Graphing Solar Output

Steven Lake

Hi all. I've got an issues I need some help with. I've got one of my panels I think is either dying, or I've got something that's screwing up its output. So I'm wondering if there's a took, a device, or something that will allow me to take 1-5 minute snapshots throughout the day and graph how much power the panel is putting out so I can see if there's a pattern to the losses (which if there is, I'll know what time of day and can be there at the time it's happening and see if it's anything I can correct), or if it's something consistent enough that I can say it's faulty and replace it. The reason I ask is I'm not 100% certain it's either a partial failure on the part of the panel, or there's something else going on here (bad positioning, random shade, etc) that's causing the issue. Either way it's not putting out what it should, so I'm just trying to do trouble shooting before scrapping it. If it's something easily fixable there's no point throwing it out. To me it doesn't matter if it has to hook to a laptop or is stand alone, because it'll be unattended the entire testing period. Mostly because I don't have the freetime to baby sit it for even a few minutes, let alone an entire day or more. Anyhow, any help or suggestions you can offer is welcome.

BB.

Re: Graphing Solar Output

I like to use a DC current clamp meter.

If you have several panels in series, the only way to see which panel may have failed, is to measure the operating voltage across each panel. If not ~Vmp, then you have found the bad panel/electrical connection... But that is difficult to measure unless you just pierce the insulation to measure voltage (not good for outdoor wiring).

Once you have identified a bad panel/string... You can take each panel and measure Isc... If your Isc is ~1/2 or less of rated value (in full sun), you have found the bad panel (if you have several panels, you can measure Isc for each and figure out what Isc is for the sun at that moment of time).

I had a bunch of failed panels in one array. In my case, if you looked very closely under the glass (top side) right above the J-Box, there was a faint coffee (rust) colored "stain" from where the J-Box wiring overheated.

About the only repairable item in the J-Box is the bypass diodes (diodes can fail shorted--which will affect your output, or open--which you probably would not see electrically from system operation). Most panels do not have blocking diodes (that I know of).

-Bill

Steven Lake

Re: Graphing Solar Output

Sorry for the slow reply. It's harvest season which means I'm kinda slammed. Anyhow, I like the Ammeter. That would come in handy for stuff other than just this. But how would I graph this over the day as I don't have the time to just sit around and watch it? I'm sure this would probably spot immediate issues, but I already put a multimeter to this and the power seems to be doing what it should for this one panel (there's only one here and not an array, btw) in this location. But that's only over a 5-10 minute period and I'm trying to determine if it's time of day specific, or what. I guess that's kinda what I'm after. IE, if it's working fine from 9am-11am and is DOA until 4pm I want to know that so I can be there when it DOA's to try and find out what's doing it, because while it passes inspection with the multimeter and amp meters I have (albeit not as fancy as the one you suggested) while I'm there, it's not producing enough to charge the bank the way it should. It did in the past, and quite well actually, so for it to suddenly have this happen has me confused, hence why I'm trying to trouble shoot it this way, if you understand my reasons for this. I have my suspicions what it is, but I need numbers and time stamps to be sure.

BB.

Re: Graphing Solar Output

I would start with just mid-day power measurements with a hand held DC current clamp DMM. In general, a failing/failed panel will output about 1/2 or less current vs its working neighbor.

Once you see a failure--It is done and you are contacting the factory/retailer for warranty exchange.

In general, I would bet that most solar panel failures (like the power section failures of power supplies/inverters) are "hard failures". The need 100% good connections to conduct rated current. As soon at the connection goes "high resistance" (bad trace bonding, cracked cell, etc.)--It pretty much is over for carrying and useful amounts of current. The point of high resistance usually becomes a point source of heat too--Which quickly helps the failure mechanism along.

You can get more expensive meters that will have some logging functions on them. But then for check solar panel output, you almost want two current channels at once--One channel measuring the amount of available sunlight, and a second channel to compare with--Otherwise, it is difficult to just look at a single set of measurements and figure out what is happening (unless the output is zero or suddenly drops to near zero on an otherwise clear day).

Interestingly, my panels failed around this time of year--And went from looking "OK" to wow, something is not right in about 2 weeks. Once it went to not looking right (about 1/2 output), the failure stayed there and the installer just took quick Isc measurements (very low) and notified the panel company that over 50% of my panels failed.

-Bill

ggunn

Re: Graphing Solar Output

Steven Lake wrote: »

Sorry for the slow reply. It's harvest season which means I'm kinda slammed. Anyhow, I like the Ammeter. That would come in handy for stuff other than just this. But how would I graph this over the day as I don't have the time to just sit around and watch it? I'm sure this would probably spot immediate issues, but I already put a multimeter to this and the power seems to be doing what it should for this one panel (there's only one here and not an array, btw) in this location. But that's only over a 5-10 minute period and I'm trying to determine if it's time of day specific, or what. I guess that's kinda what I'm after. IE, if it's working fine from 9am-11am and is DOA until 4pm I want to know that so I can be there when it DOA's to try and find out what's doing it, because while it passes inspection with the multimeter and amp meters I have (albeit not as fancy as the one you suggested) while I'm there, it's not producing enough to charge the bank the way it should. It did in the past, and quite well actually, so for it to suddenly have this happen has me confused, hence why I'm trying to trouble shoot it this way, if you understand my reasons for this. I have my suspicions what it is, but I need numbers and time stamps to be sure.

If you isolate the panel and its Voc and Isc numbers look good when there is full sunlight roughly perpendicular to the module, it is very likely OK. If you wanted to be more detailed in your analysis you could get some wire wound resistors and plot the IV curve for it, but plotting and timestamping its output over the course of a day is unnecessary.

zoneblue

Re: Graphing Solar Output

BB. wrote: »

notified the panel company that over 50% of may panels failed.

Wow i heard the failure rates were creeping up post recessiono, but 50%? must have been a bad batch. Was it soldering or cracked cells?

BB.

Re: Graphing Solar Output

Bad production or bad run... BP 4175 panels made in USA became BP 4175B panels made in China--And now BP is out of the solar biz.

http://www.npr.org/2013/03/07/173656739/bp-bows-out-of-solar-but-industry-outlook-still-sunny

It took me a few months to get replacements (there was a good size queue for replacement panels--apparently).

The panels that failed all had a very light coffee (or rust) colored stain just under the glass over the sealed J-Box. Why, never heard.

-Bill

Steven Lake

Re: Graphing Solar Output

Thanks for the input guys. Yeah, this seems like one of those bloody ghost problems I seem to excell at. When I'm there it's all roses and kittens, but overall daily output has ceased to be enought to keep the units we're using charged. Oddly enough, while I can't say this is the issue, while heading out to pull beans I did spot what looks like a turkey vulture using my solar rig as a tanning bed. You can't exactly generate a ton of energy when you've got a 20lb bird splayed out on your panel. o_0;;

BB.

Re: Graphing Solar Output

I would look at battery and Charge Controller connections... "Loose" and corroding connections in your high current DC circuit path (battery posts/flags/etc.) are probably more likely to go intermittent vs solar panels (check panel j-boxes if not sealed/array connections too).

Steven Lake

Re: Graphing Solar Output

Got it. Bad battery. Evidently something happened and the battery decided it didn't want to hold a charge anymore. Put a new battery on it and everything is happy again. Wish I would have thought of checking that sooner. o_0