FilterGuy
Solar Engineering Consultant - EG4 and Consumers
Excellent video. I have watched a lot of Mikes's videos, but I had not seen this one. This sums us why I am so adamite about getting this right.
Need to bond neutral + ground for off grid shed using PIP 3048LV-MK 48V 3kW 120V Inverter
- Thread starter hondaman82
- Start date
If you remove all screws except one, the ground fault current will find its way back via the inverter that does have the neutral-ground bond. For example, in your diagram, if you remove only the bonding screw from the top inverter
- Ground fault current from the top inverter (black wire) will travel via the grounding wire (green) to the neutral-ground bond in the top inverter and then to the neutral of the top inverter.
- Ground fault current from the bottom inverter (red wire) will travel via the grounding wire (green) to the neutral-ground bond in the top inverter and then to the 3 pole transfer switch and then to the neutral of the bottom inverter because the neutrals of the top and bottom inverters are connected to each other in the transfer switch.
In FilterGuy's version of your diagram, ground fault current will go through the ground bus in the main panel and then to the neutral-ground connection in the top inverter.
If you remove all screws and have AC input connections, you cannot make connections to allow ground fault current to return to the inverter without connecting the inverter's input neutral to its output neutral, which MPP does not support.
If you remove all screws and are completely off-grid, you will need a neutral-bond connection in the critical loads panel because it will then be your main panel.
Last edited:
The multiple neutral path is the issue I was trying to solve in my diagram by jumping the in and out neutrals. Multiple neutral paths are a much less serious issue than objectionable current and they are probably acceptable since MPP does not approve of jumping the input and output neutrals. I wonder if they know that their wiring diagram does create multiple neutral paths.
raurre
Solar Enthusiast
These units are well designed when used appropriately. That is why I purchased this model it is designed for split phase 120/240. It would seem that any electrical engineer would account for this. That is why I have chosen the wiring diagram in the my previous post because there will be no alternant paths back to the source for the EGC. The only time the EGC would carry current should be very briefly in the event of a ground fault. This should result in the Inverter going into a fault and shutting down or a overcurrent device breaking the circuit.
The most important issue is that if both inverters bond the neutral to ground when in battery mode, there will be neutral current flowing on the EGC and metal parts will be energized as diagrammed below. This will happen during normal battery mode operation. You should be able to determine if there is current on the EGC by using a clamp meter. If the inverters are smart enough to have only one inverter bond the neutral to ground, then everything is OK but other posts indicate that is not the case. Will posted that both inverters do the bond. Maybe he could check the EGC for current.
Orange highlight is line current from the top inverter in battery mode to the load. Yellow highlight is normal neutral current from the load back to the top inverter. Purple highlight shows the objectionable current. It travels on the neutral wire to the bottom inverter and then crosses to the EGC via the neutral-ground bond in the bottom inverter. From there it has multiple paths on the EGC back to the top inverter's neutral. This objectionable current can be eliminated by removing the bonding screw in the bottom inverter so that there is only one neutral-ground bond in the system.
It would also be interesting to test for current on the EGC with the bottom screw removed since there is a ground loop because the inverters have an EGC connection on both input and output. If there is current on the EGC even with the bottom screw removed, the solution is to follow FilterGuy's suggestion to remove the EGC connections on the inverter outputs (but you still need to remove the bottom screw).
It doesn't matter what MPP or Growatt says wiring if they don't know US code. Mike Holt pic and link below. These aren't grid tie inverters. Change the word generator for inverter.
Link to article 250 on grounding
chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/viewer.html?pdfurl=https%3A%2F%2Fwww.mikeholt.com%2Finstructor2%2Fimg%2Fproduct%2Fpdf%2F17NCT2-1475-sample.pdf&clen=1701441&chunk=true
Link to article 250 on grounding
chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/viewer.html?pdfurl=https%3A%2F%2Fwww.mikeholt.com%2Finstructor2%2Fimg%2Fproduct%2Fpdf%2F17NCT2-1475-sample.pdf&clen=1701441&chunk=true
Attachments
Zwy
Solar Wizard
Everyone keeps showing diagrams with the transfer switch after the inverters. Putting a 3 pole ahead of the inverters between the main panel with N-G bond and inverters eliminates many problems with loops and the N-G bonding issue. Only pass thru in the 3 pole would be the EGC, neutral would be switched in the transfer switch to allow bypass and the inverter will switch neutral with the internal transfer switch when in inverter mode.
Edit to add: EGC should run directly to panel after the inverters. Each inverter output EGC runs to that panel. No need to have EGC on AC input side of inverters.
Last edited:
Zwy
Solar Wizard
Not really. Forget any AC input for now and look at this as off grid. With 2 units in parallel, both run a neutral to a panel after the inverters where neutral is attached to the neutral busbar which is not N-G bonded. All neutrals from the system are attached to this busbar.
First, current returns to the source. Source is the inverter. The phase unbalance current on neutral will return to source, the inverter that is outputting the higher current due to 120V load. It won't return to the other inverter.
The danger is that if both inverters bond neutral to ground in battery mode, there will be objectionable current flowing in the EGC (green ground wire) as shown in the diagram below.
