverter output voltage fluctuating and decided to remove the N-G bond at the loads panel. That is when the lights glowed with switch off, it is in one of the videos.

No problems with LV6548 inverters, plenty of problems with EG4 6500EX's. I'll be putting the LV6548's back in.

 

[AntronX]

I saw the inverter output voltage fluctuating and decided to remove the N-G bond at the loads panel. That is when the lights glowed with switch off, it is in one of the videos.

My guess is hot to case to ground coupling via inverter AC output filter capacitors.

 

[bones1]

My guess is hot to case to ground coupling via inverter AC output filter capacitors.

How would you rectify if this is the case?.

 

[AntronX]

How would you rectify if this is the case?.

Bond neutral to ground, code be damned. Make sure inverter enclosure is tied to ground as well or you may have some zapping potential on it from EMI filter caps.

 

[timselectric]

How would you rectify if this is the case?.

Return for replacement.
Do not add extra N/G bonding. Creating a fire Hazzard is never the solution.

 

[Zwy]

Bond neutral to ground, code be damned. Make sure inverter enclosure is tied to ground as well or you may have some zapping potential on it from EMI filter caps.

Did that yesterday, although I never found the case to be hot.

It was when the lights were strobing last night, I removed the bonding screw and had the light glowing with the switch off.

I'm re installing LV6548's tonight. Signature Solar can decide what they will do with the 6500EX's.

 

[Zwy]

Return for replacement.

I'm tempted to drive to Sulpher Springs and return them thru the front window.

Do not add extra N/G bonding. Creating a fire Hazzard is never the solution.

Not a fire hazard but can be objectionable current on G. This can lead to a possible ground fault not tripping a breaker.

What I can't determine is why I'm seeing over 7V on N to G when the 6500EX inverter is source, never see it with the LV6548.

 

[AntronX]

Do not add extra N/G bonding. Creating a fire Hazzard is never the solution.

Can you explain how exactly or you are just reciting your code book?

 

[timselectric]

Not a fire hazard but can be objectionable current on G.

Objectionable current on the grounding system is a fire Hazzard. Equipment grounding conductors are not sized to carry the full constant current of a circuit.

 

[timselectric]

Can you explain how exactly or you are just reciting your code book?

See above
And you said you didn't care about the code. So, it would be useless to recite it.

 

[AntronX]

Objectionable current on the grounding system is a fire Hazzard. Equipment grounding conductors are not sized to carry the full constant current of a circuit.

I suspect this is what's happening right now with no NG bond at the loads panel inside. He's measuring 7 volts N to G at loads panel unbonded. This could mean that there is current on the ground wire going out to the NG bond at the pole and returns via N wire. Could be due to miswired appliance where N is wired to G putting return current onto the ground wire. His LV6548 that did not exibit this problem did have NG screw in one of the units which would be enough to NG bond both inverters via inverter case ground.

 

[timselectric]

I suspect this is what's happening right now with no NG bond at the loads panel inside. He's measuring 7 volts N to G at loads panel unbonded. This could mean that there is current on the ground wire going out to the NG bond at the pole and returns via N wire. Could be due to miswired appliance where N is wired to G putting return current onto the ground wire. His LV6548 that did not exibit this problem did have NG screw in one of the units which would be enough to NG bond both inverters via inverter case ground.

If there is current on the grounding system. It should be located and corrected. Adding a second bond, just conceals it for the moment. And makes it seem safe, until something goes wrong. And then someone ends up getting hurt.

 

[AntronX]

Adding a second bond, just conceals it for the moment. And makes it seem safe, until something goes wrong. And then someone ends up getting hurt.

I can agree with that!

 

[bones1]

So put one of the missing inverter screws that we were told to remove back in, just one?

 

[bones1]

What about the new firmware that supposedly removes the internal bonding from the ex6500 models, if you believe that.

 

[timselectric]

What about the new firmware that supposedly removes the internal bonding from the ex6500 models, if you believe that.

Which firmware do you have?
Stationary or mobile?

 

[timselectric]

Or pre upgraded firmware?

 

[bones1]

Or pre upgraded firmware?

Stationary firmware, both bonding screws previously removed.

 

[timselectric]

Stationary firmware, both bonding screws previously removed.

This is a common neutral setup.
Neutral pass through/ bonding relay should remain in neutral pass through mode.
The only N/G bond should be provided by your existing service. In all modes.

 

[Samsonite801]

So put one of the missing inverter screws that we were told to remove back in, just one?

Yeah, the NEC states that there should only be one N-G bond (for all the house circuits), as close to the power source as possible (usually in standard non-inverter application its located at a main panel in a system using grid source). Since there are 2 power sources (in a dual inverter setup), the best compromise is to just pick one like the master inverter and go with the bond on that one...

If you have grid power as well, then when the inverters are in bypass mode, the N-G bond relay in the inverter opens, then you should have a main supply panel N-G grid bond upstream that is closed for when the inverters are off.

 

[timselectric]

Yeah, the NEC states that there should only be one N-G bond (in the structure), as close to the power source as possible (usually at a main panel in a system using grid source). Since there are 2 power sources (in a dual inverter setup), the best compromise is to just pick one like the master inverter and go with the bond on that one...

If you have grid power as well, then when the inverters are in bypass mode, the N-G bond relay in the inverter opens, then you should have a main supply panel N-G grid bond upstream that is closed for when the inverters are off.

That's not how these units work. There have been recent modifications to operate in a common neutral setup. They have the stationary firmware installed.

 

[Samsonite801]

That's not how these units work. There have been recent modifications to operate in a common neutral setup. They have the stationary firmware installed.

Can you explain more about how the new setup works? I heard about some changes with how they ship with grounding screws removed now on rough scale, but I haven't learned about the technical detail and logic of this change, so I cannot speak for the 'new' way of doing it I guess now.

I still have some LV6548s and have worked with several of the 6500EX models shipped with the 'old' setup, so I guess I have something new to learn about it then.

Do they just count on you bonding upstream at the grid input supply panel or something now? Or is bonding on the inverter output?

 

[Adam De Lay]

Do they just count on you bonding upstream at the grid input supply panel or something now?

Yeah if you have grid input, then you rely on the neutral/ground bond which should already exist in or near your MSP. If you don’t have any grid input, the recommendation is to make a ground/neutral bond in your critical loads panel (or your first panel downstream of the inverter).

 

[timselectric]

Can you explain more about how the new setup works? I heard about some changes with how they ship with grounding screws removed now on rough scale, but I haven't learned about the technical detail and logic of this change, so I cannot speak for the 'new' way of doing it I guess now.

I still have some LV6548s and have worked with several of the 6500EX models shipped with the 'old' setup, so I guess I have something new to learn about it then.

Do they just count on you bonding upstream at the grid input supply panel or something now? Or is bonding on the inverter output?

The unit provides no bonding. It uses a common neutral setup. Input neutral is continuously connected to output neutral. Similar to how the top tier units are made.

 

[Samsonite801]

The unit provides no bonding. It uses a common neutral setup. Input neutral is continuously connected to output neutral. Similar to how the top tier units are made.

Sounds good. I guess I still don't understand what the firmware upgrade has to do with it though...