Thermoelectric coolers (TEC or Peltier) create a temperature differential on each side. One side gets hot and the other side gets cool. Therefore, they can be used to either warm something up or cool something down, depending on which side you use. You can also take advantage of a temperature differential to generate electricity. The thermal tape listed below works very well to attach heat sinks to the hot side.
This Peltier works very well as long as you remove the heat from the hot side. After turning on the device, the hot side will heat quickly, the cold side will cool quickly. If you do not remove the heat from the hot side (with a heat sink or other device), the Peltier will quickly reach stasis and do nothing. We recommend using an old computer CPU heatsink or other block of metal to pull heat from the hot side. We were able to use a computer power supply and CPU heatsink to make the cold side so uncomfortable we could not hold our finger to it.
Note: It is imperative that a heat sink is used on the hot side of the module. Running the module without one can cause damage to this part. If it’s too hot to comfortably touch, you’re in the danger zone!
In order to get the best performance from a Peltier module, you need to be able to remove as much heat as possible from the hot side of the device. Several customers report that a PC water cooling setup is the best way to get dramatic heating and cooling results. If you're not using a water cooling setup, you need a substantial heatsink and fan to keep the module from reaching a warmish, useless equilibrium. No matter what you use, it needs to be very firmly attached with thermal grease/epoxy in the middle. Running the device without a heatsink on the hot side can permanently damage the module.
Reversing the polarity will switch which side is hot and which is cold, so you could, for example, heat & cool the inside of a box with one Peltier. A simple DPDT switch could be use for this, or if you prefer to do it electronically, a H-Bridge would work.
If it requires power, you need to know how much, what all the pins do, and how to hook it up. You may need to reference datasheets, schematics, and know the ins and outs of electronics.
Skill Level: Rookie - You may be required to know a bit more about the component, such as orientation, or how to hook it up, in addition to power requirements. You will need to understand polarized components.
See all skill levels
Based on 4 ratings:
1 of 1 found this helpful:
I made a Seebeck Generator and it was awesome. See it here: https://youtu.be/wqgAroBRL0Q
...but it came with no brand name, no manual, no specs, no instructions, nothing... And also found after ordering these there are cheaper yet equal or better quality tecs elsewhere. But they are well made and work great. Finally blew one after heating to over 430 degrees Fahrenheit fyi
Hi, To save on wasted materials, we generally do not ship printed material with our product. We do still share as much pertinent information as possible on the product pages.
I have to agree with the some of the previous reviewers comments on this device, those being that it is well made and works great.
The way I tested for the cold and hot sides was by connecting a 1.5 volt AA battery, positive to red and negative to black and holding the wires in place for 5~10 seconds. Even at that low voltage and current I was able to easily figure out what side was what.
After, I mounted an old CPU cooler with fan to the hot side using the "Theragrip Thermal Tape" sold here and all is well with it cooling my laser diodes since I bought it a few weeks ago.
This reasonably priced electronic marvel has proven to be not only a captivating instructional device for my middle school students (my 6th graders think that I am a magician when I make a propeller spin using hot and cold water to create a temperature differential to power a motor) but also as a necessary heating or cooling component for some of my 12th grade Capstone project students. It also serves as an additional instructional example of the "two way street" nature of electrical components such as motor/generator and electromagnetic induction.
Looking for answers to technical questions?
We welcome your comments and suggestions below. However, if you are looking for solutions to technical questions please see our Technical Assistance page.
Log in or register to post comments.
So is the max operating temperature listed here (180C) the maximum temp the hot side of the pad should get too? I am confused because in the description it says the if its too hot to touch then "you're in the danger zone". I guess what I want to know is the maximum temperature the hot side can achieve before it damages the part.
Will this only draw a maximum of 7A? What would happen if I hooked this up to a 12v/15A power supply? Would it break or just max out at 7A?
This thermoelectric cooler will only draw as much current as it needs, up to a maximum of 7 amps. You could connect it to a 12 volt, 1000 amp power supply and it will never draw more than 7 amps.
Sorry its a bit late, but thank you.
What's the appropriate way to power this peltier module?
I have two different AC adapter with same spec (12V/3A). It seems like one has short circuit protection, but the other doesn't. With the one with protection, I couldn't power the module by connecting it with the power supply directly maybe because it created a short circuit and the power supply did not allow current flow. With the one without the protection, I was able to power the peltier module; however, after powering it for a few hours the power supply suddenly exploded (fortunately no hurt or damage to house). So I should have powered the peltier cooler in a different (and safer) way. So what's the best way?
How fast do these return to room temperature?
I am an amateur at best so sorry if this comment comes across as "less informed" (stupid) :P
I am working on making a small heater/cooler for a Phd student within textile design and I need this small space to be cooled down to about -10°C I was wondering if this peltier along with a large heatsink and fan could do the job? If not, can one connect multiple units to create a greater effect?
Aren't these also supposed to provide power in the presence of a temperature differential?
If so, I'm curious how much you could generate from, say, an armband of these while running on a cool day (trickle charger?). Maybe a cup of hot beverage (umm, useless "HOT" led indicator). This could get interesting.
At the very least, cooling without mechanical action is intriguing.
I asked reukpower (author of
http://www.instructables.com/id/How-to-build-a-thermoelectric-lamp/ ) about how much power he could generate using his project, his answer was that his candle powered peltier thermalgenerator was capable of 0.8W but you could get much more out of these if you could find better cooling...
I looked into this myself not too long ago, while possible to get current out, these are designed to pump heat. If you want to generate power, you'll need one optimized for that purpose.
I have a 100 watt peltier in my junk box, i hooked an LED to it and heated one side with a hot plate and iced the other, couldn't get the LED to light for the life of me.
