John is embarrassed that the free solar panels he worked so hard to give away have some duds, but, THEY’RE FREE!!! Most of them work! This page is to help you sort the good and the bad and help you with fixing the bad, if there are enough to bother with. First of all, get your stack. Bring some extra straps as they may decide to shift on the pallet along the way.
If you test your stack and only have a few duds you may decide to let it go. They still work as shingles, after all. On the other hand, some of us may have a good many duds or just want to use every free watt we can so we want to know if they can be fixed. The answer is a definite maybe. I will go through some of the failures and repairs that I found and let you decide the answer for yourself. Read carefully and you can avoid breaking things worse than they are.
A good blacksmith can fix anything but a broken heart and the break of day. If you aren’t a good blacksmith you can take out your frustrations on duds.
First of all, let’s talk about what you need to do and what tools you will need. I will also mention a few things to NOT do. Here is what you need. A multimeter…a cheap digital meter is fine. Some alligator test leads will help make up for the likely fact that you only have two hands. Some solder and a soldering iron. A can of sticky goo and a condiment bottle to spread it. A lot of these panels need to be resealed, so get the goo. I got mine at an ACE Hardeware for $16. Bottles are a buck at Walmart or Dollar store. You may want some strong cheater glasses, too. Not shown: a hair dryer. You might need a putty knife and a file…I’ll explain later.
I started by sorting mine by ones that looked perfect and ones that needed to be resealed. I eventually decided that they all need to be resealed, so I went with a pile with good cells and a pile with cracked cells.
It should be a nice dark gray around the edge of the panel. If not, the seal is delaminated and there may be some moisture in there. If there isn’t, there will be and that isn’t good.
Panels with cracked cells may work perfectly well with full output, while the one that gave me the greatest aggravation looked perfect. One would hope that the panels were originally installed with skill and care, but I suspect that when they were de-installed upon recall that they were not treated so well. I envision distance tossing contests and young idiots jumping up and down on panels. Some of them are not pretty and one of mine I judged to be beyond reasonable repair.
A badly cracked cell. With simple cracks, the power has two paths to bus bars. With spider cracks you get islands that may give you nothing if the silver lines are broken at the cracks. This can cause reduced power or failure for the whole panel. When testing “crackers” , sort them and install in series with units of similar output.
OK, those are some things for which to watch. Let me mention a couple of things to NOT DO. I mention them because I already did them and it wasn’t pretty. It was, indeed, counterproductive. First of all, don’t try to remove the glass from the panel. The gray stuff around the edge is brittle and comes loose pretty easily, but between the glass and the cells is some really serious stickum. I tried cutting it with a guitar string, which resulted in a curly Super Slinky. One more fraction of an inch and you are going to make your day worse. And don’t waste your E strings. Spoiler alert: we’re going to tunnel in from behind.
At some point we should test. I am not here to teach you how to use a meter or even how to not ruin it, but will give a little guidance. The precision meter shown earlier cost exactly nothing (with coupon) at a tool store sometimes known as Horrid Fright. You can probably find something for $10 if you don’t have one. The first instinct is to set it to volts and see if the panel is putting out. I suppose that is useful, but wrong. Let’s start with volts, though, as an instructional. Put the meter range selector to DC Volts and 20. According to the label on the back of the panel you should get around 6 volts in bright sun. The wires are on the back of the panel. If you lay the panel face down in the grass whilst attaching your test leads, don’t expect much in the way of output. The blue side of the panel needs to be facing the sun. That is where the test leads with the alligator clips will be handy. Clip them onto the MC3 connectors and run them out to the meter. Note that a few panels may be missing the MC3 connector. In that case, strip back a bit of insulation and be thankful you have the alligator clip. Clip the far end of the leads to the meter leads. It doesn’t matter which way… the meter will just read -6 instead of 6 if you have it backwards. If you are doing this indoors, don’t have great expectations for output, but there will be something.
Got it? Is it working? Good. Now let me tell you why this is not the best way to do it. Some of these panels have weak connections which will allow the unloaded voltage to pass through with no problem. Problems start to show up when you put a load on it. SO, we want to test with the ammeter instead. Outdoors, you will need to test in the 10 amp range. That usually has a different place to plug in the test lead and its own setting on the meter dial. Make these changes BEFORE you connect to the panel. In the photo below, the test is inside on a rainy day, so the meter is set for 2000ma, with the red lead in the normal position. Here is the worrisome bit about using the ammeter mode. If you forget to change back to VOLTS and then go check your car battery, you will hear a “fffffttt” noise and a little puff of smoke will come out of the meter. As you probably know, all electronic devices operate on the principle of circulating smoke and electrons. When the smoke comes out, it won’t work any more.
