This is an old project, but I have not shared it widespread yet, so here goes.
The city where I live has a "Citywide Curbside Cleanup" every year, but I refer to it as "Junk Days." I have seen people come from neighboring states, with trailers and spotlights, through all hours of the night going from house to house and rummaging through the unwanted refuse left on the curb. These items are things that were not allowed to be put in your normal trash or recycling, so typically larger items like appliances, carpeting, lawn tools, etc.
One of my favorite finds was a 4ft fiberglass satellite dish (reuse). I had built solar cookers as a child using cardboard and aluminum foil. I knew this dish was the correct shape for the task.
Step 1: Collecting Resources
I know from experience, that it is difficult to get aluminum foil to adhere without creases. This dish surface was also not smooth, it was textured, so aluminum foil is not the optimal reflecting surface to use. Being an amateur telescope builder (see my Pseudo-Nasmyth mount), I am familiar with the Keck telescopes and their segmented design. I have a couple of mirror lenses for cameras and also a Schmitt-Cassegrain telescope, so I am familiar with a primary mirror with holes in the center. I also knew that hard drive platters were made of glass or aluminum and coated with a very fine surface for the heads to float over.
I work in the IT industry, so I know a few people with old computers. And they know people with old computers. Everybody upgrades and few people dispose of the old ones. So, I started collecting old hard drives (minimizes e-waste). It takes a long time to disassemble them without damaging the disks, but after you do a few dozen, it gets easier. I turned all the old stripped cases back into recycling for cash, and I kept the Neodymium magnets for other projects (upcycling). I collected stacks of hard drive platters.
Step 2: Layout and Adhesion
The surface of the dish is textured and curved. This means a straight layout in a hexagonal pattern would not work. Some size adjustments would be necessary. I was unsure what type of adhesion to use, but I knew I did not want to screw them on. If the glue was strong enough, hopefully the adhesion would warp the platters to increase the efficiency of the focus. So, I tried contact cement. It went ok, but did not hold as I anticipated. Next I tried RTV silicone (automotive), and it worked well.
Step 3: Complete Layout
The rest of the dish was done with standard white tub caulk. The hex pattern worked to some extent, and the areas where overlap or shortages occurred, I substituted different sized platters. Most of the platters are 3.5" from standard hard drives. A few are 3.25", some 2.5", and three 5.25" platters. Total on this dish is 149 hard drive platters. I also had to change the alignment to stay clear of the four mounting bolts in the center of the dish.
Step 4: I Like to Move It
The Sun moves through the sky, so aiming the dish correctly to optimize cooking will take some frequent movement. The ultimate goal is to make it automated. For starters, moving along the horizon (azimuth) and up/down (altitude) was the way to go. I purchased a linear actuator for the altitude and an old 12v motor with a gearbox (like a windshield washer motor). I found gears to fit the bill and a chain drive for the azimuth. I cut some stock steel and arc welded it together. A quick coat of primer and paint and it was ready to go. It is shown mounted on my Meade LX200 Giant Field Tripod.
Step 5: Feeling Hot, Hot, Hot!
The time of year is March. There is still snow on the ground. But it was time to experiment. These are photos from its first setup in the sunlight. I first tried a wooden shim, which quickly ignited. I had a touch-free laser thermometer, but its upper limit was around 1000 degrees F. Look at the reading, there is no decimal point in there. Awesome!
Step 6: Cooking
The logical next step is cooking something. The first attempt was two hot dogs. It took a little longer than I anticipated, but there was still snow on the ground. The reason for the slow cooking was the color of the item being cooked. Darker objects absorb the sunlight faster, so new attempts would use darker materials. I think marshmallows will take quite a long time.
Step 7: I Am in Control
I was using the bare wires and holding them against a 12v battery to position the dish. Not a dangerous process, but it would be easier with switches. I purchased some Double-Pole, Double-Throw switches and cut a CNC panel to control it for the future. This then had just one set of wires going to a single 12v battery (or 12v power supply). The 'auto' label is for future solar tracking hardware that will auto-align the dish to track the sun.
Step 8: Summertime
I took an old tin can (dog food, vegetables, who knows?), and painted the outside black. I filled it half full of water, and in less than 10 minutes it was at a full boil. Here is the video!
Step 9: Burgertime
I found an old 6" cast iron skillet. Very black. I held it in place with a couple vice-grips, and it reached a consistent 750 degrees F in the sunlight. Frozen burger patties will cook to well done in about five minutes. Yes, I ate quite a few burgers that day.