Introduction: Making a Beefy Peltier Cooler!

Picture of Making a Beefy Peltier Cooler!

I wanted to make a more efficient cooling element for my electrically powered cooler so I went ahead and ordered the necessary parts from

Heatpipe heatsinks are really good at cooling CPUs (those 2 pictured can handle 160Watt TPD easily) so they are perfect for cooling 40mm Peltier chips. I used a 12volt 6amp rated Peltier for this project.

Step 1: Putting the Heatsinks and Peltier Together.

Picture of Putting the Heatsinks and Peltier Together.

Using thermal paste that came with the heatsinks, I sandwiched the Peltier between both and bolted everything together while making sure the Peltier was evenly placed between them.

Just a note, the printed side of the Peltier is actually the cold side once the polarity of the wiring is followed.

Step 2: Cutting an Opening on the Cooler Lid.

Picture of Cutting an Opening on the Cooler Lid.

Marking out the shape of the heatsink, I cut an opening in the cooler's lid. I used a sharp knife to clear the burrs off the edges then my vacuum cleaner to remove all the mess.

Step 3: Bolting the Heat Exchanger to the Lid.

Picture of Bolting the Heat Exchanger to the Lid.

Using 3 inch long stainless steel bolts, washers, nuts and wing nuts, I secured the entire exchanger onto the lid. Once fitted I installed the fan for the cold side heatsink to circulate the air.

Step 4: Putting the Electricals in Place.

Picture of Putting the Electricals in Place.

I soldered the 12volt and 0volt wiring respectively to give a low impedance and reliable solution. The temperature meter I stuck onto the lid with black silicone adhesive. The thermal sensor I stuck inside the cold heatsink fins.

For power, I used a 12volt plug to allow easy use in the car.

A quick test shows the internal temperature drop to below 9C with an ambient of 28C within half hour. The heatpipe heatsink is very good at cooling a 65watt Peltier! I also included thermal images of this preliminary test.

Step 5: Adding Thermal Insulation.

Picture of Adding Thermal Insulation.

Most folks don't know this but windshield sun screens, the silver type, are a very good heatshield and insulator. I cut a section of this from my stockpile and silicone stuck it onto the lid, both upper and underside.

Step 6: Fixing the Handle.

Picture of Fixing the Handle.

Due to the large size of the hot heatsink, I needed to replace the carry handle with a strap. I used an old bag strap to complete my Peltier cooler.

Step 7: Testing.

Picture of Testing.

Letting it run with nothing in the cooler, the ambient being 29C, in half hour it reached to below 17C. I'm quite happy with this cooling performance.

Peltier coolers work best with pre-chilled foods since they can't freeze food on their own in hot environments. Mind you though the cold heatsink is pretty large but I don't mind the loss of storage volume given how well it works.

An interesting note is that at 9 volts the cooler will get colder but take a bit longer to achieve this. This is due to the heating losses within the Peltier being directly proportional to the current used.

Step 8: Improvement!

Picture of Improvement!

The inside fan is a nasty heat source and I didn't want to degrade the cooling capacity so I just plain removed it. Now the temperature difference is so much more! Amazing!

I hope you found my instructable interesting and for a total cost of 700TTD, I have a cooler that is rugged and can keep groceries cold until I reach a refrigerator. No more need to buy stupid ice and having to clean the resultant mess.

Update July 2016:

Do to the ridiculous size of the cold heatsink, I reduced its size to about 30% with no loss in performance:


corradini (author)2017-10-01

VERY interesting point that higher voltage takes longer but gets colder. (Makes you think about what "efficiency" means, from both a thermodynamics and a practical standpoint. :-) So it seems it'd be smart to start out with a lower voltage, to get a quick drop in temperature (and save power), then up the voltage somewhat - if you want things colder. You could do it manually, or for the extra points (and expense) throw a 'duino or similar into the mix.

