Panel misbehaving

where2 · 11-24-18, 11:34 AM

For a few weeks now, I've had one panel misbehaving with an output down around half the daily energy of the others in the system. Following Pinball's instructions in another thread, I reset my microinverter yesterday to see if a reset had any positive results.

When I first noticed this panel had an issue, the thing I noticed was that the output voltage looks to be ~33% below normal. I'm using 60 cell panels, which seem to run 24-26V when the microinverters are running their MPPT algorithm. This misbehaving panel is running around 17V, while the panel next to it is running 24V. Recognizing each panel contains bypass diodes, this leads me to a hypothesis that one of the three bypass diodes is not functioning.

These are Evergreen panels, purchased for $0.78/W after Evergreen went belly up. I'm on my own as far as remedies at this point. (However, I knew that 6+ years ago when I bought the panels). If it's just a blocking diode issue, Newark or Digikey should easily be able to get me up and running again with a replacement diode at minimal cost.

I figured I would post my hypothesis, and see if any of the other PV users have had any similar issues with their systems? I'll go up on the roof and see if I can check that panel with a digital volt meter, if I can get the cover off the back of the panel where the diodes live.

pinballlooking · 11-24-18, 01:12 PM

One good check is to swap the micro inverter with a know good on from a panel next to it that will 100% let you know it is not the microinverter. (this is what Enphase support will have you do)

I called Enphase support on the one that was 20%. They were able to tell from their side it was the microinverter without swapping and approved the RMA over the phone. They had me ship the bad one back to them with a prepaid mailer.

I have had 3 microinverter replacements and one was just producing 20% lower than the others. In this case it was the microinverter not the panel. The other two just had 0 power output.

where2 · 11-24-18, 07:26 PM

Thanks for the suggestions Pinball. Occasionally, it's NOT the fault of an Enphase microinverter.

In my case the microinverter was reading exactly what it was seeing. When I disconnected the MC4 terminals between the microinverter and the panel, the output was indeed well under the expected voltage. Voc on these panels is 38V at STC according to the specs on the back of the panels. So, I pulled the cover off where the bypass diodes are, and found that area was fully potted with sealant. Good design, however that makes troubleshooting more difficult.

Reluctantly, I pulled the panel out of the array to get to the back side and dig the potting material out of the junction/wiring box. Once I had most of the potting material removed, I noticed at the other end of the panel was a slight bulge in the PVA material on the back of the panel where one of the wire traces turned 90° and obviously had a solder joint in the trace material. I plugged my cheapie digital volt meter into the output leads (with the panel still facing the roof deck), and when I pressed on the bulged spot in the PVA, the voltage jumped from 17v to 24V. It almost felt like pushing a bubble membrane key switch. Hey, that's the problem!!

So, I carefully cut a postage stamp sized flap open in the PVA where the bulge was, soldered the trace back together (with my soldering gun using an extension cord to the second floor roof). I knew I had a solid connection when the voltage was at 24V watching the meter while I soldered (panel facing the roof). I sealed the open flap in the PVA with 100% silicone sealant, then covered it with tape. I replaced the potting material I had removed from the junction box on the back of the panel with 100% silicone sealant. It's not as neat as the original was, but it will work. I paid particular attention to getting the silicone into all the corners to get a good seal. When the sun comes up tomorrow we'll see if we're back in business. I suspect we will be. Time will tell if my solder connection lasts longer than the factory connection.

If this had been a typical string inverter system, that open connection would have persisted for months. Having microinverters, I was able to determine which panel had issues, and pinpoint where to start troubleshooting.

The most concerning part of the whole repair was having the panel out of the mounting system. Since this was a middle panel, the clamping system relies on the adjacent panel being in place. I was a bit concerned to leave the house to get supplies recognizing that any sudden gust of wind could upset the adjacent panels which were no longer adequately secured. In the end, I had a friend pickup the silicone and bring it to the house while I stayed on the roof minding the panels and the loose one I had upside down on the peak of the roof.

pinballlooking · 11-24-18, 07:41 PM

I have seen where others had replaced bypass diodes. This is first one where someone had a bad solder joint.
No parts fix even better. Thanks for sharing in case someone else has this same issue.
Great job on the quick fix.
Without panel monitoring I would still have the one that was putting out 20% less. Find that on a string inverter.

