Evergreen cell grading - how to rate culls?
By DanG, Section Solar Posted on Fri Jan 28th, 2005 at 08:44:35 PM MST
peak power output test jig ideas

Just wondering about fabricating a test jig to better sort the EBay cells I've already purchased. Without spendy lab equipment, what is best way to sort via output curves? Known: Retail Evergreen PV panels are offered in 4 grades of PV cell quality 1.31, 1.42, 1.52, and 1.6 amp peak per cell. Panel specification table ---> Pp W = 94    102    110    115 Vp V = 31.8   32.4   32.7   33.0 Ip A =  2.96   3.15   3.36   3.48 VocV = 40.0   40.0   40.0   40.0 IscA =  3.65   3.75   3.84   3.86 w/o spendy lab equipment, what is quick and easy way to gauge output curves? Is there a calculable constant from above table that I could use to accurately sort 600 cells into grades if each cell were exposed to equal light source in a test jig? I can't depend on constant sunlight - my guess is a cool yet intense 5000K light source then read current produce on a dummy load. I know I can just measure 100 cells I & E outputs then calculate apparent power and sort accordingly - only the peak power points under load varies between grades of cells & I'd love to have a quick go/no-go referance. Any thoughts on what load to use, or possible (non-calculated) power output monitoring? Easy enough to put tape over analog meters legend and write in my own :) Any brilliant shortcuts come to mind?

Evergreen cell grading - how to rate culls? | 8 comments (8 topical, 0 editorial)

Re: Evergreen cell grading - how to rate culls? (3.00 / 0) (#1) by picmacmillan (rmacmill@sympatico.ca) on Fri Jan 28th, 2005 at 03:04:11 PM MST (User Info) http:// www.frecklefarmloghomes.com

i don't know that it needs too much attention....in my opinion the output differences are very minimal by the time you count your losses etc....shoot, even as a cloud passes the cells there is significant losses just in that alone.....even when the panel is completed, cast your shadow over the panel and you will see a significant drop....i think because of all the losses in solar that maybe you just give a good visual inspection, maybe with a light...look for any vertical cracks in the cell itself and start soldering..i made a panel from the broken ebay cells you can get and it has pretty good output even with 1/3 of the cell missing....just my laid back veiw...good luck....pickster http://www.frecklefarmloghomes.com

Re: Evergreen cell grading - how to rate culls? (3.00 / 0) (#2) by iFred (ifred2006@yahoo.com) on Fri Jan 28th, 2005 at 05:19:03 PM MST (User Info)

You could put it under light, get a reading for both current and voltage. Current test in series and voltage test in parrele. You should get about 1/2 voltDC per cell @ 3Amps DC max. >> all energy used to produce this comment or post came from solar and wind energy! It works!

Re: Evergreen cell grading - how to rate culls? (3.00 / 0) (#3) by DanG on Sat Jan 29th, 2005 at 11:01:00 AM MST (User Info)

Actually looking for something more accurate then eying them. I want to process a batch of 10lbs first; I've got over 40lbs of PV cells here to go through so it's not like I will have too few cells to complete arrays when the testing is done. And handling and testing 2000 fragile PV cells will drive a person up a wall rapidly - the laid back creq around here can relate to that... Ideal: digital output comparator that lock highest power output reading while dialing up PV loading from short circut to open circut. Even 10-bar LED calibrated to show 1.8 to 2.8 watt readout that adjustable for lower wattage but same divisions, maybe a log function even. The PV cell will slip into V shape hinged cradle, close V to connect to meter - and the clear plexiglass cradle will be attached to lightbox. Now vary output load and measure current - but the output voltage/current (power) peak is what I want to sort by. Any ideas on locking that down?

Re: Evergreen cell grading - how to rate culls? (3.00 / 0) (#4) by Victor on Sat Jan 29th, 2005 at 03:54:07 PM MST (User Info)

If you just sort by peak power then series sorted cells you may end up operating the cells at less than peak power due to mismatches of each cell"s power curve.  Off the top of my head I would be tempted to sort by short circut current with a go no go check on the knee of the curve [ Parent ]

Re: Evergreen cell grading - how to rate culls? (3.00 / 0) (#5) by DanG on Sat Jan 29th, 2005 at 10:47:52 PM MST (User Info)

thanks, thats got me looking at peak power voltage (31.8 - 33VDC) in the 72-cell panel specs - set load voltage for .44 to .46V displayed and then randomn sample each lot to chart the current response to my light source to calculate sort thresholds. The short circut current also depends on my light source; IscA range of 3.65 to 3.86A in full solar maximum hopefully will scale down accurately to the light source used. Alot of these cells are cosmetique culls w/ lithography off center or back conductor paste misapplied, a single lace edge w/ 1/64" silicon sheared away for an inch or missing a tab, etc. Many appear culled for irregular edge scallop that can match a 1mm bump on one cell to a -1mm dimple on another, or somewhere on assembly line 20 at a time got a corner nicked by robotic magazine loading or movement. I saw very few crushed cells in the main lot I purchased. Because there are so many superficial defects the foundry chose not to warranty for 25 years, I am hoping many are top-output cells to use on temporary install PV arrays, dry camping or construction site usage, the lower output cells will be the primary PV banks on my permanent install. Again - A call for smooth ideas on making a 2200 cell sort error free!! [ Parent ]

Re: Evergreen cell grading - how to rate culls? (3.00 / 0) (#6) by ghurd on Mon Jan 31st, 2005 at 09:40:47 AM MST (User Info)

How about a couple big old halogens, connect the cell to a 0.150 ohm resistor bank, measure the output current?  It would be under load at about the design V, I, and P. Thats about how the factories do it to seperate the cells into a few matched batches.  That is why (partly) you can get a 55, 60, and 65w panel all in the same frame and same size cells. A weak cell will lower the output.  I don't have a guess how much. G-

Re: Evergreen cell grading - how to rate culls? (3.00 / 0) (#7) by DanG on Tue Feb 1st, 2005 at 09:06:53 AM MST (User Info)

Heyas - apologies on dyslexic first post data, correction follows... >> "1.31, 1.42, 1.52, and 1.6 WATT peak per cell" Looking like I will use 25' 24ga solid wire wound in old-fashioned slider coil potentiometer to get variable .10 to .16 ohm load for peak power & the shortcircut test will be first as suggested. Thanks again. Now I'm just waiting for a sunny blue-sky day to proof the 50 sample cells from the first lot and see how skewed my light box results vs. sunlight to compensate...

Re: Evergreen cell grading - how to rate culls? (3.00 / 0) (#8) by DanG on Wed Feb 2nd, 2005 at 04:53:13 PM MST (User Info)

http://www.nrel.gov/measurements/device.html everything you wanted to know about PV testing but were afraid to ask

Login

Make a new account Username: Password:

Total Views

92 Scoop users have viewed this posting.

Related Links

· Also by DanG