(I'd be curious as to whether it'd then stay at the lower temp if you reduced the voltage again. That'd obviously depend greatly on the real-world properties of the cooler and system, but not sure whether it's just ultimately the heat-loss-proportional-to-voltage factor anyhow.)

mtairymd (author)2016-03-20

Nice job - looks very professional. The windshield screen was a nice touch. As for the last step, people often ignore the heat of the fan in a sealed/insulated enclosure. With the fan, you are also increasing the heat transfer coefficient from the walls to the inner ambient which lowers the thermal resistance to the hot side. This lower resistance might account for some of the additional heat rise. It's fun to run experiments like this and look at the data - yep, I'm an engineer.

corradini (author)mtairymd2017-10-01

Now THAT (thermal transfer from walls) was a clever insight that nobody else picked up on. I had to think about it for just a sec, but yeah - the more air circulation is going on the more you're using the walls of the box as kind of (really poor, but still) 'heat sink fins'! The walls are inherently warmer than the contents of the box, so moving air around must increase the speed of external heat transfer into the box. You want the inside air nice and still.

Mjtrinihobby (author)mtairymd2016-03-20

Thank you greatly for your compliment! From the way your wrote your comment and the technical content I had a good idea of your line of work. You see, I'm an engineer too.

mtairymd (author)Mjtrinihobby2016-03-21

I figured you were as well :). I've only used them once at work which was an application where we needed to keep a component locked at a certain temperature while the ambient varied from from 0 to 50C - used as a heater and cooler.

I did think of an interesting application for a car but I haven't run the calculations to see if it's feasible or economical - most likely neither. What if I mounted Peltier coolers to the exhaust manifold (hot side) while the cold side was placed on a heat sink in the air stream under the car. I assume the temperature difference would be in excess of 250 C. The energy would be used to replace or reduce the load on the alternator. Ignoring the catalytic converter issues, what do you think?

Mjtrinihobby (author)mtairymd2016-03-21

Ah.. The Peltier is terrible at power generation. Plus it will fail rapidly at exhaust temperatures since the soldered joints would liquefy. I like what you did for your work project. I was worried cycling a side of a Peltier from cold to hot would cause it to experience thermal overstress. Maybe two separate heat exchangers, one for heating and one for cooling.. Great way to double the service life! I hope you keep your comments coming on my instructables and share ideas. It's great to communicate with a like mind!

mtairymd (author)Mjtrinihobby2016-03-22

Yeah, I know the COP is less than desirable.
It looks like the melting point of 60/40 Tin/lead solder is around 188C. I think the solder joints could be handled by
placing the peltier farther downstream of the engine. I'm just trying to think of a way to
capture some "free" energy. I figured it was a bad idea since auto
manufacturers are not using it

ccopyrites1986 (author)mtairymd2016-03-24

Curious, what solder is in question here? The heatsink solder I assume?

I could see if the tailpipe had a welded plate that the peltier is in contact with and the heat sink (on the cool side) being solid copper/aluminum or whatever is your favorite mounting onto the plate (sandwiching the peltier). There would be no solder required other than the wires.

Note: I'm assuming you won't need as much of a sophisticated heatsink because you have high wind speeds cooling it down and you could get away with a larger solid piece because it's hidden under your car.

Note 2: I had a similar idea but with a 'funneled' wind generator. a small fan (or series of fans) mounted into a tin can or pipe to shield it, when you drive it could potentially help power a smaller backup battery or something?

mtairymd (author)ccopyrites19862016-03-24

I think the OP is concerned with the electrical connections between the silicon and the electrical conducting plates. I think these are typically solder joints.

rmoody3 (author)mtairymd2016-03-24

Just because the car companies aren't doing it doesn't mean it's not a good idea.

quick_dry (author)mtairymd2016-03-24

you have to be careful where you remove heat from in a car - particularly the exhaust manifold as it is an area tuned for the exhaust pulses and flow e.g. to enhance scavenging.