If you are testing the panel outside with it flat on bench, you may see 3-5 amps. The rated short circuit current is nearly 8 amps if you have perfect alignment on a perfect day. Unfortunately, 8 amps is enough to fry those skinny test leads, and you know what happens when the smoke comes out. Note that the cell is cracked and I am pointing to an area that turned brown from heat, enough to pock the glass. This probably means that only one of the 3 silver conductors is connected and this one is carrying the full load. Be suspicious! You don’t want a panel that works briefly, considering all the effort you are going to expend getting this thing wired up and screwed down.
Press here and there. Flex the panel a little. Run the heat of the hair blow dryer around and see what happens. This is a different panel with nothing suspicious about it. Pressing here causes no problems. BTW, most of the problems I have found have been on the left side.
But you remember with the panel with the brown spot at which I was pointing? When I press there the panel quits! If you just want to go through testing and come back later for repairs, put a dab of masking tape over the bad spot and move on.
If you are testing outdoors, get good output and can’t aggravate the panel, place it in the good pile, but don’t put it in service. We’ll get to why, later. If the output changes distinctly, but does not go to zero, there may be a cracked trace. Each cell in the panel has those 3 silver traces on each side and sometimes a connection goes bad. On the edges it is easy to fix. If not on the edges, you may have to decide for yourself if it is worth the trouble after you get a little experience on the easy stuff.
Ready for repairs? Here is the first thing to NEVER DO (aside from the earlier bit about prying off the glass). Never open the junction box on the back. Oh, it is tempting and it looks easy enough. Besides, you just HAVE TO KNOW what is in there. Let me save you the trouble. Two metal foil traces from the solar cells come through the backing and solder to a tiny circuit board. Across those leads is a bypass diode, which lets all the other panels in a string work if this panel should die. And then there are the black wires that hang out the back. Pretty simple. But there is the oozing black silicone adhesive/sealant, too. When you finally pry the lid open you will probably rip the circuit board loose from the panel. This can probably be fixed, but it won’t be pretty and it will be SO unnecessary because I told you DON’T DO IT. Remember?
If you suspect something amiss inside that little box, it is simple enough to test from the outside. Get the meter and put it in diode check mode. The symbol is like an arrow pointing to a crossbar. Something like ->]-. Place the panel face down so it should make no output. Place the meter probes across the panel’s output leads one way and then the other way. One way there should be the baseline digit 1 showing and the other way you should get something like 130. (Your meter may be different, but the idea here is that one way nothing happens and the other way something happens.) A reading on the order of 001 would indicate a rare short circuit and the baseline 1 digit would indicate the diode is busted. That kind of problem you might want to put off to the side somewhere.
OK, we know how to test and have learned about several kinds of failures, what do we do about it? Gear up, crew, we’re going in!
Some people may want to keep things pretty and they may want to remove the laminate from the frame. I think the frame is robust enough to accommodate a certain degree of butchery, but if you want to keep the frame intact, here’s what you do. Get a putty knife, the longer the better. Now take a flat file and sharpen JUST ONE SIDE until it is sharp enough to sharpen a pencil. I’m talking about just one side of the end, the part that does the scraping, because we are going to push that up under the frame to cut through the black silicone stickum that holds it all together. Remember we put a bevel on just one side of the blade? Place the panel glass down and put the bevelled edge of the knife down against the backsheet. It should slide without digging in to the backsheet. Bend the blade just enough to clear the frame. Press hard and you will feel when it hits the edge of the frame after cutting the silicone.
I know I am not your Momma, but never put the sharp stuff pointed in the direction of raw meat, especially yours or your friends’. You will be pressing hard and if anything goes wrong you don’t want to cut the meat. When you feel it hit the stop, slide back and move over. Repeat until you have made it all the way around the frame. Hold off at the junction box until we start separating. When you get the bottom edge free, start to open up and use the blade to finish off areas you might have missed. Oh, yeah, make sure the wires are out of their guides.
Yes, it will be easier to work on the laminate without the frame in the way, but remember you have to put it all back together and buy some together-stickum to do it.
Now for the actual repairs. What I found is that the 3 main bars coming off the end of each cell (at the ends) lay across a wider bar running at right angles. In theory these are securely soldered, but in practice they may not be! I don’t know why, but it could be any combination of manufacturing, heat cycling or stomping by persons one might want to throttle if given the opportunity. The big bars then intersect more big bars that run across the panel to the center. There you find another right angle turn to go up to the junction box. I have found dubious connections at each point. Most have been on the left side (label side), as defined by looking at the panel with the mounting holes up and facing the cells. When we are working with the panel face down, junction box up, the left side will be on the right side. It’s the side with the labels, if that makes it easy.