Neutral-ground bond is shown in light blue. Orange highlight is line current from the top inverter in battery mode to the load (I did not draw the load). Yellow highlight is normal neutral current from the load back to the top inverter. Purple highlight shows the objectionable current with black arrows showing the flow. It travels on the neutral wire to the bottom inverter and then crosses to the EGC via the neutral-ground bond in the bottom inverter. From there it flows on the EGC back to the top inverter's neutral. This objectionable current can be eliminated by removing the bonding screw in the bottom inverter so that there is only one neutral-ground bond in the system.
I believe there should be a grounding electrode (GEC) in this configuration.
Here is the diagram without the highlighting.
With AC input in bypass mode there will be parallel neutral paths as shown in the diagram below. This is not something you are supposed to have but I don't think it is really dangerous in this case and the only way I can think of to get rid of it involves connecting the inverter input neutral to its output neutral, which is not supported by MPP. So, we will probably just have to live with the parallel neutral current.
Orange highlight is line current from the grid to the load (I did not draw the load). Yellow highlight is normal neutral current from the load back to the grid. Purple highlight shows the parallel neutral current back to the grid.
Here is the diagram without the highlighting and also showing the battery.
Attachments
Zwy
Solar Wizard
No, it won't flow on the EGC as there isn't any bonding after the inverter. It can't flow where there isn't a connection.
Incorrect, current flows from source back to source. Alternate or parallel paths or objectionable current are cases where current has a path back from the intended path. The balance current on one phase will return to it's source.
You are looking at this incorrectly. The neutral runs back to source, the transformer on the pole. This is not a parallel path as both inverter transfer switches are tied to the neutral in the main panel. All you are doing is using 2 wires instead of 1. No different than using multi-stranded wire compared to solid wire.
When discussing parallel paths or objectionable current, what is being referred to is neutral current traveling outside the intended path, usually on the EGC.
raurre
Solar Enthusiast
This is the schematic of the wiring I currently have in place. This is the setup it is being run on. I have run it in all modes and there is no current on the ECG. ( I have followed this thread after I made this statement and I have found a flaw in where I had placed my amp meter. I tested the ECG connection between the transfer switch and the sub panel. There is not current there. But in inverting mode there is current on the ECG between the inverts and the transfer switch. This was my fault for not being thorough in my testing. I hope this helps to clear this up. ) 
Last edited:
Zwy
Solar Wizard
I'd rather run the 3 pole double throw before the inverter AC input with a main panel before it with N-G bond. This breaks the N-G bond as neutral is switched at the inverter transfer switch and N-G bond is at the inverter. This makes it a true separately derived system.
That's good. It may mean that the inverters are smart enough to bond only one place when more than one inverter is used, which is what we are trying to figure out. When you tested in battery mode, were you using a 120V load and was current flowing on the corresponding neutral?
FilterGuy
Solar Engineering Consultant - EG4 and Consumers
I am afraid @mccljs is correct.
The interesting thing is that the normal current will 'only' be the imbalance. Consequently, if you have a single 240V load, there will be little or no objectionable current. However, systems are rarely perfectly balanced. In the drawing below, there is only one 120V bulb for a load so it is imbalanced.
The purple dashed line is the normal flow. The yellow dashed it the objectionable current.
Exactly. And the two bonds create a parallel path for ground current as shown above.
FilterGuy
Solar Engineering Consultant - EG4 and Consumers
That system will have an objectionable current... But it might measure out as a small current..
The 'normal' current on neutral will be the difference between the two hots and the 'normal' path will take the neutral current to the inverter that is supplying the greater current. However, because of the dual bonds and ground loops in that wiring, the current will be split among many paths. With all the ground loops, I count at least 3 separate paths for the objectionable current as well as the normal path. As a first approximation, the objectionable current on any one of the paths will be the differential current divided by 4.
It would be good if we could get someone to test this in a completely off-grid configuration like the diagram below so that there would be only one objectionable path that will carry 1/2 the current and be easy to detect. I should have made it clearer that the load in the diagram is a 120V load and that the test should be done with only a single 120V load (no other loads 120V or 240V).
Zwy
Solar Wizard
You must be confused, you show a completely different diagram and not the one being referred to.
FilterGuy
Solar Engineering Consultant - EG4 and Consumers
OK.... Sorry. Is this the diagram?
That one is confusing because the 'highlighted' current does not go through a load.
It is correct to say that there would be no objectionable current with a single 240 V load (or any perfectly balanced load).
However, with any imbalanced load there will be an objectionable current. (Unbalanced loads is the typical case if you have 120V loads)
The primary differences between the drawing I showed and the one above is that there is no Grid on the AC input and no xfer switch.
If we look at the same structure with an imbalanced load or just a single 120V load, we see that there is still an objectionable current path through the neutral of the Slave inverter onto the EGC circuit..
The purple dashed is the 'normal' current.
The Yellow Dashed is the 'objectionable' current.
Yours is a much better diagram because it shows the interior of the inverters and an explicit 120V load instead of the implied one that I had.
There is an interesting discussion in this thread MPP LV6548 Ground/Neutral Safety?
