Solar Panel Surgery

So.   I have a small off-grid (for power) garden shed at a property about 600 feet from my house.  I decided a couple of years ago to cut the utility power to it and make a full solar setup.   I started with a pair of Kyocera KC120-1 panels, and then the following year, doubled it to 4 panels.    At the end of last summer, I realized that my output was half of what it was after I added the second set of panels, and didn’t worry about it.   Today it was 72 degrees out, and I decided to investigate.


First I had to figure out what was wrong.   All of my connections were good.  The wires are in good shape, and I had good continuity all the way through the array…. obviously, I have a bad panel.

The panel would read ~19.4VDC when checked with no load… which is exactly what it’s supposed to be at give or take.   As soon as a load was hooked to it, the voltage dropped to ~11V… Not good.

I isolated the top and bottom of the panel in the connection box.  The way this panel is wired is 2 x ~9.5VDC strings of cells that are wired in Parallel to produce the rated voltage.   One of these string was reading 9.84V steady.  No fluctuation… the other, however, was reading anywhere from 6 – 9V, and was all over the place.

By doing that I had isolated which side of the panel was giving me the problem… Next was to find out why.  I ran my hand up the buss wire and found in an isolated area, that if I pressed on the back of the panel, the Voltage would jump to 9.74V and stay steady, as soon as I moved my hand, it would bounce around again.   Diagnosis : Broken solder joint…. repair requires panel surgery.

20160227_150947 (Medium)
Panel removed from the array.

I took the panel to my work shop, and started surgery.  I figured if I broke the panel, it was already broken, so I wasn’t out anything.  I did look at ordering another one, however the only ones I could find online were shattered panels.  One working, and one not.    Worse case, I can’t fix it, and I can put the other panel that was paired with this one on my house with my grid tie inverter and at least get something out of it.

20160227_151724 (Medium)
The offending area.

20160227_152309 (Medium)
Following the bus wire

I have to give Kyocera credit here.  This panel was made in 2000, and it looks absolutely pristine.  I don’t know the industry standard for encapsulation on solar, but I can tell you how Kyocera’s is done.  The top layer is a plastic covering.  It’s between hard and soft plastic, it’s malleable, but you’re not going to just rip it, under that is what I can only describe as an adhesive foam layer… it’s its REALLY good adhesive.   From there, the tabbing wires and appeared the cells as well are all coated in what looked like a layer of plastic, but may have very well been an epoxy or enamel coating to keep from shorting out.  It was somewhat difficult to get through it to get tests points made for testing continuity.   Kudos to you Kyocera.

After cutting out the back of the panel with a razor knife and a pair of pliers (the encapsulation on these panels is really amazing.)   I started testing resistance across the bus wire to find my problem area.

20160227_153630 (Medium)
Good continuity here.
Tested here to get a baseline, and to test the tabbing to the contacts in the terminal box.

20160227_153644 (Medium)
Good here too.
.2 ohm… doesn’t get much better than that.

I moved to the other side of the solder joint… and what a difference!
20160227_153654 (Medium)
That’s right… 254kohm… Problem located!
20160227_154329 (Medium)
That’s supposed to be attached.

So I found my problem, and busted out the soldering gun and solder.  Honestly, the gun was a bit much for this project, but it was all I had to use.

I reflowed the existing solder, and added to it as well to help re-enforce the joint.

20160227_155022 (Medium)
N
ew liquid metal.

Now… solar cells do not do well with moisture and oxygen getting into them.  There is a reason that these panels are encapsulated so well.   Since I’m not working in a Kyocera factory… I used the best thing I could think of for the job that wouldn’t break the bank and I could get to easily…. Epoxy.

20160227_154318 (Medium)
My $7 encapsulate.

I added the epoxy everywhere that I had cut into the encapsulating layer, and then spread it out with a small paint brush.

20160227_155022 (Medium)
Re-encapsulated…. kinda.

I let the epoxy set up, and then took the panel back to the garden.

20160227_162834 (Medium)
Reinstalled… looking so dapper.

I checked the charge controller before I hooked the panel back up for a baseline on what my charge wattage was.
20160227_162943 (Medium)
11W… Well the good half is shaded currently.

Then I hooked the panel back into the array with its twin, and checked the charge controller again.

20160227_163052 (Medium)
85W!   The project was a success!

At full tilt last summer I saw about 380W at the peak being pumped into my batteries.  I finished up this project as the sun was setting, but it’s obvious to me that this fixed the issue that I was having.

20160227_163318 (Medium)
The only thing you can see from the panel showing it’s been repaired by Joe Blow and not Kyocera.

All in all.  I’m happy with how it turned out.  I didn’t have to buy all new panels for a new array, I didn’t really have to buy anything other than the epoxy.  I was out $7 to fix my $135 solar panel (bought used on ebay.)

Now I can suck down the power this summer, and never have to worry about my batteries becoming too discharged and the DC power getting switched off to protect the batteries.

I would NOT recommend that someone tries this unless you have a pretty good idea as to what you’re doing, and you know the dangers it can pose.   It could have been possible for me to short out the entire panel, cause a fire, shatter the glass, break the cells.   But if you have panels that are acting up… it isn’t rocket science on how they work, and with a little knowledge, most of them can be repaired completely, or at least brought back to a working state.

I’ll have to see over the years how my epoxy holds up, as well as the panel itself over time.  Who knows, the epoxy may end up eating the solar cell.  That’s why I like doing things…. to learn!