If you take heat out of a turbocharged car in the exhaust manifold, you'd lose engine power - since the turbo relies on the energy of the gases coming out of the engine.

There are very good reasons why there are so many manufacturers of engine component heat coatings - some to insulate and keep temp out, many to keep the temperature "in" (i.e. not transferring from the gases to the pipes around them)

As for the fan idea, that is what aircraft do with a RAT - on a car you'd only want something like that if your alternator died, otherwise you'd be better pulling power for backup batteries from that. Don't forget, if you'd getting power from the free stream of air the car is moving through, the car will spend more energy to move a certain distance

Mjtrinihobby (author)quick_dry2016-03-24

good points!

For peltier on exhaust, I would fear that the quantity of heat generated saturate both sides of the peltier, then no more voltage will be generated.

but if you can keep the cold side cold enough, it could be interesting if it is 65 w generated on each module !!

Maybe. I think the "beefy" heat pipe heat shown here would do I nice job of keeping the cool side at a decent temperature.

Mjtrinihobby (author)mtairymd2016-03-22


DonCenzo (author)2016-03-24

Hello to the engineers out there! I'm thinking of building a "Man Cabin" That will be totally off the grid - Rainwater cistern, Solar powered LED battery charger, for lighting & a tablet, and to use repurposed PC case fans to circulate/recirculate fresh air, etc. Can you offer guidance on the sizing of a solar cell array to recharge a car battery that would maintain the cooling system described in this 'able? How scalable is the peltier cooling/heating concept? Can it be used to cool down a full sized picnic cooler? My concern is that the cooling element takes up a lot of the interior volume of the chest. Could it be affixed externally and cool interior air recirculated through ports?

I have had several commercially made peltier coolers over the years. This looks like a great project for me and my grandson to upgrade the hunting/camping cabin in the backwoods on my property. Thanks!

Very cool!

Pun intended!

"Not an Engineer"

Mjtrinihobby (author)DonCenzo2016-03-24

Honestly, once you can afford it, buy the true compressor type, refrigerant based portable freezers from amazon. They cost 380USD and upwards depending on the size and weight 20lbs for the smallest unit. These are the most efficient and run directly off 12VDC and are in every sense of the word, a perfect little refrigerator.

For solar powering a home you can read my instructables:

corradini (author)Mjtrinihobby2017-10-01

>> the true compressor type...

That sounds like the voice of experience, and probably valuable to the guy who may need to seriously rely on one. OTOH, this *IS* Instructables... >;-)

HoppyH1 (author)DonCenzo2016-03-24

While I am not an engineer by degree, I do have over 15 years of college covering a very wide range of subjects, including that discipline. I also live off grid (for the last 5 years) and use 9 ancient solar panels of the amorphous Silicon pattern charging 6 car batteries, but I do not use this system exclusively as I have learned to my regret that no single system is reliable enough when your survival depends on it---I nearly died my first winter. As such I have backup systems for everything I need when it comes to heat, light and power, sometimes as many as 7 different systems of backup.

That said, you can build an affixed external cooling chamber for your ice chest, as many of the really old propane refrigerators are basically that way. You will however, have to insulate your cooling chamber and affix it soundly to your cooler, which will make it far from portable and of little practical value (other than you get the joy of making it yourself) over purchasing an RV fridge that uses propane and electricity for cooling like I have.

As a final note, I strongly suggest that you check your State's laws on rain catchment systems as some States, like Colorado makes that practice a felony

benwade (author)HoppyH12016-09-18

I just thought I'd respond to your comment about rainwater collection being a felony in some states. Just to clarify, this was never true, it is an internet meme based on one man who was arrested and jailed for 30 days after having been denied a permit to build a rainwater collection system. He requested a permit, was denied, and went ahead and built three (3) giant rainwater collection systems. He was arrested, not for collecting rainwater, but for doing construction without a permit. Of course, this was "spun" as evil governments trampling on the "little guy."

