Power-One solar Inverter driving me NUTS!!!

[truclacicr]

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

PVI-3.0 3.6 4.2-OUTD-US Manual

https://powertripenergy.com/wp-conte...4.2-manual.pdf
https://ressupply.com/documents/abb/...-US_Manual.pdf

CASR series Current transducer

https://www.lem.com/sites/default/fi...asr_series.pdf

LEM's product range

https://www.lem.com/en/product-list?keys=

The red 'GFI' (ground fault) LED indicates that AURORA Inverter is detecting a ground fault in the DC side of the photovoltaic system. When this kind of fault is detected, the AURORA Inverter disconnects from the grid and the corresponding fault indication appears on the LCD display. AURORA Inverter remains in this condition until the operator presses the ESC key to re-start the grid connection sequence. If pressing the ESC key doesn’t clear the ground fault check the ground-fault, fuse located in the switchbox. If AURORA Inverter does not reconnect to the grid, contact Power-One Technical Service.

As I see it, we should be looking for a fault on the PV side.

I believe that GRP is Photovoltaic GRound.

The CASR/25/SP3 transformers are measuring the current in one leg of each PV circuit. Ground fault detection requires that the differential current be sensed, ie we must sense the current in both legs. Therefore we must look for a detector on the PCB that has the bulk capacitors. This detector must be connected to BOOST1, BOOST2 and GRP.

The CT...SP4 transformer appears to be sensing the AC current, so I don't think we should be looking here.

[mitsu2k]

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

I went over these pictures again and maybe I found something or not. LOL
If you look at that spot I marked, it looks like a dark area (like a light burn mark) compared to the others.

Take your DMM in diode mode and check the 3 legs of all the IGBTs and diodes on this board. See if they are shorted or leaky.
How to IGBT? Here: https://www.incbtech.com/circuit-ben...s-ins045e.html

Great, I will do that as soon as I get home. Thank you.

[mitsu2k]

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

PVI-3.0 3.6 4.2-OUTD-US Manual

https://powertripenergy.com/wp-conte...4.2-manual.pdf
https://ressupply.com/documents/abb/...-US_Manual.pdf

CASR series Current transducer

https://www.lem.com/sites/default/fi...asr_series.pdf

LEM's product range

https://www.lem.com/en/product-list?keys=



As I see it, we should be looking for a fault on the PV side.

I believe that GRP is Photovoltaic GRound.

The CASR/25/SP3 transformers are measuring the current in one leg of each PV circuit. Ground fault detection requires that the differential current be sensed, ie we must sense the current in both legs. Therefore we must look for a detector on the PCB that has the bulk capacitors. This detector must be connected to BOOST1, BOOST2 and GRP.

The CT...SP4 transformer appears to be sensing the AC current, so I don't think we should be looking here.

Thank you, so I can understand, would we be looking for a leaky cap or a bad CASR/25/SP3? When you say "Bulk Capacitors" do you have an idea of which ones those might be? I can take more and better photos if need be.
These caps in the photo are connected to both boosts and GRP. I have never removed these caps but I did check them and they did read the 30 uF as expected. However, they could certainly be leaky. They did test OK with my ESR meter but I know that is not a proper test for leakage.

[mitsu2k]

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

I went over these pictures again and maybe I found something or not. LOL
If you look at that spot I marked, it looks like a dark area (like a light burn mark) compared to the others.

Take your DMM in diode mode and check the 3 legs of all the IGBTs and diodes on this board. See if they are shorted or leaky.
How to IGBT? Here: https://www.incbtech.com/circuit-ben...s-ins045e.html

I also need to mention that this area circled showed signs of heat as well. All that rubber silicone coating stuff was really really brown in this area and especially underneath that blue relay. I cleaned it up and checked all the components in that area such as diode, there was a couple vets, a PNP, and NPN, and some resistors as well as the relay itself although they showed significant signs of heat, they all tested just fine. I actually remove them and tested them so it's to avoid any misreadings from in circuit testing. Not sure why this area was getting hot. I wish I would have taken a photo before hand, but I got ahead of myself and cleaned it up.
I found a phot with that blue relay removed and photo taken during the cleanup process.

[CapLeaker]

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

I think we are on the right track. Now its to figure out on what board the problem is. Then go from there.
Leakage could be on FET's, diodes, IGBT's, some burnt crap somewhere, a chaffed wire etc.

[truclacicr]

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

Look at page 4 of this document:

https://cdn.badcaps-static.com/pdfs/...95a1307f0b.pdf

Notice that the positive and negative PV conductors pass through a detection device that senses the net current and produces a GFCI interrupt to the controller if the sensed net current is non-zero. That's the way that an RCD / ELCB works.

[truclacicr]

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

That Meder device appears to be a 12V reed relay.

https://cdn.badcaps-static.com/pdfs/...147bd57807.pdf

[mitsu2k]

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

Look at page 4 of this document:

https://cdn.badcaps-static.com/pdfs/...95a1307f0b.pdf

Notice that the positive and negative PV conductors pass through a detection device that senses the net current and produces a GFCI interrupt to the controller if the sensed net current is non-zero. That's the way that an RCD / ELCB works.

Can you please elaborate on this? I am not as professional and experienced as you. If you could do a markup in the photo that would be great.

I think you are talking about the current sensors on the left hand side back of the board on the PV side of the board.

[mitsu2k]

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

That Meder device appears to be a 12V reed relay.

https://cdn.badcaps-static.com/pdfs/...147bd57807.pdf

Yes you are correct. I did remove this relay and check it. It is good. I also soldered some wires to the pins/legs
and extended them out of the unit so I can check if this relay is getting the 12 V power when the inverter is powered on. It is not, and I suspect that it should if I’m not mistaken.

I simply checked it by applying 12 V DC to it and checking if the contacts closed and they did. I also checked all the small components adjacent to it and they all seem to test good. I even removed all of the SMD mosfets and transistors and tested them outside of the board with my transistor tester and they all tested good. Just for the heck of it, I had ordered new ones anyway and I put the new ones in but it didn’t seem to solve anything so I put the old ones right back

[mitsu2k]

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

Look at page 4 of this document:

https://cdn.badcaps-static.com/pdfs/...95a1307f0b.pdf

Notice that the positive and negative PV conductors pass through a detection device that senses the net current and produces a GFCI interrupt to the controller if the sensed net current is non-zero. That's the way that an RCD / ELCB works.

This link will not open for me for some reason.

edit: scratch that, it just opened.

[truclacicr]

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

The block diagram on page 4 of that other inverter is showing that there is a device which is sensing the total current flowing through that loop. Let's call it a current transformer.

Let's say panel #1 is producing 100A and panel #2 is producing 120A. This means that the total current flowing through the loop is 100A - 100A + 120A - 120A = 0. This produces zero flux in the transformer, so the output of the secondary winding is 0.

Now let's imagine that current is leaking to building ground from one of the panels. The current through the loop might now be 100A - 99.9A + 120A - 120A = 100mA. There is now some magnetic flux in the transformer, which induces a current in the secondary, which triggers the GFCI.

In your case the respective conductors are BOOST1, BOOST2 and GRP (combined negative outputs from both panels). You need to find some component through which all these currents pass. At least that's how I see it, otherwise I'm stumped, too.

Note that my discussion is relevant for AC current transformers, but DC current transformers work a little differently. They typically use Hall effect sensors rather than a secondary winding.

[mitsu2k]

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

The block diagram on page 4 of that other inverter is showing that there is a device which is sensing the total current flowing through that loop. Let's call it a current transformer.

Let's say panel #1 is producing 100A and panel #2 is producing 120A. This means that the total current flowing through the loop is 100A - 100A + 120A - 120A = 0. This produces zero flux in the transformer, so the output of the secondary winding is 0.

Now let's imagine that current is leaking to building ground from one of the panels. The current through the loop might now be 100A - 99.9A + 120A - 120A = 100mA. There is now some magnetic flux in the transformer, which induces a current in the secondary, which triggers the GFCI.

