Join the conversation on the friendliest most helpful DIY solar energy forum
Found an interesting article about a guy building a cheap solar powered refrigerator using a Peltier junction for $35. I know they have ice chests like this already as I have one.
It's just cool to rig one up yourself. I've picked up Peltiers on Ebay before for spares.
I noticed that many people on this blog want to store milk, vegetables, seeds and medication in a refrigerated atmosphere but have yet to buy 2 refrigerators. Remember the “two is one and one is none” motto; you need 2 ways (or more) to cool your food. You could make a Zeer pot, like I did but there is too much humidity and the pots aren’t big enough for 2 gallons of milk.
I can’t buy another (let alone storing a 19 cubic footer) $700 fridge as a spare. The other problem is – I might not have AC to use in the future, regardless of all the solar panels I have. So, I needed a small DC powered fridge that will hold at least 2 gallons of milk and some other stuff.
I thought of what MD Creekmore would do – a “Do It Yourself” project. (Mr. Creekmore has a propane fridge but, even though he is a great inspiration, I want to limit my power needs to solar since, that is the only long term solution and also because I have solar) Just like the “MDC made” water filter or electric generator – a person can make a refrigerator. Of course, there will be some technical discussion of the actual properties of said fridge. It will be a bit more involved than the water filter project and, for good reason – you want something a bit more complex.
My most important consideration is that it can be service-able by me, so, the fridge is not going to use Freon. I am going to use a “Peltier junction” cooling unit – already built for the purpose. These units are less efficient heat pumps than an evaporator/condenser cooler but are solid state, small, low power and can be ganged together for more cooling power. Did I mention cheap? At only $35 per unit – you could buy 2 and have one for a spare.
I added fans to the peltier heat sinks (see photo) and they are the only moving parts. They are blowing air “into” the heat sink – an important point here; there is more exposure to cooling air than trying to suck air “out” of the fins. You can have lots of spare fans and they are easily replaced. I wedged some round head screws between the tapered fins to mount the fans on each heat sink. The power cord goes to a car cigar plug (negative ground) with a 10A inline fuse.
For anyone who is not familiar with the peltier device – it is a heat pump that is made like a silicon chip. A DC electric current will get the chip to “pump” heat to one side of the chip leaving the other side cold. Heat sinks (and fans) help to distribute the hot and cold.
The second part is the “containment box” that would consist of the good ol’ beer casket – where they get an icy bath until they meet their drinker. Yes, a cooler, the cooler I selected is 18″ x 10″ x 10″ (1.07 cu. ft. approx.) about 20 quart. A cooler has all the “fridge” properties like waterproof food grade inner walls, foam wall insulation (except the lid), rugged dent resistant outer walls. The other benefits of the cooler are: light weight, portable and you probably have them lying around in a corner from all the camping trips. This cooler will actually fit 3 gallons of milk but the 2 side containers will touch the walls.
Now, a cooler can be expected to work as good as a cooler was designed to work – it melts 2 bags of ice keeping 12 beers cold for a couple of hours. I figured that I would improve the efficiency a little bit. I also figured that this could make for a great article so I wanted to gage any progress I made. I decided to use a 7 dollar temperature monitor with remote sensor, which eliminates the “open fridge door” variable in checking the temperature inside. The sensor was taped to a plastic box (see photo) to keep it off the “floor” of the fridge so that air temperature only would be measured.
Now, what to check, since I am designing a fridge; the “cooling cycle” is the metric to gage. This just records the temperature as the fridge runs – how fast it cools. The other metrics are – lowest temperature and the “warming cycle” which is keeping track of the temperature after the cooler is shut off.
As I add insulation I will keep track of how the numbers stack up. The lowest temperature is the temperature when loses (heat gain) are equal to cooling effort. The cooling effort is (approximately) 60 Watts – that is what the heat pump and both fans consume. (A person could mount 2 cooling units which would double the total cooling power)
The first task involved adding the peltier unit to the cooler. The smaller top heat sink has 2 “fins” that slide out from the grooves on the side. I cut the hole (to the size of the smaller, top heat sink) into the plastic (see photo) with a metal Stanley cutter; the cover is really thick in some areas.
I tested the “bare fridge” (see graph) and found that I could only get a 17F degree drop in temperature. The graph shows a fast rise after the heat pump is turned off and the lowest temp (47F) does not really work for keeping milk cold.
The second test involved adding a cheap space blanket (see photo) to the outside of the cooler. I used cellophane tape to attach an old (12 year) Mylar space blanket to the top and bottom of the cooler. The improvement was noticeable; the test ran longer as the heat pump was shaving 0.3F and 0.2F bits for a while. As long as the temperature kept falling – I would run the cooler.
I, next foamed the cover (see photo) and then tested (see graph) a third (and fourth time), the low temp was now down to 40F deg. – good enough for milk. The fourth test was exactly like the third except for shorting the peltier leads together at shutoff, this extended the cool time about 35 minutes at the low end.