P.S. The guy was quite well-to-do if not rich.

corradini (author)benwade2017-10-01

Not quite accurate. I don't know about felony vs. misdemeanor vs. civil violation, but it IS true that collecting rainwater was generally illegal in Colorado. (Link to article in Denver Post - which I presume is not fake news - below). This has been modified, VERY recently, but historically Colorado - in its state constitution! - held that water in the air was owned *collectively* by the people of the state and that if an individual collected rainwater that it infringed on any senior rights of people "downstream". Water rights in the American West is a VERY complicated, touchy, and much-litigated issue.

John CarloT3 (author)2017-06-24

Nice build but i don't think the cold side fan is causing significant heat.

It's just that you should disspate more heat on the hot side. There's a reason why most peltier cooler setups have smaller heat sinks on cold side.

If we want a cooler/chiller/freezer we want to dissipate the warm heat, so we put a larger heat sink at the hot side. We do the other way if we want a heater.

I didn't try looking it my self but you can check the size of an aircon's Evaporative Fins (cold side) compare to the Condenser Fins (hot side). See which one is smaller in size.

TutuA (author)2016-04-18

I made one peltier cooler box similar to yours yesterday with a Polystyrene box. The cooler box consists of 2 pieces of peltier module that consume 120W in total.

But I think the key of getting lower temperature is insulation.

Since you opened a big hole on cooler lid, I doubt that hot air would back from that hole.

Mjtrinihobby (author)TutuA2016-04-18

correct! insulation is the key!

steven4872 (author)2016-03-22

Heat pipes come in two version a diode version and a non diode version. The diode version is the cheapest to make. It conducts heat in only one direction, up. The heat source is on the bottom where it boild a fluid in the heat pipe. The resulting vapor moves up the pipe, condenses at the top and then the condensation moves down to the heat source. Diode heat pipes don't work well when the heat source i at the top.

Non diode heat pipes do exist but they typically are more difficult to make and cost. more. i don't know anything about the heat pipes you purchased but my gues is that the bottom or cold side may not be working any better than an empty copper pipe. You hot heat sink should work fine. You may want to look cloely at the information that came with your heat sinks.

Mjtrinihobby (author)steven48722016-04-14

It is already built and working perfectly to my requirements. No further changes will be made.

Kinnishian (author)steven48722016-03-24

Non diode pumps can be found in objects that are expected to move- for example, laptops. But they can be much smaller for laptops. I'd look towards old large gaming laptops specifically. If you have two samples, you can cut one open and loop for a wick in it to determine if it's the wicking kind that works in most configurations.

I am not sure the use of the heat pump makes sense for the cold side. Depends on whether the fluid boils at the cold temperature. If it doesn't, you've lost all the benefits. I feel like OP should address that

david.roseman (author)2016-04-01

So....where do you put the food when the entire inside has a heatsink!!!?

starphire (author)2016-03-27

The commenters noting that heat pipe orientation can be crucial to its performance depending on its internal design are quite correct. But there's an even bigger issue with using a heat pipe on the cold side of your Peltier device, as a heat pipe will not actually function as such outside of its intended temperature range - that is, the working fluid that changes phase to a gas and then recondenses into a liquid *must* get hot enough to vaporize, and the working fluid and internal pressure is normally chosen for a hot side that will be well above room temperature and a cold end that is near room temperature. You would need a specially customized heat pipe to operate correctly in an unconventional temperature range such as this.

I have no doubt that you are seeing a good transfer of thermal energy from the cold side to your cooler, and if you're happy with it now that's great. But I believe that if you were to cut or drill a hole in each heat pipe there - thus disabling its vapor/condensation cycle completely - you would see no change in performance at all. Because you would still have copper tubes connecting your cold plate to a large fin surface area, so it will still function as a heat sink. However, I don't recommend that you actually do this unless you know what's inside, as the working fluid may be hazardous if released.

