Non-PV Solar Power




Edit: Sweet, I'm on hackaday!

Hello, in this Instructable I will show you how to generate solar power using inexpensive solid state parts and without PV (photovoltaic) cells or panels. I wanted to make this project to see if I could beat the dollars per watt costs of commercially available solar panels.

You only need a few inexpensive parts to build one. It is pretty easy to build and very simple. As far as I know, my idea for the combination of thermoelectric coolers and Fresnel lenses is original... but it's probably not.

Sorry about the rubber bands and foam but remember this is an experiment on my part. Rubber bands are just the best thing for me! Likewise, I recommend you also build a prototype and if (when) it works you should go on to more fancy enclosures. Remember this is just a rough guideline and you can modify it however!

How it works:
This generation method uses a Peltier cell to generate electricity. Peltier cells are designed to be used as heat pumps. When you apply power to a Peltier cell, it begins pumping heat, and one side becomes cold and the other, hot. However, you can do the opposite and generate power from a temperature differential on the sides. To generate this differential, a Fresnel lense focuses light onto one side of the Peltier, and it becomes hot. The other side has a heatsink attached along with a fan that is powered by the Peltier.

Lenses concentrate light, which is absorbed by everything, and when light is absorbed, heat is created. You can easily burn yourself and other things with the lens. Don't leave this out uncovered because the sun moves and it might focus on something. I disclaim all responsibility for anything ever.

Step 1: Parts and Tools

For this project you will need:
a Peltier cell (I got a 50W one for $5 off of eBay, they are still there)
a Fresnel lense (mine's about the size of a sheet of printer paper, there are a whole lot on ebay)
a CPU heatsink (any type of heatsink will work, as long as it's big enough. I got mine at a garage sale for $2.50)
a 5V fan (more info on that later)

Tools/other materials:
Thermal grease
Black paint (I used semi-flat spray paint and it didn't burn)
Soldering stuff
Wire cutters
Reflective material

Note on the peltier: The wattage doesn't affect the power output, but it does affect the internal resistance, which in turn affects the voltage output. Higher wattage = lower voltage, higher amperage. You want a lower wattage peltier so the voltage output is high enough.

My dad built a DIY projector and one of the Fresnel lenses he bought was scratched so he gave it to me.

Warning about the Fresnel! Don't go much bigger than the sheet-magnifier size because you will fry your Peltier. We are probably already running it over it's heat limit, which means it may not last as long.

We have a bunch of foam lying around from building RC foamy airplanes.

The heatsink used to have a fan on it but it was 12v and barely ran off of 5v, so I hunted for 5 volt fans. I was impatient and couldn't find one that I could purchase locally, so I went to Fry's electronics and picked up a laptop cooling pad for $10. There is another one there for $5 that will work fine. It had two fans built in, so I took out the 6 screws, pulled it apart, and pried off the two fan blades/circuit boards. As you can see, they were smart and made the fan axles built into the case instead of standalone fans... so I outsmarted them and dremeled out the axles! You could go my route and use a band/jig saw or dremel, but I highly recommend buying a 5v fan online to save yourself the trouble.

Step 2: Prepare the Peltier

First, you need to get your stuff ready. You want to have maximum thermal conductance between the non-hot side of the peltier and the heatsink. You should first remove any mounting equipment the heatsink might have, and while you are at it if it has a fan, take it off too. It might come with a bit of thermal goo on the bottom. This probably isn't the right size for your peltier. Go ahead and scrape it off, and then clean the surface of the heatsink with alcohol and a paper towel.

Put the peltier on some newspaper and put a coat of black paint on one side of it. I used semi-flat spray paint, when I did the first coat it congealed into blobs on the surface, but the second coat covered all of it. I did put on too much though and it's a bit thick.

Now, put enough thermal grease onto the peltier to *just* cover the whole thing. You don't want to use too much.

Figure out where on the heatsink you want your peltier, and stick it on. I recommend putting it in the center. Stick it straight on and wiggle it around a bit to make sure the grease is spread properly.

Step 3: Attach Things

Now you will want to firmly attach your peltier to your heat sink. I used iron wire to hold it on and keep it from sliding as you can see, but really, any thin material that can resist the high temperatures and you can wrap around will work fine. Gluing or taping is not recommended because it will not stick to the smooth heat sink surface.

After this you should attach the fan. What I did is like I did with the "baseplate" as you can see in the next step. I put rubber bands on the base of the fan and around to keep it on. I had to put a bit of cardboard between the base and the heatsink fins to keep the fan blades from hitting the sides of the heatsink. I bent the cardboard so part of it would slide into the fins and this worked perfectly, almost holding the fan in on it's own!

Step 4: Frame Creation

The first thing I did before making my frame is I cut a baseplate from foam that surrounds the Peltier and goes on the bottom of the heatsink. Measure the size of your peltier, and cut a hole in a piece of foam as big as the bottom of your heatsink. Make sure to cut holes for any wires or anything to come out!

I had to attach the baseplate with rubber bands, which works fine for me. I recommend, if you can, using any mounting hardware that is usable. Just make sure the bands don't go over where the light will be focused! The baseplate is important because it's really hard to mount the peltier/heatsink anywhere without it. Glue/tape just doesn't stick to heatsinks... It also makes it so you can put a reflective box around the Peltier. We'll talk about this once we've made our frame and figured out our focal point.

Now you will want to make a frame/enclosure for everything. In the pictures you can see what I did, feel free to improvise. There are many ways to make one. You could make a box, or a foldable stand, or do what I did, which is sort of a combination. The strategy is the same for all of them. Figure out the rough range of measurements between the peltier cell and the lense for the focal point, because you will want the peltier-lense distance to be adjustable.

Don't measure the focal point where the light is at the smallest point, that will just burn off the paint. Flip around the lense until you are pointing the side at your "measuring surface" (I used a napkin so it wouldn't burn) that creates a square sized spot, and measure it where the spot is a bit smaller than the total size of the Peltier - remember, there are two points where the spot is the size of the Peltier. Use the farther one or the reflective shroud will actually block light!

Once you have figured out the focal-point range, cut your frame. Mine has a slit that you can move a foam panel back and forth on, to adjust the focal length. The light-collecting assembly sits on this. See the photos for more details.

One thing to remember when making your frame/enclosure etc. is that heat management is EVERYTHING, because this generates using heat. If somehow the heatsink is getting heat that is not through the peltier (maybe light is falling onto it) then you will be losing a lot of power.

Step 5: Finishing Up

Now that you have your frame and everything built, we will make that reflective shroud I mentioned. I did this by cutting four pieces of foam that were half as wide and a bit longer than a side of the Peltier, gluing them on in the formation that you can see, and double-sided-taping aluminum foil to the inside. Hint: Mylar from the inside of potato chip bags is a better reflector.

I actually extended it so I could put a plexiglas window to trap the heat, but it was right at the focal point so it distorted the lense... maybe I'll drill a hole.

To give the fan some extra power, I wired together a 3v 150ma solar panel for $3 and used skewers to hold it up, and put it in series with the Peltier. Power for usable applications can still be drawn from the + and - of the peltier because the solar panel is just for the fan.

I used springy-wire splicy things to connect the wires. The circuit has the peltier and solar cells in series to generate ~5V for the fan.

Step 6: Notes

From my setup, my highest recorded voltage is exactly 2.7v. The internal resistance of the Peltier is ~3 ohms because at 7.5v input it draws ~2.5A. This means I could get 2.7/3 = .9A out, and 2.7*0.9 = 2.43W.
So, now I will figure out the cost for the average person to build it.
Heatsink = $5
Peltier = $5
Lens = $5
Fan = $5

Solar Cells = $3

Total: $20-25

This is equal to right around $10/W, which is on par with the cost of commercial solar panels. Pretty good for an untested experiment. However, some disadvantages of this are:
-My enclosure is too flimsy to be held vertical so it can harvest power at the middle of the day
- Needs to be well aimed to be effective
-Less efficient than PV

Yeah, this technology probably isn't ready for prime time. But it is interesting and very fun/cheap to make.

One thing you might be wondering, is why is the power output so low?
Well, first of all, the wattage of a Peltier is not the power it will output. It is the power it will draw at it's rated voltage. Another thing, Peltiers are not designed to generate electricity. They are optimized to pump heat. To really take advantage of the energy available, you will want to get a Seebeck unit. These are the same thing, but optimized to generate. With a seebeck unit, power output would go up by many times. Same with price.
So, I will continue working to improve the output. There are a lot of things I can and will be trying. Make this project your own!

Here are ideas I had:

Use something like the LM2623, LM2371, Mintyboost circuit, or LVBoost to get this up to 5v to charge/power USB devices and Li batteries. You probably will have to parallel a couple of whatever you use to get full output and decent efficiency.

Use one of the projector-TV fresnels and focus it onto a 3-by-3 grid of Peltiers. Put them in series and no inefficient boost converter needed.

Build a solar tracker to harness the output all day.

Get sheet aluminum about 10" by 10", put peltier+heatsink in center, paint other side black, insulate it and put plexiglas over it and it will be a lot like PV.



    • Pets Challenge

      Pets Challenge
    • Backyard Contest

      Backyard Contest
    • Sensors Contest

      Sensors Contest

    103 Discussions


    1 year ago

    Couple of thoughts on economics:

    (A) I'm wondering (although I don't think so) whether it's not just cost-per-watt when comparing to PV, but also dealing with area. If you need more square (feet/meters) to generate the same wattage, that's also a possible 'cost' to factor in.

    (B) More significantly, you'd for sure have higher up-front (capital + installation) costs. Not only would lenses (or parabolic mirrors) have significant mounting requirements, but they'd need to track the sun - which means motors and joints and things. Not a criticism - I love the idea and understand that your point was just a general experiment! - just an observation.

    (C) Finally - a Peltier module is not a Seebeck module - you want a "TEG", not a "TEC" (thermoelectric generator vs. cooler), more or less. Peltiers aren't as efficient when run 'backwards' to generate power from heat. (Quick info here - and I have no connection to this company:


    2 years ago

    Keep going with this, a very good instructable. You are on the right track.Use a lens not a Fresnel lens and implement all you other ideas. Large panels are not good in high winds and helistats are too complex and fragile.


    2 years ago

    can you test a satelite dish with a mirror finish to be able to generate power at wider range of light angle? or you could try putting the peltier on a tower and have a circular array of mirrors on the ground. on TV years ago i saw an experiement in nevada desert where they used a large array of mirrors to melt a solid compound into liquid which then generated steam power. Does anyone knows have knowledge of this exeriment?

    1 reply

    Reply 2 years ago

    The mirror array is called Heliostat. There are many versions of it.


    3 years ago on Introduction

    perhaps you could group the solar panels around the Peltier, so you get some concentrated light from the Fresnel lens. I love this project though!


    3 years ago on Introduction

    Vertical are already marketing photovoltaic panels is up to 20 times more efficient than a conventional solar panel and takes up much less space. A major revolution in the production of solar energy, especially for those who have little space available.

    The first manufacturer to sell this system is called Sun-Säule.

    More information


    3 years ago on Introduction

    Ya están comercializando paneles fotovoltaicos verticales, es hasta 20 veces más eficiente que un panel solar convencional y ocupa mucho menos espacio. Una gran revolución en la producción de energía solar, principalmente para los que no tienen mucha superficie disponible.

    El primer fabricante en vender este sistema se llama Sun-Säule.

    Más información en


    3 years ago on Step 6

    Maybe you can use heat sink for home heating in a winter.


    I would like to see someone build a 3-stage solar power generator: a photovoltaic cell connected to a peltier/seebeck unit connected to a stirling engine. Actually, the peltier/su could go on either end of the stirling engine. I wonder how much power one could get out of such a setup.

    2 replies

    Reviewing you comments of 5 years ago, I know that the back of solar panels gets very hot and that heat diminishes the output of the panel....great idea to use the back of the panels to generate power and thereby cool the panels.

    I know I'm replying to myself, but oh well. I had a rethink of my idea, and came up with this setup: take an old satellite dish, line the concave side with small photovoltaic cells (1st power source). The pv cells, being shiny, would reflect some light, and being mounted in a satellite dish, would focus the reflected light onto a seebeck unit (2nd power source) which is mounted on the stirling engine (3rd power source). Would this not work?


    5 years ago on Introduction

    I don't know much at all about optics. Would it be feasible to use a hook up a series of lenses and heat pumps into an array? Would there be any advantage to this? (I'm new to the DIY-energy community.) Along with that, would it be possible to use a more more concave lens, closer to the Peltier cell?

    1 reply

    5 years ago on Introduction

    im getting interested in peltiers as backup, not replacement to solar.i think your math is off tho, you stated $10/watt is on par with solar. i can get solar for .70 to $2.00 a watt (depending on panel manufacturer) anywhere (ex. home depot 105 watt 12v panel for $159.00= approx. $1.60/watt)


    8 years ago on Introduction

    Just a thought, if you use a peltier cell and a freshnel lens to generate electricity and you have a bank of them in say a 1m X 2m area, will the set up generate more or less electricity per given area compared to solar PV's?

    1 reply

    Reply 7 years ago on Introduction

    Depending of the elements you will be using -You can expect more or less 1Kwh per 1m X 2m area of peltier.
    But the problem is - You willl need a pretty BIG freshnel lens or BIG solar oven and 1m X 2M heat sink too.... that's a huge thingy.


    8 years ago on Step 6

    Nice job - I've been thinking of heating a peltier element with my solar oven to generate a little electricity, but haven't tried it yet.