Introduction: Recycling CPUs Processor Heat

Picture of Recycling CPUs Processor Heat

This Project involves conversion of waste heat released by the processor of the computer to Electrical energy using the peltier Effect.
The electricity is generated by the peltier tiles which are placed across temperature difference between the bottom of the laptop, which is hot and the ambient air which has a moderate temperature.

I chose to do a research on the peltier effect as I had earlier heard from people that a lot of heat is wasted which is generated by the processor. Today we are mostly using non-renewable sources of energy to generate electricity which will be depleted in the next 50 years. So I decided to recycle heat generated by the computer processor as I had studied “Law of conservation of energy” which says that energy can neither be created nor destroyed but can be converted from one form to another.
In the peltier module, when one side is heated and other is cooled, through the temperature difference it generates electricity.
I had calculated the area of bottom of the laptop and calculated the area needed to recycle energy. My project has generated about 1.2V en the first build. This voltage was enough to run a USB computer fan.
I have took 12 peltier tiles and placed them on a top of a Gel pack. On above the peltier tiles There was the bottom of laptop.

If this technology is further developed, we can save over 50%-70% of the battery life and thus save electricity.

Question / Proposal-
Hypothesis: If I can capture enough heat from the processor of computer and convert it efficiently to electricity, then I can save the battery lifetime.
Objective: To make a processor heat recycler that generates electricity using the heat generated by the processor.

Step 1: Technical Design

Picture of Technical Design

Now, we will see the technical design of this project.

Firstly, We need a heating source and secondly we need a cooling source. In my case, the laptop's processor is the heating source and the Gel pack is the cooling source.The electricity is generated by the peltier tiles which are placed across temperature difference between the bottom of the laptop, which is hot and the ambient air which has a moderate temperature.

When the heat is transferred from the laptop to the gel pack, electrons flow and generate electricity.

Step 2: Materials

Here is the list of materials you need for this project-

1) A Gel Pack

2) TEC Peltier Modules [TEC-12704 or TEC-12706]

3) A set of wires

4) A Laptop

5) A USB computer Fan

6) A laptop Pin [Optional]

7) A Multimeter [Optional]

Step 3: Construction of the Circuit

Picture of Construction of the Circuit

In this step, we will join all Peltier modules together and connect everything together.

Now place the peltier tiles on the Gel pack. Connect them using wires in series. The peltier tiles should be placed in such a arrangement that it covers all the area of the gel pack.

Step 4: Connecting the Charger and the Fan

Picture of Connecting the Charger and the Fan

Firstly connect the voltmeter to check if the current is flowing.

Now, Connect the laptop pin and the fan to the circuit. Make sure they are in parallel to each other.

Step 5: Testing the Project

Now, place your laptop above the peltier tiles just as shown in the video. Now your fan would run.

You can also charge your laptop by connecting the laptop pin to the circuit.

Congratulations! Now you are one of the person to successfully recycle computers processor heat and turn it into electricity.

Step 6: Results & Conclusion

Picture of Results & Conclusion


The results prove my hypothesis, I had tested the project when the temprature of the laptop was high and when it was low.

Temperature Difference (T1 - T2) 17C 25C 35C 48C 58C 79C

Voltage 0.76V 1.2V 1.9V 2.2V 2.4V 4.02V

Now i calculated that how much efficient was my project to save power. So i found that it saved about 30%-40% of the laptops power.


In conclusion, I succeeded in generating electricity from the heat released by the processor.

I have proved my Hypothesis that I can capture enough heat from the computers processor and convert it efficiently into electricity.

I have tested my project several times and this is the proof


Main power supply-
100% in 4 Hours --1

Recycled Heat supply-

36% in 4 Hours --2

Adding both (1+2)-

136% So 100% charged in 2.7 Hours

So I have proved my Hypothesis





Official Website- .


uhvr (author)2016-10-07

lol g3t r3ckt

nikhilct (author)2016-04-27

a student cant get ur project copy....for reading ur innovative ideas because of ur payment.....sir so i kindly request u to give ur papers free for poor students as open source will help sir... please its my humble request....

sathish nirmal (author)2015-08-10

can i know the cost of tech-12704 peltier module

Just4Fun Media (author)2015-05-26

Great instructable! I built a similar system that is self powered off of the computers waste heat! And I was able to reach core tempatures of 4.2 degrees Celsius!

MRCmegarallycar (author)2014-06-09

Sure he did something but he did it wrong and refuses to listen to anyone else.

And what was done wrong? All I see is a great idea that worked and I am now determined to figure out how to use. Wattage would be a nice thing to see, but please don't turn offensive for any reason.

How can you say this. Can you prove me worong??

th3tguy (author)2015-01-03

Could you tell me where you got all the peltier modules, and for how much? And I'm sorry for sounding a little dumb, but could anyone help and explain how he connected the wires together and to the fan/computer?

