Prototype panel is finally finished!

[Jeff7]

Link to the site.

I've got a writeup there, and a decent amount of pics. All thumbnails are fairly small, but the big pics are, for the most part, saved at a high quality setting, so they're not only big, but they have minimal compression artifacts. But of course, this also makes some of them big downloads, so be aware of that.

Thus far, the panel has sat outside for 4 days and nights, and has had no condensation at all, at least not on the inside. I did keep it in a freezer overnight once, and when I got it out, it formed a layer of ice really quickly, but that was only on the outside.

There's a storm coming tonight, and I intend to leave the panel out to endure it. If these things are going to stay outside, they'd best get used to some abuse.

One odd thing I've noticed - the epoxy I used turns bright yellow after staying in the sun for awhile. Hopefully its adhesion properties aren't affected.

[Jeff7]

Oh yes, output info of the finished panel:

2.10V, 2.61Amps in peak sun. The cells I used were from a pile I set aside while sorting - ones which weren't quite normal, or cracked at places. Basically, I didn't want to sacrifice any top-notch cells on a prototype.

The polycarbonate caused a 9.5% drop in output.

[pyrocasto]

I think I'm gonna try the same thing with chips of cells. I found some conductive metal tape to put them together.

[old55olds]

I had thought of epoxy too. But UV resistant epoxy was not available in this town. On talking to others I was told that urethane would be preferable as it stood up to UV better.

ken

[Volvo farmer]

Looks nice. Very professional.

Here's something I found the other day.

http://www.eets.co.uk/Manufacturing03.htm

So what I want to know is what materials are they laminating with these? The top thing is probably tempered glass, but what is the bottom material and what are they pouring or heating around the cells stick it all together?

Somebody buy one of these and let me come over and borrow it for a day will ya?

[wooferhound]

I used 30 year Calking adhesive

[Jeff7]

What kind of metal tape? Is the adhesive side conductive too? I found some braided copper ribbon tape once, at Electronic Goldmine, but I found that it had a fairly high resistance for just some copper tape - over an ohm per inch.

It would be nice to find a good way of using the beat-up solar cells; I've got quite a few of them, and I'm not even done sorting the mass of cells I bought. I'm guessing there's about 15 pounds of cells I rejected, either due to cracks, or because they were outright shattered.

[old55olds]

Apparently they use Tedlar. The back seems to be a silicone encapsulating compound.

Ken

[Jeff7]

Well, good news - the panel was outside, and got an inch of snow on it, and sat through some high winds. This morning, the sun hit it, and the snow melted off - and there's still no condensation or moisture visible inside it.

[tecker]

I use Buss wire It's pre tinned and wouldn't take up a lot of room. I've been tring to get artsy with the pieces something that is out of my scope.I suprised my self to find that I'm even less adept than I though.  

[pyrocasto]

1ohm per inch! This says it has .005 ohms per inch, but I'm still looking for more.

http://www.emsdiasum.com/microscopy/products/sem/conductive.aspx

It's little more than half way down the page.

[Jeff7]

Maybe I should have posted this in the Diaries category. Oh well.

The panel has survived complete submersion for several hours, with no leaking, and still producing power all the while.

Next day, I subjected it to 88C (190F) temperatures, by setting it on top of a wood pellet furnace. The epoxy got a bit soft, as did the caulking. The polycarbonate top bowed out a little bit too, but not a single seal broke.

However, condensation did finally form on the polycarbonate. I did use plenty of helium in the panel, though some water might have remained around the perimeter.

I did get a bit cheap and impatient though, and for the perimeter spacer, I used a wooden dowel instead of aluminum. That may well have served as an inefficient sponge, and the high heat might have baked the moisture out. When the panel was tilted, and the water hit the dowels, it just disappeared right into them, bolstering the theory that they were rendered extremely dry by the heating.


At least, that's what I'm hoping was the case. This design, while probably the most expensive of the designs on this forum, also seems to be quite robust, able to withstand extreme thermal stress (-15C to nearly 90C) without exploding or leaking, and also able to endure direct hits by 1"+ ice cubes.


I've never been a good scavenger, nor have I had "connections", so I've had to go the normal retail route to acquire my parts. I have yet to draw up plans for a full-scale panel - once I do that, I can determine the cost of producing it. All I need now is some time, and a large, flat area where I can lay out a big sheet of paper to do an accurate scale drawing. And I'm lacking both. Some day though, I hope to see this thing through - and hopefully, it'll be sometime before commercial solar panels reach 25 cents per watt, or else I'll have a lot of expensive, brittle wind chimes.

[Jeff7]

Moderators, if you want to move this thread to the Diaries section, go right ahead. At this point might be more relevant there.

Looks like I've hit a snag with production of larger panels. I decided that 24V panels will reduce the amount of work - fewer panels to seal up - and also make wiring easier. Problem: they're big. Onlinemetals.com, which has good prices on sheets of aluminum (they also have polycarbonate, which I haven't compared pricewise to anywhere else yet), only sells sheets up to 48" long - at least, that's all that they can ship by UPS. The sheets I'd need would be either 28" x 59" (4x18 cells), or else 41" x 41" (6x12 cells), which seems that would also exceed UPS' maximum shippable size.

Net result - if I want the larger sheets, Onlinemetals.com can provide them, but:

"Please note that with pieces of the size that you are looking for, we

cannot ship via UPS - they must be shipped common carrier.  Assuming

that you just want one or two of each size, your shipping cost for 5-7

day service is approximately $75.00."

Now I assume that more pieces wouldn't substantially increase that cost, but I'm not sure. That, and when I sent the e-mail, I misplaced a decimal point on my request for aluminum. Oops.

I checked some local places for aluminum, but their prices are so high that paying for shipping at Onlinemetals still puts me ahead.

Guess it's time for a dreaded cost/risk analysis - fewer expensive 24V panels, or 2x more 12V panels, with more wasted space (about 5cm on each side around the perimeter) and more risk of an improperly sealed panel.

[Jeff7]

This update comes bearing a public service announcement:

The Evergreen solar cells do not like being wet. After a day under water, the silkscreen lines on top will start to peel off.

Now for the back story.

After the panel sat on the stove and endured that heat, I put it back in the sun for a few days. I also put it in a pan again, for an extended submersion test.

I didn't like what I saw - it was taking on water. I let it fill up nearly halfway. It took about 4 days to do it, as these were very tiny leaks.

After awhile under the water, that's when I noticed the silkscreen lines beginning to peel off of the cells.

Some things I found out:

The polyurethane foam doesn't seem to be working out. Either it cracks under the stress of extreme heat (not good), or else it simply doesn't bond well to aluminum. That was one factor that allowed water in. So it won't be used in the next design revision.

Second factor: the epoxy I used is fairly inflexible - regular 2-ton epoxy. It too likely cracked as the panel expanded, and the polycarbonate flexed outward. So I either need to find a way of completely immobilizing the various components, including neutralizing thermal expansion (not really possible without changing a few laws of physics), or else find a better adhesive.

Devcon's site lists the Tensile Elongation of 2-ton epoxy as 1%. Definitely not flexible at all.

Devcon has something called Flex Welder, which has a TE rating of 130%. They've also got Metal Welder, which has a TE rating of only 50-75%, but it's also supposed to be better than the Flex Welder at bonding to both aluminum and polycarbonate. Metal Welder's spec sheet also claims "Excellent environmental resistance".

I guess I've got to figure out where I can get that Metal Welder, as it sounds like some pretty capable stuff. Hopefully it doesn't carry a prohibitively high cost with it.