LeonardoL47 (author)2016-03-27

I think you need increse the heat disipation for better termal jump. If you can down the temp of the hot side of peltier cell in 5° or 10°, your cooled side probably down in 5° lower temp.

LeonardoL47 (author)2016-03-27

Tell me if i wrong, that Heatpipe optimal function was desinged with the base down and the disipator up, because the gas in heat pipe were liquid in the base and gaseous in the upside. If you put inverted, the cooled heatpipe probabliy did not was eficiently. Maybe you better performance with an solid Alu disipator and a coolerfan.

Sorry for my bad english

AeSix (author)2016-03-24

So, a few concerns and improvements - otherwise extremely nice tutorial.
First, way too much thermal compound. There should only ever be enough to lightly lubricate the two surfaces and fill in gaps/divets and to smooth over the surfaces. The idea is that the machined surfaces of the heatsink and (usually CPU, but heatsink in this case) are full of microscopic fractures, bumps etc, and that the two surfaces will not mate up 100%, possibly with slight differences in angles causing a nearly insignificant wedge of air between the two surfaces. The thermal compound is to be used to fill in those gaps, as minimally as possible without leaving more gaps. Using so much that it squishes out of the sides of the surface areas is too much. Thermal compound, though intended and designed to conduct as much heat as possible - it is still an added layer of material which still provides an amount of insulation to the system. That insulative factor could potentially become higher than than the conductive factor of the thermal paste, or at least higher than using no thermal compound at all. (it can mildly retard the ratio, or make the efforts of the compound completely useless and more damning than not using any at all). Sorry for the long winded spiel, I've seen so many PC builds and other projects thwarted for efficiency by this simple mistake, I feel the need to explain when ever I see this happening.

The next thing I would suggest is using a 90* bent heatpipe for the inside half of the system, this would allow for the same thermal transitional area without stabbing into the middle of the cooler taking up space for food items. In fact, having now that I think about it, having two that are bent at 90 degrees may save a considerable amount of space. The two heatsinks could be mounted together sandwiching the lid between them (minus the space for the two heatsinks to touch directly) If done right, the hollowness of the lid could be used as airflow delivery as well. A few holes drilled in the lid with the fan protected by the heatsink on the inside, and the heatsink protected by the fan on the outside (which the fan is then protected by a metal fan grate). Again, just an idea for improvements. You can do an image search for "Low profile heatpipe" to get an idea of the type of heatsinks I'm referring to.

Another thing I thought of is, if you don't want to use the hollow of the lid for air dispersal and circulation, it could be beneficial to use inflatafoam around the heatsinks and bolt holes to re-add insulative value to the areas with holes cut in the lid. WARNING: DO NOT allow the foam to seem into the fins of the heatsink or it'll ruin the thing. Painters tape is enough to prevent the foam from going between the fins, but it may be difficult to peel off later. Another method may be to cut the hole, fill with inflatafoam and then once that's set and cured, cut out the space needed for the heatsinks and bolts. A butter knife or even plastic knife is all that is needed to cut the set foam, and a chopstick will push right through it for the bolt holes.

I hope I've been more constructive than critical for you! :D

cparsons7 (author)AeSix2016-03-24

So only going to comment on the thermal paste part and I will say I too was of your mindset with this matter, turns out we're wrong, based on a cpu not a peltier, there was a video done on this very subject on Linus Tech Tips channel not too long ago and basically the only issue with too much is a risk of shorting something makes no difference to the thermal transfer (on the hot side at least) which when you think about makes sense as this tried and tested method (I would think) comes from back in the 90's when it probably did make a difference

mtairymd (author)cparsons72016-03-25

AeSix's logic is correct. It is a simple calculation when you just consider the grease by itself "L/(k*A)" - not the full story. Thermal paste thermal conductivity (k) runs from 0.5 to 6 W/m-K which is obviously much higher than air at 0.025 W/m-K. The interface area (A) is fixed based on the construction. Increasing the thickness (L), increases the thermal resistance. When you watch the Linus video for the "too thick application", notice that the grease squirts from the sides due to pressure applied to the joint. Therefore, the final grease thickness isn't too bad or "too thick". If I apply a .125" thick layer grease with low pressure, I've created an insulator. This has happened on large surface area applications. So yes, too much grease can technically cause a problem. Linus' conclusion for his application was correct. BTW, there are multiple factors to consider such as flatness, surface finish and particle size of the filler in the thermal compound but I think we've all discussed this way too much :).

