Optimisers for Sol-Ark other than Tigo?

[Hedges]

Arc fault may have a sensitivity adjustment.

I'm viewing TIGO monitoring to try and figure out why I'm getting arc faults. I'm seeing a strange phenomenon. 5 minutes before the Sol-Ark reports an arc fault the optimizers stop reporting data. I have 1 minute data resolution with the premium monitoring service..

Do you have a scope, ability to see what's happening on PV wires?
I think PLC is pulses of injected current, so pickup with a CT, or more likely a CT you make with suitable toroid, might be the way. (I've observed attenuated signal at 20 kHz when trying to measure inrush with CT; what frequency is PLC from Tigo?)

 

[GregTR]

There is no PLC from Tigo, Tigo's comms are on separate wire from CCA to TAP and wireless from TAP to the individual optimizers.

 

[Oldphile]

Has anyone brought this issue to Sol-Ark's attention?

I did and they just brushed it off as a wiring fault.

 

[Oldphile]

Arc fault may have a sensitivity adjustment.



Do you have a scope, ability to see what's happening on PV wires?
I think PLC is pulses of injected current, so pickup with a CT, or more likely a CT you make with suitable toroid, might be the way. (I've observed attenuated signal at 20 kHz when trying to measure inrush with CT; what frequency is PLC from Tigo?)

There's no sensitivity setting. I don't have a scope. As @GregTR said TIGO does not communicate on the PV wires when using a CCA/TAP. The TIGO RSS transmitter uses PLC, but that's a different product.

 

[Oldphile]

I have a support ticket opened with TIGO. I'll report back with the results.

 

[Oldphile]

Well, I didn't get anywhere with TIGO, they just say the modules are working correctly. Yesterday gave me some new insight. The TIGO optimizers went offline 12 times for a duration of 5 minutes +/- 1 minute, one of which the Sol-Ark reported as an arc fault. I turned arc fault off after the fault that occurred at 10:26:46 The faults are occurring and sometimes Sol-Ark recognizes them as a potential arc fault.

 

[zanydroid]

I did and they just brushed it off as a wiring fault.

Did either TIGO or SolArk have a suggestion on how to root cause this? The current situation leaves too much for false positive/false negative/loss of protection required by code.

If you want to you can try bisecting the system to see if a sub-segment of the string has a higher/lower probability of triggering a problem.

(There is still the possibility of a wiring fault IMO)

 

[Carlos_Sol-Ark]

Well, I didn't get anywhere with TIGO, they just say the modules are working correctly. Yesterday gave me some new insight. The TIGO optimizers went offline 12 times for a duration of 5 minutes +/- 1 minute, one of which the Sol-Ark reported as an arc fault. I turned arc fault off after the fault that occurred at 10:26:46 The faults are occurring and sometimes Sol-Ark recognizes them as a potential arc fault.
View attachment 201007
View attachment 201008

Just butting in here. From my understanding, our arc fault detection monitors the PV current and voltage waveforms. It is constantly monitored to see if there is a waveform pattern that matches the pattern of a typical arc fault. If it sees this pattern you will get an F63 arc fault on the Sol-Ark. I believe this is also how most arc fault detection works on other safety products as well because detecting arcing( to open air, wood, other wires ect.) is historically pretty hard. This is just what I have heard by no means can I confirm this as a 100% certainty, but I have no reason to doubt it at this time. We do recommend you have arc fault detection on by default for safety and code compliance.

Normally the standard procedure is if you are getting an arc fault, make sure you manually clear it, and if it happens again its time to start checking termination points, J-Boxes, MC-4 connectors and wires. You can isolate strings by taking them all off and then adding them back one at a time each day and seeing which string gives you an Arc Fault on the Sol-Ark.

I am not in the finger-pointing game by any means but if you are having TIGO events several timesa day, its possible that when the RSD is shutting the PV power down it can at times resemble the pattern our ARC Fault detection is looking for. Just a guess.

