A friend of ours recently decided to upgrade the solar photovoltaic (PV) charging system in his small camping trailer with another PV panel. Not having loads of money to spend, he purchased a used 120-watt Kyocera on E-Bay. But buying the panel during winter meant that he didn’t get around to installing the panel for a few months, so it didn’t occur to him that he might want to check the panel before his chance to either return it, or ask for a refund, would expire. As luck, or the crookedness of the seller, would have it the panel did not output any power and, as so often happens with electrical devices, I received a frantic call. The option of returning the panel no longer applied and he appeared to be financially and electrically out of luck.
First, I contacted Kyocera and had my friend send off a warranty claim on the panel, just to see if it was still replaceable. Kyocera quickly shipped a working refurbished panel and when installed it functioned perfectly. My friend was so happy with the outcome that he gave me the broken panel as thanks. After sitting around in our shed for a few months, waiting for its place in the electrical repair queue, I tested its output with my favorite digital multimeter. This indicated that the panel was putting out only half of its proper voltage.
The voltage test was performed first on the two outermost screws in the junction box, the main positive and negative terminals (shown with wires connected).
When it became apparent that only half the usual voltage was present I next tested the two screws on the left and the two on the right. These each connect to 18 series-connected PV cells on either side of the panel. One side tested to a proper 10+ volts when exposed to sunshine while the other side indicated zero volts. This meant that there was a broken circuit somewhere on one side of the panel. Since the panels are covered with glass on the front and a thick coating of plastic on the back, the only ways to test for electrical output within the panel are visual inspection (looking for an obvious flaw, corroded spot, or actual break in the thin silver-colored “traces”) and “back-probing” through the plastic with sharp test probes.
It’s easy to see these electrical pathways through the glass side of the panel, but if you want to find where to stick a probe through the plastic backing you’ll need a powerful light source. You could use the sun itself, but working indoors I simply used a very bright flashlight. I first traced the major connection pathways on the plastic backing using a pencil. Then, placing the panel back in sunlight, I put one meter probe on the negative screw terminal and started probing the large, conductive, back surface of each cell with the other probe. I marked each probing spot with a pencil for later caulking.
It turned out that I had increasing voltages as I moved toward the final positive output trace, with voltage falling to zero only at the final trace itself. So, scraping the plastic backing away at the last known point where I found a voltage, I soldered a jumper wire to that point and ran the wire to the junction box’s main positive terminal. At first I had a poor connection, until I realized that the conductive trace had a very thin plastic coating that needed to be sanded. When the soldered wire held well I tested the output and found a full 21 volts at the outer terminal screws. Caulking the wire to the back of the panel, and sealing all of the tiny probe spots I made completed the repair.
It doesn’t look terrific, but with about an hour of work, and the use of ordinary step-by-step thinking, we were able to add another panel to our system at nearly no cost. Sometimes it’s not WHO you know, or even WHAT you know, but what you can work out for yourself.
Since then another PV panel, shown above as the far left, top panel (a mere 8-watt monocrystalline model) suddenly failed. Since I had been using that set of nine panels wired in series to produce high-voltage DC, one failed panel completely shut down the entire array’s output. You can see the corroded spot between two of the silicon cells in the photo below. The glass surface, back plastic, and side seals all seemed perfectly intact. But for some reason corrosion was causing an open circuit at this point. Since the panel was under a 25-year power warranty I asked for, and quickly received a replacement which I wired back onto the rest of the array.
Once again I scraped away the plastic until I reached good-looking wire traces on either side of the corroded cell. Soldering on a short jumper wire proved to restore voltage to the output terminals, although at a slightly reduced voltage due to bypassing one cell. I will no doubt reuse this panel for some other upcoming project.