In your case the respective conductors are BOOST1, BOOST2 and GRP (combined negative outputs from both panels). You need to find some component through which all these currents pass. At least that's how I see it, otherwise I'm stumped, too.

ok makes sense. My brain is still trying to comprehend this information but I am sure within a few hours of processing it will make sense to me and then I will have a new "set of eyeglasses" so to speak, which will allow me to take a clearer and more educated look.

On a second thought, does this mean the megger might not help me since the leakage could be as small as 100mA?

I see the two CASR/25/SP3 devices that are current transducers on the boosts. Could these be the detection devices? If so, what if I swap them with the two on the AC side to see if it makes a difference or would that be a moot point?

[CapLeaker]

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

If there is any path, the megger will show it.

[truclacicr]

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

I see the two CASR/25/SP3 devices that are current transducers on the boosts. Could these be the detection devices?

AFAICT, those devices only sense the current in the BOOST leg. There is nothing on that PCB which senses the current in the GRP leg.

[mitsu2k]

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

If there is any path, the megger will show it.

I got the megger and I am getting some interesting results. However, I just got called out on a job. I’m gonna try to finish it up really quick and come back and continue the testing then I will report back the full results. Thank you guys so much.

[mitsu2k]

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

I’m actually driving right now but I don’t wanna keep anyone in suspense so I will give you just some preliminary findings for now. When I tested the large board with the mega after about one minute I got about 2.4 mega ohms both on the positive and the negative. However, those numbers continued to climb and after about five minutes they were in the 40 megaohm range. When I connected the other board, with the floor, large capacitors the readings seemed to stay locked at a little over one megaohm. I still have to do some more testing when I get home but what do you think of these results so far?

[CapLeaker]

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

That’s what I thought is going to happen. First board you get the rising MOhms and with the second board you get the shitty 1Meg reading. Keep going!

[mitsu2k]

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

Test results with second board completely disconnected. Testing form Pos/Neg DC input terminals to ground. Both sets of inputs seemed to yield same results so I will just list the results

Positive terminal
Initial reading at test start - 1.6 megaohms
ONE MIN - 2.4 megaohms
THREE MIN - 3.1 megaohms
FIVE MIN - 3.2 megaohms (seemed to top out at this reading)

Negative terminal
Initial reading at test start - 1.6 megaohms
ONE MIN - 2.8 megaohms
THREE MIN - 8.7 megaohms
FIVE MIN - 24 megaohms (continues to climb but I stopped at 5 min)




TEST with other board connected.

1.0 to 1.1 megaOhms on both positive and negative. Test for one minute

When unplugging GRP, Boost 1, Boost 2, and BULK the readings jump up to 10 megaohms or so. So either this board or the transformers seem to be the culprit.

[mitsu2k]

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

I went over these pictures again and maybe I found something or not. LOL
If you look at that spot I marked, it looks like a dark area (like a light burn mark) compared to the others.

Take your DMM in diode mode and check the 3 legs of all the IGBTs and diodes on this board. See if they are shorted or leaky.
How to IGBT? Here: https://www.incbtech.com/circuit-ben...s-ins045e.html

I did check those large IGBTs for shorts and I found none. As for “leaky”, I don't know how to check for that. Also could not find any shorts on the few SMDs that I checked but I was not thorough. Will do it again. What about those 3 blue caps? How can I check them for leakage? Will the megger do it?

[mitsu2k]

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

This other board is at least a 4 layer PCB. Very hard to trace things. I don't know how to accurately test the blue caps in the attached photo. I did put the megger to them and they came back over 4000 megaOhms. I also checked as many of the SMD components as I could, and I could not find any short. The seven LGBT looking devices have absolutely no markings that I can see on them. I did test them for shorts and could not find a short, but that's the extent of the testing that I was able to do. Those heat sink clamps are very difficult to get off and I haven't attempted to do that yet. Maybe there's a marking underneath them but I'm not sure. Anyhow, I have taken some better close-up photos of that board and attach them.