Master_Squid (author)th3tguy2015-04-06

Looks like the peltier modules are in Series. Meaning the + and - are chained together.

TommyT1 (author)2014-11-15

Perfect, if all you wanted to do was power a fan for free.... i love it

JonathanO1 (author)2014-09-19

The gel pad will also heat up over time, thus reducing the efficiency of the heat exchange.

blodefood (author)2014-08-10

I am no electrical expert, but I can see that the 'ible demonstrates a useful idea. If there is any saving however small, I could suggest that it has demonstrated potential and needs further study. Take into account with regard to the cooled gel pack whether it was cooled outside in cold weather (free cooling), whether there was a loss of heat in bringing the gel pack inside the house by opening the door, AND if the heat produced by the computer offset (however miniscule) the amount of heat desired from the home heating system to bring the interior temperature of the house to the desired level. You can offset your heating costs with incandescent bulbs because they produce heat as well as light, but while you pay less for gas, solar, geothermal or other heat, you will pay more for hydro in the process. You might as well try doing the same thing with computers or other appliances. Just saying.

ProfessorCrey (author)2014-06-27

Hi ssarthak598,

Its a very nice project, but as the fellow commentators have pointed out your efficiency calculation appears to be off. Aside from the argument based on known efficiencies for peltier effect devices(the thermoelectric effect is very inefficient) we can run the numbers easily too. A modern computer processor in today's laptops draw power at around 30W. So 36% efficiency would amount to recapturing 10.8W as electricity. You can show that even with a perfect heat engine that the maximum theoretical efficiency is Eff= 1-T_c/T_h ==> (T_h-T_c)/T_h. Given your temperature ranges above(remembering to use kelvin) that gives a theoretical max of 22%. You don't have to believe the theory blindly if you don't want to. Measure the amperage and calculate the electrical power using P= I*V and I'm sure you'll see a number much lower than 22% even.

Again, its a nice project but its important to have the correct perspective when analyzing the things we make.

lazyboy395 (author)2014-06-21

a little more work and you could probably even patent this

billbillt (author)2014-06-09

I think you did a great job no matter what all the downers say about it.. At least you got away from the lazy and just talk like most of the downers are and did something to investigate,,, I think this approach has great possibilities ... Double plus good again,,,

ssarthak598 (author)billbillt2014-06-10

I am currently researching on this technology deeply :)

ssarthak598 (author)billbillt2014-06-10


billbillt (author)2014-06-09

got my votes.....

Orngrimm (author)2014-06-04

Volts are worthless if there are no amperes behind. The voltage you measured are the idle voltages and not loaded voltages. Connect a reasonable resistance for a DC/DC-converter like 20Ohms to the peltiers and measure the voltage AND the current (at the same time with 2 multimeters). This way you get the wattage produced by a peltier of yours. And wattage is what counts.

Then if you have this, you can check the watt-hour of your battery and do a first calculation:

Lets say every peltier generates about 1W of power (which may be a bit high but anyway) and your Battery has 95Wh (My HP notebook has such a battery). Also say the total DC/DC-conversion gets made very well and you get 90% from the Peltiers to the 18V your notebook needs. Then again 80% efficiency from 18V powersupply of the Notebook over charging-controler into the battery... Leaves you at 72% efficency from peltier to battery.

Lets assume that your 12 Peltiers really deliver 1W each so 12W total. With the guessed efficiency of 72% from peltier to battery that would be 8.64W to the battery. 8.64W load-power to a 95Wh battery gives a charge-time of just shy of 11h.

Under the assumption that your Pelties really provide 1W each (I doubt that... More in the vincinity of 0.2-0.3W) then you could recuperate about the guessed 8-9W from the dissipated heat. If your Notebook uses about 19W like my HP you recuperate about 45% (for 1W-peltier-power each) to 11% (0.25W peltier-power each).

Both numbers are not too shaby to be honest.

But to know where we stand, we would need wattage and not idle voltage.

Also a suggestion: Be sure you have a REALLY goot heattransfer from the Notebook to the peltiers... Looks like you just placed the notebook atop the peltiers... Use heatconducting paste (Attention! Messy!!!) to increase the heatflow from the notebook to the peltiers... Also placing peltiers on a metal table would help to dissipate the heat very efficiently...

ssarthak598 (author)Orngrimm2014-06-05


Here I can give you the results after testing my project several times-

Main power supply-
100% in 4 Hours --1

Recycled Heat supply-

36% in 4 Hours --2

Adding both (1+2)-

136% So 100% charged in 2.7 Hours

blightcp (author)ssarthak5982014-06-05

Restating times does not help answer the simple questions we are asking. A simple multimeter in your pictures would tell you how many milliamps you are producing. The poster above is correct, it is very easy to get a high voltage but any current will pull down the voltage quickly.