11-24-18, 11:34 AM	#1
where2 DIY Geek Join Date: Apr 2013 Location: Sunny Florida Posts: 401 Thanks: 74 Thanked 83 Times in 73 Posts	Panel misbehaving For a few weeks now, I've had one panel misbehaving with an output down around half the daily energy of the others in the system. Following Pinball's instructions in another thread, I reset my microinverter yesterday to see if a reset had any positive results. When I first noticed this panel had an issue, the thing I noticed was that the output voltage looks to be ~33% below normal. I'm using 60 cell panels, which seem to run 24-26V when the microinverters are running their MPPT algorithm. This misbehaving panel is running around 17V, while the panel next to it is running 24V. Recognizing each panel contains bypass diodes, this leads me to a hypothesis that one of the three bypass diodes is not functioning. These are Evergreen panels, purchased for $0.78/W after Evergreen went belly up. I'm on my own as far as remedies at this point. (However, I knew that 6+ years ago when I bought the panels). If it's just a blocking diode issue, Newark or Digikey should easily be able to get me up and running again with a replacement diode at minimal cost. I figured I would post my hypothesis, and see if any of the other PV users have had any similar issues with their systems? I'll go up on the roof and see if I can check that panel with a digital volt meter, if I can get the cover off the back of the panel where the diodes live. Last edited by where2; 11-24-18 at 12:32 PM..

11-24-18, 01:12 PM	#2
pinballlooking Super Moderator Join Date: Jun 2013 Location: SC Posts: 2,923 Thanks: 172 Thanked 564 Times in 463 Posts	One good check is to swap the micro inverter with a know good on from a panel next to it that will 100% let you know it is not the microinverter. (this is what Enphase support will have you do) I called Enphase support on the one that was 20%. They were able to tell from their side it was the microinverter without swapping and approved the RMA over the phone. They had me ship the bad one back to them with a prepaid mailer. I have had 3 microinverter replacements and one was just producing 20% lower than the others. In this case it was the microinverter not the panel. The other two just had 0 power output. __________________ Current project Aquaponics system , Passive Solar Greenhouse build To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts. Solar Install 12.5 Kwh- To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts. Mini Split installs - To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts. EV Chevy Volt - To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.

11-24-18, 07:41 PM	#4
pinballlooking Super Moderator Join Date: Jun 2013 Location: SC Posts: 2,923 Thanks: 172 Thanked 564 Times in 463 Posts	I have seen where others had replaced bypass diodes. This is first one where someone had a bad solder joint. No parts fix even better. Thanks for sharing in case someone else has this same issue. Great job on the quick fix. Without panel monitoring I would still have the one that was putting out 20% less. Find that on a string inverter. __________________ Current project Aquaponics system , Passive Solar Greenhouse build To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts. Solar Install 12.5 Kwh- To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts. Mini Split installs - To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts. EV Chevy Volt - To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts. Last edited by pinballlooking; 11-24-18 at 08:44 PM..

11-24-18, 07:26 PM	#3
where2 DIY Geek Join Date: Apr 2013 Location: Sunny Florida Posts: 401 Thanks: 74 Thanked 83 Times in 73 Posts	Thanks for the suggestions Pinball. Occasionally, it's NOT the fault of an Enphase microinverter. In my case the microinverter was reading exactly what it was seeing. When I disconnected the MC4 terminals between the microinverter and the panel, the output was indeed well under the expected voltage. Voc on these panels is 38V at STC according to the specs on the back of the panels. So, I pulled the cover off where the bypass diodes are, and found that area was fully potted with sealant. Good design, however that makes troubleshooting more difficult. Reluctantly, I pulled the panel out of the array to get to the back side and dig the potting material out of the junction/wiring box. Once I had most of the potting material removed, I noticed at the other end of the panel was a slight bulge in the PVA material on the back of the panel where one of the wire traces turned 90° and obviously had a solder joint in the trace material. I plugged my cheapie digital volt meter into the output leads (with the panel still facing the roof deck), and when I pressed on the bulged spot in the PVA, the voltage jumped from 17v to 24V. It almost felt like pushing a bubble membrane key switch. Hey, that's the problem!! So, I carefully cut a postage stamp sized flap open in the PVA where the bulge was, soldered the trace back together (with my soldering gun using an extension cord to the second floor roof). I knew I had a solid connection when the voltage was at 24V watching the meter while I soldered (panel facing the roof). I sealed the open flap in the PVA with 100% silicone sealant, then covered it with tape. I replaced the potting material I had removed from the junction box on the back of the panel with 100% silicone sealant. It's not as neat as the original was, but it will work. I paid particular attention to getting the silicone into all the corners to get a good seal. When the sun comes up tomorrow we'll see if we're back in business. I suspect we will be. Time will tell if my solder connection lasts longer than the factory connection. If this had been a typical string inverter system, that open connection would have persisted for months. Having microinverters, I was able to determine which panel had issues, and pinpoint where to start troubleshooting. The most concerning part of the whole repair was having the panel out of the mounting system. Since this was a middle panel, the clamping system relies on the adjacent panel being in place. I was a bit concerned to leave the house to get supplies recognizing that any sudden gust of wind could upset the adjacent panels which were no longer adequately secured. In the end, I had a friend pickup the silicone and bring it to the house while I stayed on the roof minding the panels and the loose one I had upside down on the peak of the roof.