6000xp with "PV isolation low" warning

[Kent86]

Hello,

I recently installed the two MIdnite solar SPD's on my 6000xp system, one for each MPPT (one for each string). I shut down the system, installed them and turned everything back on in the order recommended by the manual but the inverter kept making a clicking noise and the solar wouldn't activate in the inverter. The SPD's are the 300 volt DC models and my strings are well below the threshold. Then the inverter would show error code 20 "PV isolation low". I restarted the inverter multiple times with the same result. So I decided to disable the PV isolation setting and everything immediately cleared and started working as it should. The issue is- why would the PV isolation (GFCI capabilities, as I understand it) have to be turned off for the PV circuit to function with the SPD's? Is it possible that the SPD's are being seen as a ground fault?
As an electrician, I really liked the idea of the PV circuits having ground fault protection and want it enabled.

 

[Hedges]

What is the voltage of your PV string?

SPD do have capacitance. These HF AIO drive AC superimposed on PV wires. Maybe it is AC leakage that causes the fault.

I was able to test small MOV with a HiPot, but larger surge arrestors tripped the 1 mA threshold. Had to test them differently.

With SPD disconnected, measure voltage on AC range from PV+ to ground, also PV- to ground. Since you have SPD connected and isolation disabled, test it that way too.

Try putting AC ammeter between ground (neutral) wire of SPD and ground. That should measure leakage directly.

Ground fault of PV strings for transformer isolated inverters was a 1A fuse, but transformerless may be more sensitive. Utility really wouldn't like it if we fed DC, even 1A, into their transformers.

 

[Kent86]

string 1 is about 150 VOC and string 2 is 200 VOC.

Thanks for the advice, I'll check out out tomorrow and get back

 

[Hedges]

The 300V SPD should be fine as far as no DC leakage, so see what AC current there is due to capacitance.

Oh, AC voltage. But the 60Vrms 90Vpeak I'm guessing you'll see, plus 200VDC, is still less than 300V, shouldn't conduct with 450V MOV. Let's see what AC voltage you measure too.

https://www.midnitesolar.com/pdfs/spd_ryan_frontBack-spanish.pdf

The one I'm using on AC line clamps to 200V.

Testing Metal-Oxide Varistors (MOV)

Surge arrestors are used on AC and DC side of PV inverters for protection, to clamp high voltage transients. One inverter I have (actually several) are in the SMA family SUNNY BOY 5000-US / 6000-US / 7000-US / 8000-US. These come with a kit of MOV to protect the DC side: The MOVs are about...

diysolarforum.com

 

[RCinFLA]

Disconnect the PV array positive and negative lines and check their resistance to grounded aluminum frame.

Leakage at the edge of panel from backing plastic delamination at the edge allowing humidity to infiltrate is a common cause of GFCI tripping.

Should have greater than 200k ohms from pos and neg leads to metal frame of panel.

 

[EG4_Jared]

string 1 is about 150 VOC and string 2 is 200 VOC.

Thanks for the advice, I'll check out out tomorrow and get back

Additionally, with string 1 at 150v and string 2 at 200v, this is very close to the minimum operational range of the MPPTs at 120v. It's possible that the PV input was indeed low during the testing.

 

[Hedges]

Additionally, with string 1 at 150v and string 2 at 200v, this is very close to the minimum operational range of the MPPTs at 120v. It's possible that the PV input was indeed low during the testing.

"DC INPUT VOLTAGE RANGE 100-480 VDC"

https://eg4electronics.com/backend/wp-content/uploads/2023/10/EG4-6000XP-Inverter-Spec-Sheet.pdf

6000xp ... error code 20 "PV isolation low"

Did that show "PV isolation low", or "PV insolation low"?

Did the programmer mean to say "PV isolation low", or "PV insolation low"?

 

[EG4_Jared]

 

[Hedges]

string 1 is about 150 VOC and string 2 is 200 VOC.

Thanks for the advice, I'll check out out tomorrow and get back

Are the strings same orientation and same model panel?
If so, all in series for 350Voc would be good.
If not, you just need more panels

(to reach reliable operating voltage.)

Your point that turning off isolation testing makes it work indicates leakage issue, not low voltage.

 

[Kent86]

Are the strings same orientation and same model panel?
If so, all in series for 350Voc would be good.
If not, you just need more panels

(to reach reliable operating voltage.)

Your point that turning off isolation testing makes it work indicates leakage issue, not low voltage.

I have an east/ west array. East is seven 190 watt panels and west is six 390 watt panels.

 

[Kent86]

I really don't think it's a ground fault issue with the panels themselves, as the system was working fine with PV isolation turned off before the SPDs were installed

 

[Kent86]

I double checked the warning and it definitely says "PV isolation low"

 

[Kent86]

Also, the voltage rating of the Midnite solar SPD is 0-300 volts

 

[Kent86]

I really don't think it's a ground fault issue with the panels themselves, as the system was working fine with PV isolation turned off before the SPDs were installed

Meant to say "PV isolation turned on"

 

[Hedges]

I have an east/ west array. East is seven 190 watt panels and west is six 390 watt panels.

In hot weather, the lower voltage one might drop below MPPT minimum as EG4 Jared said. See if you can add a panel.

I double checked the warning and it definitely says "PV isolation low"

We expect Chinglish in the manuals, so why not in the firmware? Could have meant "insolation". But "voltage" more likely.

I really don't think it's a ground fault issue with the panels themselves, as the system was working fine with PV isolation turned off before the SPDs were installed

An experiment disconnecting SPD, with insolation test turned off, could help confirm. It takes multiple A/B/A/B tests to mostly rule out coincidence that some other issue cropped up at the same time.

Leakage current through capacitance of the SPD is my working theory. You can measure it, and Jared can maybe confirm threshold of test.

If capacitance is the issue, different SPD or different configuration might work. Possibly a 600V SPD has lower capacitance, can connect to PV+, PV-, GND without tripping. But it is twice as large, maybe be same capacitance and twice the voltage.

SPD strictly between PV+ and PV- without GND shouldn't trip. I think both black wires to PV+ and white wire to PV- would provide same voltage protection at twice the current.

Another 300V SPD with one black wire on PV-, other black wire on GND, white wire capped would be 2x voltage, 1/2 capacitance. That would provide some protection with half the AC leakage current.

A different brand SPD with less protection could also have less capacitance.

There may be other technology SPD besides MOV that have lower capacitance. We use things like spark gap and zener diodes for instruments (low power) because capacitance would reduce frequency bandwidth. Not sure which are best for PV.

PV panels themselves also have capacitance to ground. Try disconnecting one PV string at a time, test just the other with SPD. Maybe capacitance of larger PV array plus SPD is too much.