Before you get started, let me familiarize you with the underground highway buried under that gray backsheet. Panel face down and mounting holes at the top? Let’s start with the right wire of the junction box. It runs down almost to the plastic frame crossbar, assuming you left the panel intact, makes a right turn and goes all the way to the end of the panel, about where the last of the backsheet can be seen. It then turns straight up and the three small bars from the solar cell connect. The top row of solar cells form a chain all the way over to the left where it connects to a vertical bus bar that runs just under the frame, beyond the last visible part of the backsheet. This bus bar on the left runs the full edge and connects to the 3 lines from the bottom row of cells. The bottom row runs over to connect to the lower bus bar on the right, ending just below the center bar of the frame. This line runs left to the center where it makes a right turn up to the junction box and out the black wire on the left. Got that?
Now remember the panel that you could push on and get it to change? Whether it is pushing, pulling or heating, go to the spot where things are happening.
If it helps, measure from the edges so you can precisely find the same spot on the back. Don’t get confused when you turn over to work on the back.
If you are going beyond here, I hope you have a good soldering iron and some experience with it. I have seen soldering irons that are so bad that even I have trouble using them and I have half a century experience with this stuff. You want one that is plenty hot and you want one with a chisel or flat tip. A round, pointing one is not going to be much fun to use. I tried one with a very pointy tip and bent a little hook in the end, which made it usable. What we are going to do is melt through the multiple layers of plastic to get to the metal strips. This will be tedious, messy and stinky. Safety glasses would be a good idea, but if you are wearing cheaters that is probably safe enough. You want to melt through slowly and then run the line of the area to be exposed. You DO NOT want to be burrowing off to the side of the tracks and you do not want to be applying a lot of force, which could cut traces or break stuff.
Most of the problems I have found have been on the label end. That is good news because the frame is not in the way, there, and the frame gives you a reference point for going in. From the picture you can see about where things are. I have decided that the best plan is to take the extra effort and go ahead and resolder all six lines from the two solar cells and both bus bars. This photo shows the panel turned 90 degrees clockwise, so this is the area of the upper cell. Toward the middle and at both ends you can see the cell’s tabs on the bus bar. We need to dig a little more to the left to hit the bus bar going back to the center of the panel. In addition to a couple of layers of the gray sheet, there is also some clear sheet and the solder will not stick to anything but metal. Scrub back and forth with the iron until you see the metal shining up and apply just a dab of solder. Rub back and forth until it begins to melt onto the bus before adding any more. It is very possible to get a great shiny blob of solder in there that is not connected to anything and we really need to connect.
Early on I considered Conductive Epoxy because not everybody can run a soldering iron. In the end I have to recommend against that because it is very difficult to get those traces clean. You will think they are down to metal but they still have a clear coat of whatever Suntech used in the laminating process. You can’t be sure it is clean unless you see solder melting on it.
Oh, I guess I should have brought this up earlier, but DO NOT use ACID CORE solder. It will cause corrosion and general aggravation later. Use ROSIN CORE, which treats you to a fresh pine scent. If you use your imagination. 60/40 tin/lead is the best solder. 63/37 Eutectic solder doesn’t work very well, but was a fad for a while. Most manufacturing now uses lead free solder, which is just about useless. The only worry with lead solder is that you don’t want to be snacking and soldering. It is more likely that salt from your snacks will cause problems in the circuit repair than it is that lead on your fingers will transfer to your Cheetos and poison you. Wash your hands when you are finished, at any rate.
Now you need to get out the meter and check again. Working? Press on it, rap on it, bend it, heat it. Still working? Give it the test in the sunshine. It takes a good connection to handle real amps. Does output jump up or down when pressed and stressed? That could indicate that one of the 3 tabs is not right. Does it happen when pressing on the side we repaired or on the other side? Go to where the problem is.
You may need to go to the other end of the panel. Since there is a single bus bar there, if you can reach it you can easily determine if the problem is in the top row or the bottom row by putting one meter lead on the bus bar on the other end and then touching the other probe to one wire or the other. Easier, touch the second probe to the top half and bottom have repair areas. You should get more or less equal readings, making sure the point of the meter probe is finding a clean spot, not covered in silicone or other goo. If one half, say the bottom, does not have output, then you need to burrow in and solder the three bus connections to the cell. Note that on this side you have to dig through the frame if the laminate is not removed from the frame. If that does not do it it can get real delicate burrowing in to the individual cells. I had one like that, but 7 of the cell were severely stomped and cracked so I let that one be. Use your own judgement according to your own temperament. I don’t have much of that left.