cparsons7 (author)mtairymd2016-03-25

I did say it was based on a cpu as was AeSix's and yeah I'm not arguing that if you have a chunk of paste in there it will not change into an insulator, but when you press the two together as with linus it squirts out and thus makes all those calculations useless as it's as close as it can get and any uneven parts of the peltier would need to have extra otherwise they would not get any contact, that's all my point was

mtairymd (author)AeSix2016-03-24

Wow, that was a long critique on thermal compound! However, you are totally right.

jimmie.c.boswell (author)AeSix2016-03-24

actually that is not the way, you want to apply heat sink compound. you want to apply it in the center as a mound. so that when you press on the item, that it spreads outward and does not trap air bubbles.

Mjtrinihobby (author)AeSix2016-03-24

very good points and I thought the same things for the lessons I learnt after building it. Thanks for the tips! If I make another electric cooler then I will build it differently and be more economical with internal space. I really want to use vapor chamber (cold side) coupled with a water cooler (hot side).

ThomasK19 (author)2016-03-22

Just at a guess: wouldn't it be more efficient to use smaller sinks but with a small ventilation? Of course, this way it looks much more impressing xD

Mjtrinihobby (author)ThomasK192016-03-23


FlorinJ (author)Mjtrinihobby2016-03-24

I guess ThomasK19 meant a fan on the outside. My guess is that this would help too, more than a larger heatsink.

ThomasK19 (author)FlorinJ2016-03-24

Actually I thought of fans on both sides. I'm pretty sure they would increase efficiency more than increasing the sink sizes would do.

FlorinJ (author)ThomasK192016-03-24

A fan on the inside would generate heat. Plus, convection currents would do the fan's job anyway, on the inside.

ThomasK19 (author)FlorinJ2016-03-24

I haven't thought about convection. However, a small fan does not produce much heat if operated at low speed. It's probably not that easy to create an efficient air flow inside without a fan. Note that one mini-fridge I bought a longer time ago DOES have a fan inside. Most likely just for that reason.

DDW_OR (author)2016-03-24

when i lived in Phoenix AZ i had an Igloo electric cooler. i added a fan on the outside to blow onto the heatsink. then in the morning i put the cooler , with the lid open, into my pickup. the cooler was not plugged in. then at noon i closed the lid and plugged it into 12v outlet. the pickup cab and the inside of the cooler where both at 115 deg.

after a 3 hour drive the cab was still 115 but the cooler was at 40. i do not know when the cooler reached 40 during the drive.

Mjtrinihobby (author)DDW_OR2016-03-24

for my cooler in the car, at over 50 Celsius it takes 20 minutes for the internal temperature to drop to 20 Celsius.

TMorgan69 (author)2016-03-24

Do you have a part number for the heat sinks?

Mjtrinihobby (author)TMorgan692016-03-24

ARCTIC Freezer i30 CO Extreme CPU Cooler

FlorinJ (author)2016-03-24

Given that cold air flows downwards, I'd expect you to get similarly results with just a solid flat heat sink, reaching out sideways. The convection currents inside the box should do all the work.

About This Instructable




Bio: As of April 2017 I have decided to no longer post on instructables. The fact that several of my published works have been removed without ... More »
More by Mjtrinihobby:External Light Sensor Night Light.Replacing a Toilet Tank Float Valve. How to Make Pig Feet Souse!
Add instructable to: