Hi, I’ve had two 100W flexible solar panels connected in parallel to a 20A Tracer controller for about 18 months. They’ve been really great, keeping the fridge going and the battery fully topped up, off grid at festivals etc. However when we took the cover off the van last weekend, both panels appear to have failed. They produced between 9V and 13V but no current. The panels are clean (covered all winter) and have no signs of damage. I ran the fridge for nearly a hour to run the battery down a bit, but no change. I then covered the panels, disconnected them, disconnected the controller from the battery, then reconnected everything in the reverse order to reset the controller, but there was still no change. I also tried connecting the panels one at a time (remembering to cover/uncover them before disconnecting & reconnecting each time) and found that when I tapped the diode box on one panel, the voltage increased and it started producing about 1.5A, which was about right for the conditions. However it only worked for short time and today I can’t get it to start again. The other panel refused to work at all, it just outputs a voltage of between 9-10V. I also tried connecting up a spare panel and it worked fine, so I’m sure it’s the panels that are at fault. I know that bypass diodes can be troublesome, hence why I tried tapping them, but for both panels to fail simultaneously seems odd, unless there was a common cause. The back mounted diode boxes are not easy to access without removing the panels, but I might be able to prise them open and snip out the diodes. Does anyone know if I can operate the panels without bypass diodes, or have any other ideas as to what may have happened and how to fix it. Thanks
Not clear if you are using a multimeter or charge controller for your voltage measurements. Controller can give erroneous readings if not connected to battery; don't think your one has battery backup. You may be able to carefully scrape off the varnish on the tracks on the panel upstream of the diode box. Test the voltage with a DVM there then revarnish. On your spare panel, you should see the voltage changing with cloud cover. Sent from my MotoG3 using Tapatalk
I measured the voltage on the controller and then used a multimeter across the controller terminals and across the panel connectors. They all gave very similar readings. Yes, the spare panel voltage/current did fluctuate as I rotated it. We have no cloud today. The boxes aren’t very accessible, but I’ll have a go at prising them open and see what gives. Presumably I can do a quick continuity check of the diodes in each direction to confirm if they’re working? Thanks guys.
Yeah, multimeter in diode position should give open circuit one way and closed circuit the other. Doubt it will be them though, more likely corrosion. Are they solid panels of flexi? Sent from my Pixel 4 using Tapatalk
They’re flexis. The box lids are off and I now will need to carefully dig out the sealant that to get at the internals but the sealant looks brand new - no indication of any potential route for moisture ingress. They are mounted on the underside of the panels and well protected.
You could add diodes across each panel externally if they are in series, provided you get access to the separate positive and negative wire from each panel. Expect to find corrosion or a cracked tinsel coming off the panel . You may have to remove the panel and cut away at the covering on the back to get at the wiring... then seal up with PU glue/sealer. Cause could be vibration of unsupported diode box locally flexing the panel..
Right, I managed to scrape out the sealant from the diode box (using a mirror to see by) and could then see the diodes. Everything looks in mint condition but unfortunately the diode connections are on the other side of a circuit board. Then with the panels uncovered again I initially got 9.6V across one diode (or panel half) and 9.5V across the other. So I connected the panel up to the controller again and recorded 19v but no current. The controller voltage then reduced to around 13V and the diode voltages read 9.7V & 3.7V. I set the meter to the diode setting and got OL across both diodes (+ve to -ve) and 0V across what seemed to be the good diode (-ve to +ve) and 0.156V across the other (-ve to +ve). Does 0.156V mean that one half of the panel may have failed and it is now being bypassed? I haven’t tested the other panel yet.
By the way, I cut slots in the roof for the diode boxes to slot down into and stuck the surrounding panel down with sealant. They are fixed down pretty solidly, which is why I have difficulty reaching them from underneath.
I don’t have an answer but I’m watching this with interest as I have the same problem. I’ve been carrying around two useless 100w semi-flexible panels for 18 months or so, both produce circa 20v open circuit but near zero short circuit amps.
Don't think you can test diodes when they are connected to a power source; might be valid if panels covered. Sent from my MotoG3 using Tapatalk
I did also test the diodes while covered and got a similar result, although slightly less voltage reading. I think the test might be invalid in both cases because the diodes are still connected in parallel with the panel cells. I think further dismantling is in order today.
Call the company you bought the equipment from & ask them for advice/replacement. It doesn't matter if its out of warranty as goods must be fit for purpose and last a reasonable length of time. @Sydney & @77 Westy both seem to have fairly new equipment that should still be working - ring the people who sold it to you.
I can sympathise. I may end up doing the same. It’s a lot simpler to leave them up there than to cut/clean off the bonding material and repair any removal damage. Before fitting these two panels I used to just tape a 100W panel (now my spare) to the roof when we arrived somewhere and remove it again when we packed up. Worked for years, and still does.
You’re quite right of course but I have probably voided any warranty/support by opening up the diode box and scrapping out the sealant, so I might as well continue investigating it. I could try ringing wrt the other one, once things are back to normal.
You could try eliminating complexity from the system by replacing the controller by a load, say a 12v bulb. The controller is mainly to protect the car battery. Put the multimeter in series with the bulb to measure current. Try it on your spare first. Check that the fuse in your meter is not blown. Sent from my MotoG3 using Tapatalk
I bought the panels and a 20amp MPPT controller from Amazon in September ’17 but I didn’t get around to fitting them until June ’18. At first, I thought the performance was poor because I often camped in shady spots but it was soon apparent that they didn’t work even in clear sunlight. The panels and controller were supplied by Eco-Worthy and were supposed to have a one-year guarantee, I’d had them about nine months. Amazon were no help whatsoever and it’s impossible to contact Eco-Worthy directly. I gave up after numerous attempts.
My 100W flexi went the same way - voltage but no amps. More volts if I cool it with a bucket of water. It lasted maybe two summers and by the time I noticed the battery was dead too. I will not be replacing the panel.
Re contact Amazon asking for full refund as goods not fit for purpose. Tell them to pass your message on to whoever deals with compliance - refer to sale of goods act. If that doesn't work send them a Notice Before Action letter
If you hit or flex it and it comes back to life it will be a cracked connection. Across a hard point like the edge of a junction box. The diodes in these installations are for dealing with shadowing sections of the panel. Without diodes even a small shadow ( guy rope, twig, bird poo) on one cell shuts down the panel. With diodes it just loses the voltage of the fraction of the panel that the diode bridges. If a panel produces 19 volts in sunlight then it is basically working but has a nearly broken wire , or a corroded connection. The wet metal salts remaining after the wire dissolves provide enough current to produce a voltage into the 10 million ohms of the meter. Diode tests across a connected solar cell are almost impossible except in the dark. And even then, the panel will still produce wrong results as it appears as a resistance in the dark. My secondhand / ex skip £700 each panels have the diodes laminated in the panel, and only had a thin, flexible plastic cover bonded on over the connections. They expired after sea water ran up the sides of the wires and got into the tinsel. Removing the connection cover and cutting back the rear plastic covering of the panel exposed about 5mm of good tinsel, which I was able to solder new wiring onto. If you have nothing to lose, try it.
