Power-One solar Inverter driving me NUTS!!!

[mitsu2k]

Hello folks, this is a last ditch effort. I thought I would seek your expertise before giving up.. I have been working on this thing for over 2 months and have gotten nowhere. Here is the issue:
Inverter powers on and immediately red GFI light comes on and inverter will not complete the initialization process.
There is no error code on the screen. I have removed most of the Y capacitors and tested to no avail.
All relays are good.
I replaced the current transducer- same problem. Checked most of the transistors and diodes and all appear good. I wish I could get my hands on a schematic which might point me to the GFI circuit.

I just don't know what else to look at here. I have hit a brick wall. Attached is a photo of the board. Anyone have any ideas?

The Inverter is a Power-One Aurora 3.6. I am sure the issue is not with the panels because I did power it up with a 100VDC power supply and had the same issue
.

[CapLeaker]

Re: Power-One solar Inverter driving me NUTS!!!

GFI is the ground fault protection on the solar array. There are multiple reasons why the inverter is kicking off and some of them aren’t even the inverters fault.

We need to know more on what you tested and how. For example, did you check your solar panels and wiring? Maybe there is arcing? It’s not an easy test to do where you rip out your DMM and go for it. Does the inverter error out by just connecting it in the grid, nothing else.
Other than that, if she errors out in ground fault and doesn’t give you an error code, it’s not looking good. I do see a GRP pin… where does it go on the other board?

[mitsu2k]

Re: Power-One solar Inverter driving me NUTS!!!

I am pretty sure the problem is not with the panels as I get the same error without the panel array connected and just using a 109v dc power supply to power it on. I had also tested the panels with an insulation resistance meter and they tested fine with super high resistance. Over 50 mega ohms. The inverter itself tests the insulation resistance of the panels and it comes back 20 mega ohms, which it should be.

Like I mentioned, I removed and tested all the caps. I also removed all the Y caps and fired it up without the caps installed. I checked all the relays. Also checked most of the transistors. I visually checked the MOVs and also removed them completely. I checked for cold solder joints.
As for the GRP pins, they go to another board inside the inverter. That board is connected to the transformers and has 4 large capacitors as well as some large transistors and some other components. It's a much smaller board than this one and I have not messed with it because it is going to be a pain to remove due to compound and heat sink situation. I have the inverter assembled right now but I can take it back apart and try to get some photos of the other board. What is GRP?

By the way, I tested the caps with my Fluke 87 as well as my Peak ESR meter.

[mitsu2k]

Re: Power-One solar Inverter driving me NUTS!!!

Here are some more photos. GRP is circled in one of the photos. There are a total of 3 of those rectangular protection caps on this other board. At least I am assuming they are protection caps.



By the way, Inverter does error out even when not connected to grid.

[mitsu2k]

Re: Power-One solar Inverter driving me NUTS!!!

Here are some better photos of the other board.

[CapLeaker]

Re: Power-One solar Inverter driving me NUTS!!!

What happens if you wire the inverter to the grid and don't connect anything like panels? I assume GRP is the ground protection. I saw various GRP pins or points that always have a rectangular blue cap or similar connected to ground. I would follow these around.
A ground fault can be on the DC side, but also a ground fault can be on the AC side where the inverter connects to the grid. If something goes stupid with the DC side, there is always an error code. Since you don't have any, maybe the fault is somewhere else, like on the AC side.
Your test gear is ok and measurements too, so I agree with the inverter being the problem. The problem could be as stupid as a shorted MLCC SMD cap too.
I marked up your first picture with things I'd be looking at so you have an idea what components I mean. There are also a bunch of little neon lights…
Anyway, the problem may not lie on this board.

[mitsu2k]

Re: Power-One solar Inverter driving me NUTS!!!

Thank you! I would first like to start by answering your question about what happens if I connect just to the grid and no panels. The result will be that the inverter will not power on at all. This one requires DC power in order to function or even turn on. I will go through your suggestions this evening and report back. I so much appreciate your help. I’ll continue to update you.

[CapLeaker]

Re: Power-One solar Inverter driving me NUTS!!!

On a second thought:
It baffles me that the inverter doesn’t continue to finish the initialization process. What I think it doesn’t do in order to protect itself. So there could be something like a bad mosfet, diode or something along that line on that other board.
What I wouldn’t do, is making the unit “brain dead” as it can result in nice loud BANG and some fire flying. Don’t ask how I found out.
Anyway, I hope you got some new ideas.

