I realized that using tiny pieces for large panels was not at all feasible. It would take
considerable time to figure out patterns that would maximize the space used by the cells
themselves. A lot of solar panel area would also have been covered by the rather large
copper strips, and the cells still had no good means of being anchored, as gluing each
one to the foam would have consumed even more time. So the little pieces were abandoned.
A month or two later, I came upon some solar cells on eBay being sold by the pound - "cosmetic
econds," the listing said. "May have broken corners, cracks, chips, slight warps, off color, etc."
The description and the pictures looked promising. I also found
The Otherpower Forums. Although they don't have
quite as many members as the tech forums I usually visit, their loyal visitors had a wealth
of information and ideas available. It seemed that others had found these eBay cells and were
already building their own panels.
So with the new information, I decided to risk the purchase of a pound of the cells.
Some cells cracked during shipping, so they all needed individual inspection. Packing
solar cells properly for shipping either requires special shipping containers or lots
and lots of time and care to prepare each cell for the harsh environment it would have
to endure. I spent time over the next month researching ways of building panels, as
well as possible costs. Testing of the cells during that time showed that they
erformed very close to their rated specs, despite being rejects. Justification for
rejection ranged from minor problems, such as tiny edge chips or silkscreening slightly off-center,
to severe, like cells missing large portions of the conductive backing, or a few that had the
leads soldered from the front around to the back, shorting out the cell.
This is the stack of broken cells that
came from the first pound I sorted
through. Some here are outright broken
in half, but most just had hairline
cracks. I found cracks by gripping every
corner of the cell all at once, and lightly
twisting the entire thing back and forth.
Any cracks would make themselves known
by making a high pitched creaking,
caused by the rubbing of the adjacent
edges. Close inspection revealed what
kind of crack it was, and its severity.
Here's a stack of some of the usable cells I
wound up with. It was about a 60% survival
rate. I probably can still use some of the
cracked cells, as a few of the cracks don't
break any of the silkscreened lines, which
doesn't interfere with output. However, I
wanted to get as many perfect cells as I
could get; I'd deal with the less-than-perfect
A month or two had passed now since the buying of the original pound of cells. I figured
that these things could be used in full-sized panels without nearly the trouble of the
little chips I'd been working with before. Problem was, the seller's supply was limited.
So I decided to take a risk and load up on the cells while the getting was good, even though
I wasn't sure if I could produce a good design for a panel. I went on a road trip, from
eastern Pennsylvania to the seller's location in a town 30 miles south of Boston, Massachussets.
This meant I could see the cells myself before buying, and it also meant that they would not
be subjected to the rough handling of UPS/FedEx/USPS/DHL/Whatever Shipper Inc.
Net result of this trip? 60 pounds of cells that needed to be sorted through. Talk about a time
consuming process...but the yields are still good. Each good cell is capable of about 1.5 watts.
And the best box I've sorted thus far had a survival rate of over 90%. So at this point, I had
more than enough cells to work with; it was time to draw up blueprints for a prototype panel.
Head to Page 3 for the info on the good prototype (as opposed to the lousy one on the first