Power-One solar Inverter driving me NUTS!!!

[mitsu2k]

When you say “ground” do you mean earth ground or another ground?

[ssrkh]

u r welcome dear simple jumper circuit ground to the CT output pin that will bypass the sensing of the current and ampare will start vary but yoiu can check after that the system is vcompleting the initialization process

[mitsu2k]

Dead end again, replaced the MOVs with no change. At least I have some fresh ones in there now in case I ever do get this thing working. I am very interested in what ssrkh said above in post 160 about jumpering the output of the CT devices to ground. I see the output pin but I am not sure what ground he is talking about. i.e. earth ground or another pin on the CT? Does anyone else have an idea?

[mitsu2k]

Then our Megger test should have shown that. And maybe it did, because one side on DC had a higher MOhms reading than the other, but cording to spec this reading was supposed to be acceptable and to me it's not.
Maybe take the CT off the DC and redo the insulation test. If it changes, your hopefully be golden afterwards.

Hi CapLeaker, I'm not sure I understand. My understanding of the CT devices is that they are just measuring current and reporting back to the controller. It seems that if I remove them and test with the megger again, the readings may change because there is no pathway going through them to the rest of the circuit. Road will be blocked so to speak. So I'm not sure what this would point me to. Can you please help me to understand?

[CapLeaker]

Then our Megger test should have shown that. And maybe it did, because one side on DC had a higher MOhms reading than the other, but cording to spec this reading was supposed to be acceptable and to me it's not.
Maybe take the CT off the DC and redo the insulation test. If it changes, your hopefully be golden afterwards.

[mitsu2k]

Hi ssrkh, thank you for chiming in. Couple questions if I may please...
There are two CT devices on the DC side. There are another two on the AC side. I suspect the error is on the DC side because it happens even with no AC connected. Of course there is another CT which I have already replaced so I will not bother you with that one. Attached are some photos of these CASR CT devices and the datasheet is here: LEM Catalogue CAS_CASR_CKSR_09

Can you please tell me which pins I need to jumper? Also when you say "ground", are you referring to earth ground? Just want to make sure I do it right. Thank you!

[ssrkh]

listen you guys most of the times Ground fault comes with the CT what u need to do is just bypass CT by jumpering ground to Output of the CT so u will know where the fault is comming from and after that if u have again same fault of not performing initialization completely then watchdog is getting the bit High from the EC embedded controller so replace the controller fault will gone thanks me later bye

[mitsu2k]

I have a question about those 4 MOVs. When I measure voltage between positive and earth ground, I get about 145V. When I remove the MOVs and measure without them, the voltage increases to around 194V. So the MOVs are suppressing some voltage. I am wondering if these MOVs are getting weak and they should be suppressing more. I did find that the middle leg of the MOVs of the positive input is connected directly to earth ground. On the negative the middle leg is connected directly to the GRP terminals that goes to that second board. I have went ahead and ordered some new MOVs and I should have them tomorrow. So I guess time will tell if this is the problem. The thing that strikes me odd is the blue MOVs on the AC side both test greater than 4000 MegaOhms with my megger. But these red ones are far below that. I know this is not a proper test but the specs of these MOVs are similar, at least close enough so I would think the results shouldn't be that far off. Just my crazy brain thinking without the proper expertise and experience. What if these MOVs are actually there for GFCI purposes rather than surge? Any thoughts?

[mitsu2k]

so it looks like gfi coincides with the 4th relay then . what is that relay for ?
maybe set your meter on megohms and check between the ground and around that 4th relay looking for leakage on the board . it does look as if the inverter is seeing leakage from somewhere .

the 4 relays are on the AC side. During normal operation, you can hear the relays cycling on/off.

[petehall347]

so it looks like gfi coincides with the 4th relay then . what is that relay for ?
maybe set your meter on megohms and check between the ground and around that 4th relay looking for leakage on the board . it does look as if the inverter is seeing leakage from somewhere .

[mitsu2k]

what is the history of this thing ? like did it work then suddenly this fault ? any other factors that may have happened just prior to this particular fault ?

Pete, the history of this unit is as follows. The inverter was probably installed around 2014 based on the date code. About 3 years ago it got the E031 error which is where the relay contacts on the PCB arc and burn out. At that time I repaired it by reflowing the solder joints on the relay in question. about a year later same thing happened but on a different relay. I repaired it same way. Several months later, it happened a third time and again on a different relay. This time I decided to reinforce the traces using some copper wire to allow more current to flow so the contacts would not arc and burn out. This seemed to work very well however I neglected to add the copper wire on one of the relay contacts. a couple months after that, I noticed that the inverter was not working and the red GFI light was on. So again I opened it up and found that the one relay contact that I did not reinforce with wire had burned out. However, this time that relay was melted pretty good and was not functional. I replaced all 4 relays at that time and reinforced all the pads/contacts with wire. Reinstalled it and the GFI light did not go away. So here we are today.