Side note: If this is a 15K inverter please be sure you are using two RSD transmitter cores on your PV wires so that the PLC signal from the transmitter is stronger and better able to reach the RSD modules on your PV panels. The 15K is known to emit its own high-frequency noise that can interfere with this PLC signal, but in all cases having 2 cores fixes the issue. Best rule of thumb is every inverter should get its own transmitter and two cores for its PV strings. It is not necessarily required as many installs don't follow this rule, but it's pretty much guaranteed to prevent the issue explained above.

 

[Oldphile]

Just butting in here. From my understanding, our arc fault detection monitors the PV current and voltage waveforms. It is constantly monitored to see if there is a waveform pattern that matches the pattern of a typical arc fault. If it sees this pattern you will get an F63 arc fault on the Sol-Ark. I believe this is also how most arc fault detection works on other safety products as well because detecting arcing( to open air, wood, other wires ect.) is historically pretty hard. This is just what I have heard by no means can I confirm this as a 100% certainty, but I have no reason to doubt it at this time. We do recommend you have arc fault detection on by default for safety and code compliance.

Normally the standard procedure is if you are getting an arc fault, make sure you manually clear it, and if it happens again its time to start checking termination points, J-Boxes, MC-4 connectors and wires. You can isolate strings by taking them all off and then adding them back one at a time each day and seeing which string gives you an Arc Fault on the Sol-Ark.

I am not in the finger-pointing game by any means but if you are having TIGO events several timesa day, its possible that when the RSD is shutting the PV power down it can at times resemble the pattern our ARC Fault detection is looking for. Just a guess.

Side note: If this is a 15K inverter please be sure you are using two RSD transmitter cores on your PV wires so that the PLC signal from the transmitter is stronger and better able to reach the RSD modules on your PV panels. The 15K is known to emit its own high-frequency noise that can interfere with this PLC signal, but in all cases having 2 cores fixes the issue. Best rule of thumb is every inverter should get its own transmitter and two cores for its PV strings. It is not necessarily required as many installs don't follow this rule, but it's pretty much guaranteed to prevent the issue explained above.

I was initially investigating potential arc fault issues, but now the focus is the TIGO optimizers dropping out for 5 minute durations. The optimizers don't use PLC. The CCA communicates with the TAP on the roof over RS485. The TAP communicates with the optimizers wirelessly. I am using the Sol-Ark 12 volt output to power a relay that keeps the CCA alive until the RSD button is pushed.

I have two Sol-Ark 15K systems, one with optimizers and one without. The one without has never reported an arc fault. I hope Sol-Ark optimizers will perform better.

Thank you for your input.

 

[Carlos_Sol-Ark]

I was initially investigating potential arc fault issues, but now the focus is the TIGO optimizers dropping out for 5 minute durations. The optimizers don't use PLC. The CCA communicates with the TAP on the roof over RS485. The TAP communicates with the optimizers wirelessly. I am using the Sol-Ark 12 volt output to power a relay that keeps the CCA alive until the RSD button is pushed.

I have two Sol-Ark 15K systems, one with optimizers and one without. The one without has never reported an arc fault. I hope Sol-Ark optimizers will perform better.

Thank you for your input.

Hmmmmm It might be worth it to power that relay from its own 12V power supply. Have the power supply plug into a 120V circuit that is on the LOAD side of the inverter. It will still go dead when you hit the emergency stop button which should fit your needs. I only suggest this because the 12V output power is very small only 100mA and I want to be sure the Sol-Ark 12V power supply isnt intermittently cutting power to your relay and there for the CCA unit. Just an idea. If you feel like its still inverter tied at any point reach out to support and you can ask for me directly.

 

[Oldphile]

Hmmmmm It might be worth it to power that relay from its own 12V power supply. Have the power supply plug into a 120V circuit that is on the LOAD side of the inverter. It will still go dead when you hit the emergency stop button which should fit your needs. I only suggest this because the 12V output power is very small only 100mA and I want to be sure the Sol-Ark 12V power supply isnt intermittently cutting power to your relay and there for the CCA unit. Just an idea. If you feel like its still inverter tied at any point reach out to support and you can ask for me directly.

It's a SSR, it only draws 15 ma. My neighbor has the same setup without the TIGO issues I'm experiencing.

 

[Lt.Dan]

Hmmmmm It might be worth it to power that relay from its own 12V power supply. Have the power supply plug into a 120V circuit that is on the LOAD side of the inverter. It will still go dead when you hit the emergency stop button which should fit your needs. I only suggest this because the 12V output power is very small only 100mA and I want to be sure the Sol-Ark 12V power supply isnt intermittently cutting power to your relay and there for the CCA unit. Just an idea. If you feel like its still inverter tied at any point reach out to support and you can ask for me directly.

This is how mine is powered. A 120v circuit from the load panel runs to a din rail mounted 12v converter, to the CCA unit.

 

[Rusty959]

I have a buddy with a SMA SB 7.7 with a TIGO optimizer on every panel who also get arc faults on the inverter. So it's not an exclusively sol-ark thing for sure. He spent lots of time discussing with TIGO, who seemed to mildly accept it was a re-occurring issue with the optimizers but the only solution they gave was to disable the optimizing entirely. This did seem to solve the issue, though being intermittent is hard to verify without a significant period of time in this mode.
At the end of the day he had them re-enable optimization because since half of one string experiences significant morning shading, he's better off having it go down for 5 minutes occasionally instead of no optimization at all.

 

[Rusty959]

Also noteworthy, I used to run the same SMA inverter, same TIGO parts, literally purchased together so as close of lots as possible, and I never had the same issue. The difference of my setup to his was only 1 of my 3 strings ran optimizers where all 3 of his did.

I now run the EG4 18kpv and also have not seen an arc fault on it.

 

[Hedges]

Side note: If this is a 15K inverter please be sure you are using two RSD transmitter cores on your PV wires so that the PLC signal from the transmitter is stronger and better able to reach the RSD modules on your PV panels. The 15K is known to emit its own high-frequency noise that can interfere with this PLC signal, but in all cases having 2 cores fixes the issue. Best rule of thumb is every inverter should get its own transmitter and two cores for its PV strings. It is not necessarily required as many installs don't follow this rule, but it's pretty much guaranteed to prevent the issue explained above.

How about passing PV wires 2, 3, maybe 4 times through a single core? That should make a larger (voltage) signal. (but would be smaller current). Could also pass PV+ and PV- through core in opposite directions.

You guys should be able to check such things out in your lab, also look at the RSD keep-alive signal and your own switching noise. Then you can say if I'm all wet, or if this does work (without customer needing more hardware.)

 

[Oldphile]

the only solution they gave was to disable the optimizing entirely.

Interesting. I didn't know that was an option. As I said in post #29 the main issue is the optimizers going offline for 5 minute durations. This is with Sol-Ark arc fault disabled or enabled. I've turned this problem over to my installer. I've done all the monitoring I can.

 

[andrew-d]

FWIW, I had a ground fault on my Sol-Ark 15k + Tigo setup, and it turned out that a Tigo optimizer had gone bad and was shorting to ground, so the Sol-Ark was (correctly!) throwing an error when that happened.

The fix was to use a voltmeter and measure to ground, and divide by the per-panel voltage; that helped narrow down which panel/optimizer was likely the issue. After replacing it with a spare, I contacted Tigo support; they shipped me a replacement one for free.

I don't know either way whether an optimizer failure could manifest as an arc fault issue, but I wouldn't be surprised if so. Probably worth doing what @Carlos_Sol-Ark mentioned and isolating strings for a few days to see if there's a single bad unit.

 

[Oldphile]

I finally saw with my own eyes; the CCA green LED was off and the TIGO monitoring showed no data for a 5 minute duration at that time. Someone on Reddit said they had my issue, only worse, and it was a bad TIGO power supply. I have another power supply, but I'll wait for input from my installer. During the install when I was having issues they told me Don't touch anything.

Edit: I got PTO November 10, 2023. Today I went back to November 11 with TIGO Advanced Charts. This problem has been here since day one.