[petehall347]

Re: Power-One solar Inverter driving me NUTS!!!

as a sanity check take a look at the mains supply .
what is the model number of the unit in question ?

[mitsu2k]

Re: Power-One solar Inverter driving me NUTS!!!

On a second thought:
It baffles me that the inverter doesn't continue to finish the initialization process. What I think it doesn't do in order to protect itself. So there could be something like a bad mosfet, diode or something along that line on that other board.
What I wouldn't do, is making the unit “brain dead” as it can result in nice loud BANG and some fire flying. Don't ask how I found out.
Anyway, I hope you got some new ideas.

I tend to agree with you that it is not an actual ground fault. I believe it's something else like you said for the same reasons that you mentioned. As soon as it powers on instantly, I hear the relays click, and at the exact same time of the relays click , I get the GFI light. So it is exactly like you said. It doesn't have time to do it start up diagnosis. So I believe it is an erroneous error for that reason as well as the reason of it not giving the actual error code on the display.

[mitsu2k]

Re: Power-One solar Inverter driving me NUTS!!!

as a sanity check take a look at the mains supply .
what is the model number of the unit in question ?

Here is a photo of the model of the unit.

it is PVI 3.6-OUTD-S-US


Thank you

[CapLeaker]

Re: Power-One solar Inverter driving me NUTS!!!

apparently a

-No Code
-Ground Fault
-Red LED

means a leakage current to ground is detected in the DC section of the system.

what happens if you measure resistance from the solar panel dc inputs (both positive and negative) to the inverter earth ground terminal? Maybe put 100V or whatever the inverter has max DCV reading with the Megger to it.

[mitsu2k]

Re: Power-One solar Inverter driving me NUTS!!!

Ok but the display does not say “ground fault”. Just red LED and no code.

I am going to borrow a megger from a friend and try that. I think that’s a great idea. This might at least give me an indication if it is a real ground fault or not.

[petehall347]

Re: Power-One solar Inverter driving me NUTS!!!

have you tried a reset by holding down ESC for 3 seconds ?

[CapLeaker]

Re: Power-One solar Inverter driving me NUTS!!!

The display is dead? Does it come to life at all?

I’ve got a 10kw inverter and so far within ten years it kicked off only once with an arc fault. When my unit went into this ground fault all I had to do was to turn off DC power, turn off AC power, wait a minute, then turn on AC power, turn on DC power and all was good.

As for your unit, either there is a really something leaking over from DC to earth ground or the whole sensing circuit is fubar.

BTW... if you do put high volts DC into one of the MPPT inputs on the inverter to power it up, did you use an isolation transformer on the HVDC PSU?

[mitsu2k]

Re: Power-One solar Inverter driving me NUTS!!!

have you tried a reset by holding down ESC for 3 seconds ?

Yes I have tried that. No go

[mitsu2k]

Re: Power-One solar Inverter driving me NUTS!!!

The display is dead? Does it come to life at all?

I’ve got a 10kw inverter and so far within ten years it kicked off only once with an arc fault. When my unit went into this ground fault all I had to do was to turn off DC power, turn off AC power, wait a minute, then turn on AC power, turn on DC power and all was good.

As for your unit, either there is a really something leaking over from DC to earth ground or the whole sensing circuit is fubar.

BTW... if you do put high volts DC into one of the MPPT inputs on the inverter to power it up, did you use an isolation transformer on the HVDC PSU?

The display does light up but immediately the red GFI LED lights also. It says “initialization” and goes through all its initial measurements as far as “Vgrid, Fgrid, Riso, etc” everything comes back as the correct values. But it just loops because of the GFI LED error. I dis not use an isolation transformer. I just powered it on with to 60v DC power supplies wired in series.

To be honest, I think the GFI LED comes on right before the display even displays anything.

[mitsu2k]

Re: Power-One solar Inverter driving me NUTS!!!

By the way, when I powered it on with my DC power supplies, I did not connect the AC/grid connection. Just the DC connections. So I don’t believe that an isolation transformer would have been necessary under the circumstances but I could be wrong.

[stj]

Re: Power-One solar Inverter driving me NUTS!!!

make sure all those big MOV's are open-circuit.
they do develop resistance as they age.

[mitsu2k]

Re: Power-One solar Inverter driving me NUTS!!!

make sure all those big MOV's are open-circuit.
they do develop resistance as they age.

Yes they are open circuit. Removed and checked all of them. The four red ones on the DC side and the two blue ones on the A/C-Grid side.