I would like to add that the very last time I mentioned above, there were no error messages on the display. The 3 previous times I had the E031 error on the display and the alarm led was lit but no GFI light.

[mitsu2k]

I did notice that there is a switch on your inverter, next to the AC connection. That is to pair or not to pair them DC inputs. I think the switch isn't set for them 2 DC ports as paired. Also there support to be a jumper between the 2 DC ports going from positive to positive and negative to negative. Putting the inverter on the DC port into paired, supposed to handle more current or something. Can't remember. Got to read that manual again.

CapLeaker, The "PAR' switch according to my understanding of the manual is used to parallel more inverters together.

[mitsu2k]

I would mega ohm the cables from the solar panels to the inverter with out the cable being attached to the solar panels or the inverter especially if the positive and negative cables are connected to together as one cable run and or are in conduit because I ran into a situation where water had introduced into the installation of wires and were shorting out in the conduit so do not rule this one out I use a mega ohm tester to figure out why this device was blowing fuses ( how you test for this is you have one meter lead on positive wire and the other meter lead on the negative wire )

On the same note are you using outside rated cable that you are using for the wiring from the solar panels to the inverter if not you should be and should be outside rated conduit or if I was doing this I would be doing it this way

Your results should be the highest mega ohm reading that the meter ohm meter has and any thing else would be suspicious at least to me it would be or you have something wrong with inverter showing a voltage leaking situation

SAM, this is good thinking about wires shorting out inside the conduit. I do not think this will be the case but I will check it at night. It may shed light on that small transient voltage either way.. They are coming through my attic in a flex conduit. 12awg I believe. I have not seen any signs of water or moisture when I went up in the attic. The panels were installed by the builder or builder's contractor when the house was built so the wiring should all be the proper type there..

[petehall347]

what is the history of this thing ? like did it work then suddenly this fault ? any other factors that may have happened just prior to this particular fault ?

[CapLeaker]

I did notice that there is a switch on your inverter, next to the AC connection. That is to pair or not to pair them DC inputs. I think the switch isn't set for them 2 DC ports as paired. Also there support to be a jumper between the 2 DC ports going from positive to positive and negative to negative. Putting the inverter on the DC port into paired, supposed to handle more current or something. Can't remember. Got to read that manual again.

[sam_sam_sam]

I would mega ohm the cables from the solar panels to the inverter with out the cable being attached to the solar panels or the inverter especially if the positive and negative cables are connected to together as one cable run and or are in conduit because I ran into a situation where water had introduced into the installation of wires and were shorting out in the conduit so do not rule this one out I use a mega ohm tester to figure out why this device was blowing fuses ( how you test for this is you have one meter lead on positive wire and the other meter lead on the negative wire )

On the same note are you using outside rated cable that you are using for the wiring from the solar panels to the inverter if not you should be and should be outside rated conduit or if I was doing this I would be doing it this way

Your results should be the highest mega ohm reading that the meter ohm meter has and any thing else would be suspicious at least to me it would be or you have something wrong with inverter showing a voltage leaking situation

[mitsu2k]

Yours isn't an A model. It doesn't have that fat ass clamp contraption. When you hook that inverter back up to test, you put one or both strings to it? Are you using both DC ports? The solar panels face all in the same direction?

I only have one string. All panels face the same direction. I am using only one DC port but I have tried connecting to the second one as well, and got the same results

[CapLeaker]

Yours isn't an A model. It doesn't have that fat ass clamp contraption. When you hook that inverter back up to test, you put one or both strings to it? Are you using both DC ports? The solar panels face all in the same direction?

[mitsu2k]

Ok I did the test outlined in the manual. (Positive shorted to Negative) at 250v on the megger to earth ground.

result= >4000 MegaOhms on the megger (full scale). All good here.

[mitsu2k]

from what i can make out its behaving correctly its only the "A" models that dont halt at this fault . unless yours is an "A" model

HI Pete, I now know what you are talking about "A" models. Mine is not an "A" model. I found the manual that you must have been looking at. It appears that these have a built in Arc detection feature as well as some other features. For example, the "Autotest" feature is non existent in my inverter. I even looked under the service menu. I was able to find a key generator in order to access this menu.