Try testing one Peltiers alone and check that the data you receive aligns with your expectations.

ssarthak598 (author)blightcp2014-06-05


1 peltier gives the reading given in the picture. It gives about 500mA.

blightcp (author)2014-06-05

While you did reclaim energy, your savings of 30-40% are incorrect. Energy had to be spent to cool the gel block prior to testing. Granted, you did harvest heat energy from the laptop but not with any efficiency.

I have tried this myself with ice water and boiling water on either side. The rate at witch the thermal energy is consumed to create electricity is only 2-3% efficient when I last researched this.

The only time I have seen this style of energy generation used, is in Iceland where they have access to very hot geothermal water and the winter cold air outside. this system works to power scientific instruments at remote stations. They then switch to solar chargers once there is enough sun to power the station.

ssarthak598 (author)blightcp2014-06-05

I think your power coming may not be constant. My heat was constant. Here is the proof-

Main power supply-
100% in 4 Hours --1

Recycled Heat supply-

36% in 4 Hours --2

Adding both (1+2)-

136% So 100% charged in 2.7 Hours

zikzak1 (author)2014-06-01

I would almost think for this to be effective, you would want to either heatpipe heat off of the built in heatsink , or at least do a heat exchange from the hot air blown off of the heatsink. Just a cool gel mass and the heat off the bottom of the laptop wouldn't be near as effective, or one would think not... It would be interesting to see the amount of current and voltage generated. After thermal losses and regulation ineffiencies, I don't see how this could be much more than 10% efficient, but it looks like a cool project.

Ziggythewiz (author)zikzak12014-06-04

10% for this project would be great. I doubt he could get more than 1% with this setup.

Using the Fan vent, which was designed by actual engineers to remove as much heat from the laptop as possible is the only viable option.

If the bottom of the laptop was covered by plastic, it was never meant to conduct heat in the first place.

This is like capturing heat from a stove by placing a device on the floor rather than above a burner.

ssarthak598 (author)zikzak12014-06-01


You can try this project yourself. I took 12 peltier modules for my project. 1 module gives me about 0.6-1.0V in this much temprature difference.

YankInOz (author)2014-06-04

By the way, here is a photo of an impossibility: Using a TEC and liquid cooling to dissipate the heat to produce water but it got a bit cold. I found that a heat sink did not dissipate the heat fast enough. These results are supposed to not be obtainable. The results only took 3 minutes. The ambient temperature was 85F. Peltier devices can't do this. As you can see.

So, just keep on keepin' on.

ssarthak598 (author)YankInOz2014-06-04

I am slowly and slowly advancing it

YankInOz (author)2014-06-04

May I suggest that the reason you have extended battery life is your reduction of heat by the use of TEC being used for the Seebeck Effect. When you use a TEC to convert heat to electricity this is called the Seebeck Effect. The temperature variation between your computer and the gel is enough to convert some of the heat to electricity but unfortunately as configured would only generate mili-amps. You list your the model number of your TECs but what are the wattages? That is the critical factor. Most conversion using TECs is less than 3-6%. Your results seem to indicate 10 times that amount. Heat is one of the factors that can cause a battery to drain more rapidly.

What was your ambient temperature in the room/lab where your computer was being cooled and heat converted. It is not just the temperature of the computer and the gel but also the ambient temperature that will effect the power production.

There are ways to increase your power output and they do thumb their nose at the Carnot Cycle. My philosophy is that "scientists" who quote history without looking at the current possibilities are no longer scientists but philosophers.

It is just like the first photovoltaic cells were far less efficient than the ones today. Don't let nay sayers discourage you.

Some of my work in thermal conversion has stumped a for MIT professor whose field of study was thermal conversion for now almost 15 years. I had people quite Carnot Cycle to me and then when they witnessed my results, they were a bit less defending of old ideas.

Another anecdote: In the late 1800s man believed that if we travelled fasted than 60 miles an hour our heart would collapse and we would do. Seems true today, aye.

Just keep at it. It would be nice to sit and chat with you about your projects.

NightLord (author)2014-05-31

Could you please explain how did you calculate efficiency?
30 - 40% efficient conversion from 80°C to 20°C (room temperature) is unfortunatley thermodynamically impossible. (Carnot Cycle, google it).

ssarthak598 (author)NightLord2014-05-31

We can reusing the heat from this project. We are calculating how much energy are we saving. After i charged the laptop, it saved me 36% of my battery lifetime in a time of 3 hours.

The charging time of laptop is 3Hrs.So it is 30% - 40% efficient

NightLord (author)ssarthak5982014-06-01

How did you use power generated from peltiers? If you didn't use it to charge laptop, than there is no reason it would increase.

Don't get me wrong, still a nice project, and it could save some battery lifetime, but nowere near 30 - 40%. Maybe something like 3 -4%.

Unless i got your calculations wrong. Usually efficiency is defined as amount of electricity/amount of heat relesed.

ssarthak598 (author)NightLord2014-06-01


Its possible. Dont tell me this. I have practically tested.

When I tested, 1 module generated 3V on temperature difference of 55C - ICE(0C) and Candle(53C)

And the efficiency-

You can run the project yourself as well.

Main power supply

100% in 4 Hours --1

Recycled Heat supply-

36% in 4 Hours --2

Adding both (1+2)- 136%

So 100% charged in 2.7 Hours

antioch (author)ssarthak5982014-06-03

An amateurish energy-related 'ible by an author who doesn't even seem to grasp the basic concept of watts and voltage? Seriously?
And you run your mouth all over the comments when you don't like the tone provoked by your lack of knowledge and pomporous attitude?

When will you stop telling people not to tell you things you don't agree with?

You are hypersensitive to criticism and will thus fail and harm utterly wherever you go in live.

It's all over this 'ible, very repetitive and annoying. And do you even know what "spam" is? What DO you know?

rlee27 (author)antioch2014-06-03

I must concur that the premise is good but scientifically this is not a complete, or accurate experiment. Watts are the important measurement. Where is the energy going is the other concern. The numbers look good, but the math reminds me of a politician balancing a budget. David Copperfield math.

ssarthak598 (author)rlee272014-06-03

I will surely update them in 15-20 days

ssarthak598 (author)NightLord2014-06-01


You can run the project yourself as well. just take a laptop pin and connect it to project.

Main power supply

100% in 4 Hours --1

Recycled Heat supply-

36% in 4 Hours --2

Adding both (1+2)- 136%

So 100% charged in 2.7 Hours

ssarthak598 (author)NightLord2014-06-01


Its possible. Dont tell me this. I have practically tested.

When I tested, 1 module generated 3V on temperature difference of 55C - ICE(0C) and Candle(53C)

And the efficiency-

You can run the project yourself as well.

Main power supply

100% in 4 Hours --1

Recycled Heat supply-

36% in 4 Hours --2

Adding both (1+2)- 136%

So 100% charged in 2.7 Hours

andreyeurope (author)ssarthak5982014-06-01

NightLord is right! You can't use TEC Peltiers to produce this energy needed for laptop charging. I don't know how did you measure that, but my colleagues have tested a lot with this.

ssarthak598 (author)andreyeurope2014-06-01

I used a Volt and Amp meter for measuring the voltage.Tell your colleagues to consult me for their project. I will feel good helping them :)

Please feel free to message me or email

ssarthak598 (author)andreyeurope2014-06-01

First check this project by yourself and then comment on my project please. I dont like spamming comments please!

ssarthak598 (author)andreyeurope2014-06-01


I am using TEC modules generate enough electricity. In temprature difference of 50C, 1 module generated me 2.5V.

andreyeurope (author)ssarthak5982014-06-01

NightLord is right! You can't use TEC Peltiers to produce this energy needed for laptop charging. I don't know how did you measure that, but my colleagues have tested a lot with this.

billbillt (author)2014-06-03

double plus good.....

AMCGT (author)2014-06-03

Dear Sarthak:
Your investigation is very interesting, but the results need more work. You said you made a proof of 36% of recycled heat, but we only see a little fan blowing with energy that you got both from the computer bottom (not the most efficient place, it should be the heatsink) and a cooling gel, that you cooled elsewere spending energy. Most people will tend to be working in a confortable environment of about 20ºC not 7ºC of the gel.
Another thing. It's irrelevant the Volts that you obtain. Volts is not energy. It's just tension. You need to measure Watts, and so you had to show us the measured electrical current (W=V*I), for each temperature diferencial.
Anyway, keep working. You have the initiative. You just need to bare with critics and some failure, and eventually you'll arrive to results.

ssarthak598 (author)AMCGT2014-06-03


I will surely update my porject.


andreyeurope (author)2014-06-02

Hey Sharthak, I didn't want to offend you, and I'm sorry if I did offend you, but you're contradicting yourself. You told me in a comment that 1 Peltier module gives you about 2.5 V, and in another comment you said that it gives you 0.6-1.0 V. Could you be much clear? All you will create will be confusion.

ssarthak598 (author)andreyeurope2014-06-02

I am sorry for my bad behavior. I told that in my project normally, i get reading 0.78V-0.86V when it is running on 1 module. The peltier module gives 2.5V on difference of 50-55C when it is running on heat of candle.

I have put 12 modules together. That gave me around 9.8V-11.4V(Approx 12V).

Thanks for your patience

Warm Regards,

Sarthak Sethi

I_-FANGOR-_I (author)2014-06-01

Using a gel pack is a huge waste of energy because you have to initially cool it down. You should you use large dissipator instead.