Doing this made the peltier junction “fight” the temperature change with its own electric generation thru “temperature difference”. (This efficiency improvement would be – adding a switch to shut off/short leads) Adding expanding foam to the cover was the biggest improvement made, so far, and the cost was small – just 4 bucks.
The poly-urethane foam did not stick to the poly-ethylene cover so clean up was easy. The cover is easy to foam after you have cut the hole in for the heat pump. All the later trials do not vary in the first hour of testing – proving that the foam insulation was a big factor for the fridge.
The next step was to replace the old space blanket with a much better insulation solution. The blanket was interesting but, contains no insulation, just pure heat reflection and with only an 80% rate. I first added aluminum foil (used for grilling) to the top and bottom of the cooler. I wrapped a heavier space blanket around the cooler then doubled the blanket over a fiberglass batt wrapped around the cooler. The space blanket will “seal” the fiberglass in a sandwich and also serve as a double reflective layer. I used duct tape to hold the foil against the sides and for the blanket seal. The foil gives 100% heat reflection but tears too easily, the blanket and insulation “press” the foil against the cooler wall and protect it.
The testing (see graph) shows that the “NASA cooler” can now drop down to 32 F. It also took almost 4 hours to warm up to 60F. A side benefit of the added insulation is that sound levels are down by half, can’t hear the fan inside.
This fridge is meant to be run on solar power – just one sixty Watt panel (direct connected) will provide all its power during the day. At night – one can use a battery or plastic “ice blocks” to help keep the cold temperature. A 100 Watt solar panel would provide enough power for the fridge and a battery.
I do not have temperature regulation – peltier cooler does not stop at a pre-set temperature. Right now the cooler cannot go below the freezing mark (32F/ 0 C) no matter how much it runs since it does not have enough power but regulation could be added if saving power became paramount. Remember this fridge was home made and can always be upgraded.
I only measured the air temperature (see chart) inside the fridge; milk (and all other things) will take longer to cool down, maybe 6 hours (per gallon of milk) with this unit. It will also take longer to warm up – a cold gallon of milk will help chill the next gallon of warm milk.
I could reverse the cooling into heating by adding a small “bridge diode” to the fan leads and putting in a reverse-able coupling. The peltier unit comes with a connector – you would need to find the cord to fit the connector. Of course; I already have a solar cooker.
You could go out and buy a cooler with the peltier unit inside although I think that it would cost a little more then my unit. It might be a quite a bit less efficient (no space blanket heat reflection) and the parts replacement will be a question.
———- I had this laying around too)
I know that this fridge does not look “commercial” and I would not blame you if you bought a nice $1400 “Sun Cooler” DC fridge but if your Sun Cooler gives out – you now have the skill to convert a cooler to a fridge. Well, what do you think?
I think, if I wasn't determined to be ice less at camp I could get away with a 60 watt system.
That is a heck of a write up/post. And I salute your determination, recycling, & effort.
I picked up a PowerChill cooler off craigslist for about 50$ and have been using it over a year now. It works for weekends & extended stays. Imo to much power draw at 4 amps constant use on smaller solar systems. When its over 80 degrees I have to run it more then 1/2 the day. I use an AcuRite Thermometer to keep an eye on the inside. Thermoelectric coolers for what they can do are not true refrigeration. But there cost, or, this cost is attractive compared to an Engel or Edgestar.
I would typicaly vacation at my camp in the cooler months. And so then the cooler wouldn't have to chew as much Ah reserve. Even though I have taken to running it mostly durring charge hours off the load side. Still when I'm there for more then a few days I have to be prepaired to keep two extra coolers & make ice runs into town.
Or I have three choices; Consider a 12v/120 edgstar (500$), double my battery bank (600$) Ice runs and soggy food. Cool.
To controll the temp rob an old fridg of the thermostate, it should handle 60 watts, even at low voltage.
Also, just for your comparisons! I have an old bernz-o-matic ammonia fridge. Some ran on L.P. Gas, this one runs on 12 volts D.C. or 110 A.C. . The entire cooler is about 20 inches square and tall, the compartment is 18 deep by a 18 by 10? Those are approxaments! It has small ice trays and will keep the beer cold and make ice! No fan the coil sets high and gravity circulates air. Insulation is definitely lacking. The biggie to you, it runs on 60 watts AC or DC. So your numbers ore looking good to commercial built.
I don't know why my I-Pad posted the picture an all of that post above--it was supposed to just have been the link itself..oh well...I'd sure like to find one of those old bernz-o-matic ammonia ones. Used to have one at lake house that ran on LP--Servel was brand.
Reading some more on cheap refrigeration for solar panels, I found some interesting links on converting a chest freezer into a fridge...one guy's conversion brought refrigerator usage down to 0.1 kWh a day running about 90 seconds per hour.
The concept is rather intriguing
Check with motor home, trailer dealers they used to be 12/110/LP and a take out can't be worth much.