Are you convinced that you have your panel fixed? Put it face down and take your squeeze bottle of liquid stickum and pour it into the furrows you dug. The right stuff will be non corrosive and just as hard to dig into as what Suntech used. If you opt for silicone, use the stuff in the tubes from the auto parts store that says “sensor safe” on the label. It won’t corrode anything. The other might not, but apparently there is a reason they make the special stuff. House caulk? Better not.
After that dries, you can put the panel with your good ones and consider the drying out and resealing phase. I think that most of the panels have a breach and most of them have some degree of moisture inside. You need to devise a kiln. This can be a closet or posts with a tarp over or a large appliance box. Stack the panels loosely so that air can circulate. Put a fan in there and a small heater. You don’t want to start a fire or melt anything but you want it plenty warm in there to drive off the moisture. In a sunny dry place like Arizona, you might be able to just leave them out in the sun all day. Then take your squirt bottle of stickum and go around the glass thoroughly. As the panel cools down from the drying, some of the sealer may be drawn in, creating a new gap. Make adjustments as needed. I am going to use the hull of my solar yacht project and a propane heater for my kiln. The boat needs drying out after a recent storm incident, so we can kill two birds with one stone.
My kiln (blue hull) and the two solar yacht projects. They will be decked and Bimini’d in 335 watt panels from Sun Electronics and ballasted with golf car batteries..
Something else to check is the cable insulation. I found several with knicked cables and it seems it was the panel frame that cut the insulation. Water can get in there and corrode the wire. I found some of that, though it will take an awful long time for any major harm to occur.
Are some of your MC3s missing? Some will have the missing end plugged into their good ends. Remove and transplant, if this is the case. Use marine shrink tube, which contains a sealant, on the splice. If you come up short on MC3s, you can either order new ones to install or just find one that is missing the positive and another missing the negative and just solder them together. Also not that some connectors have the locks that clip over once the connectors are joined and some do not. If you think they won’t stay together, use electrical tape.
Because these are 4 volt panels, they will need to be in strings to your charge controller. The string might be just 3 panels going to a simple 12 volt controller or maybe 20 panels going to a MPPT controller. At the higher voltages you not only have to be aware of your immediate shock hazard safety, but assure that all connections are snug and all panels functioning properly. The long strings, up to 600 volts, associated with grid tie inverters is where these panels had their original problems. Careful installation and initial monitoring in the lower voltage strings used in off grid battery systems should not pose much of a threat. Grading panels with cracked cells by output and then matching to similar units in a string should make for a safe workable system.
I want my panels to be tested in real world conditions before going into the Solar Shed I am building. I have a small power station providing not-quite-enough energy to my barn. There it charges stationary batteries and any tractor batteries that need topping up or reconditioning. Batteries run a security camera and transmitter and an inverter for lights and tools, but not enough and it took forever to recharge with the small panels I had installed. A thin buried cable is limiting, but the charge controller can handle up to 48v, multiplying the use of that thin wire. I adapted an old rack to serve as a test stand. 1×4 purlins were put across the slope on 13.5″ centers. Shingles were short tested and then screwed down and cabled in series. A little more carpentry and I’ll get two more panels in their proper orientation. They will each be given a day or two under load and then swapped for more. The guy wire shadow is a plot by the evil power company. All of these panels have cracks, delamination or surgical repairs.
AND THEY WORK! AND THEY WERE FREE!
Update 12-7-16 Evening.
First batch of panels spent 24 hours on the burn in rack and performed well. There was a wide temperature range during the day and all survived a short circuit test. These panels were removed and will be sealed, tomorrow. We’ll get photos of the sealing process, both on the front laminate and back repairs. In the meantime, of the next batch to be tested, two had missing MC3 connectors. I think I will have enough of the connectors. In many cases the cables were just cut instead of being unplugged! I also found one that had been quick checked before, but had problems with a short test before joining the new string. That one it turns out has a broken trace on one of the cells and will need surgery. The problem was located with a meter and thumb pressure. Since I am not working on it immediately, I marked the spot with a paint pen.
Tomorrow, the burn in rack will get extended to burn in 7 panels. The charge controller can handle up to 8 in series. Maybe I’ll have photos of that, too. The big deal tomorrow will be a building supply run to get materials to begin the solar shed, a shed with a solar roof, or a solar array with storage beneath.
More information and tips will be posted as available, along with details of our installations.
An installation manual for this panel is available on the internet by search for Open Energy SolarSave solar roofing and not Suntech. Larger shingles were marketed under the Suntech name.
—Neal Collier (Back To Sunelec )