Did you try reversing the LED terminals?
I had a 35 watt module rescued from an old 6 pack soda / beer cooler, and used a candle to drive a 5 volt fan. If you heat sink / ice one side and heat the other, you can get enough power to do small jobs. If you did not get the LED to light, you may have had issues with how the led was connected or, the module was bad. I can definitely verify that you can do something with this. I used mine to have scented candles where the scent was pushed around by the fan, no batteries required.
This would be a useful companion part to the energy harvester breakout board. The chip's spec sheet even shows a Seebeck-effect generator driving it, and that's one of the things these TE coolers are!
Yes, the principle is also true in reverse. If you apply a temperature differential, you will get a voltage.
Expecting any more of these? They just seem really cool(pun intended).
I think they're pretty hot personally. But I guess it depends on how you look at it.
In Germany they have used these on the cars exhaust to generate power to help supplement the alternators current.
with the temp difference of the exhaust pipe and the ambient air gives a great temp difference to generate power.
I remember seeing a BMW demo where they recovered so many watts of thermal energy from the exhaust that the tailpipe was cool to the touch and the water vapor in the exhaust had almost fully condensed. We'll definitely be seeing more thermal insulation and recovery optimizations on cars in the future because it's 'low-hanging fruit' for the engineers.
What is the smallest battery this can run from and cool noticeably?
I used If i remember 5AA Ni-MH batteries and this thing became very cool. I did not want to touch it. Hot end became very very hot. Putting some mafths into this Ni-MH = 1.2V nominal. The batteries had a rated capacity of 2400mAh. I had 2 wells of them, and it ran for an hour with 2 5V fans attached to it. That is 1.2V x (2.4A x 2) = 6.9W of power put into it. It maintained a small Styrofoam's internal air temperature, and slowed the rate at which the ice inside was melting. Later, I hooked this up to 10 Ni-MH batteries and it got cold and hot so fast it was crazy. Make sure to heat sink/fan the hot side if you are using this to cool. Its power is based off of its potential to create temperature gradients.
Experience and Summary: This thing runs at 5V. It probably can Run at lower voltages. It will draw as much current as you let it. The more voltage you let it have, The bigger temperature difference it can achieve. The more amperes you feed it, the faster it can achieve that difference.
Hi, How to put several of these in a series with arduino ?
I got one of these things down below -20 F by connecting the hot side to a water-cooled CPU heatsink: http://www.newegg.com/Product/Product.aspx?Item=N82E16835181015
I accidentally "burnt" my finger on the cold side... it still sort of hurts, actually...
This was with a +12V input from a computer power supply.
By comparison, using the CPU fan recommended by Sparkfun, I could only get the cold side down to +38 F - that's a 60 degree difference! Good cooling is essential.
Hmm... since one of these can freeze water solid, how about sticking one on the bottom of a mixing bowl and making ice cream?
That would be because most heat sinks have the maximum difference between the hot and cold sides... which.. listed above is 69C so you simply got your hot side 60 degrees cooler than sparkfun did, so your cold side got 60 degrees colder
Well they be in stock again??
I'm just getting started with electronics, but this is on my list of things to try.
I did find some good reading here: http://www.tellurex.com/technology/ (note the pdf versions on the right as well.)
Highlights for me were:
A heatsink is required on the hot side. And cold side should be connected to something.
You can use PWM (>= 60hz) or voltage to regulate the temperature(s) of the device.
The hot and cold side is arbitrary, if you reverse polarity it switches the hot and cold side. So it can both heat and cool which is good if your ambient fluctuates like the outside world does.
Not sure if the Seebeck or Peltier effects confuse me more than Ohm's law ;-)
Time to setup my first Sparkfun order...
Why doesn't SparkFun sell a heat sink to go with this?
We're working on it.
I have been using two of these for a project that I am working on. The black wire ripped out of the plate and now it isn't working. If anyone has any tips on how to fix this, it would be greatly appreciated. I need it as soon as possible please. THANK YOU in advance
Why haven't we attached one of these to a Bitcoin miner yet?
We could harness the waste heat to power the miner itself AND keep it cool!
Although I think Bernoulli, Joule, et al might have something to say about that idea. ;-)
Still, I always thought using a computer's waste heat to generate some electricity (while also cooling the computer) seemed like a better idea than just using a heat sink and fans to heat the room, even if the efficiency is limited...
They did this in Germany with a supercomputer. Although they used water as the cooling medium and then they used the heated water to provide power to other facilities that the super computer was around. brought up both the efficiency of the machine and the computing power.
The Wiki in the link above is outdated. I bought this, and took me a while to figure out how to use it. At the college today, found an excellent source that breaks it down in a simple yet meaningful way that nearly anyone could understand.
I dont see a datasheet for this particular thermocooler on this websites page... This is one of the products where the datasheet really helps to demonstrate application...
I am trying to make a small low powered cooling device that can cool down and heat up but dont want the device to get too hot or cold, could i power this with a simple 1.5V AAA battery? and what would that do? Thanks
These Little guys are so fun to play with! They do suck tremendous amounts of power though for their size... found that out the hard way... any chance you'll be selling a heatsink/fan combo again to go with this any time soon? Or maybe a kit.... That would be interesting.
If I need to cold down water from 85 to 5 degrees Celsius in less than a minute, how many Peltier devices would a need to use? is it any efficient?
Can you tell us more about the sealing on the sides of the device? It looks to be filled in with some white substance. If you don't have specifics on the material or ingress protection standard, perhaps impressions as to what it seems to act like? I'm trying to get a rough guesstimate as to its moisture-proof-ness.
From cross product investigation, generally the side seals are made from silicon.
Silicone. Like the caulk used to keep moisture out.
Cant ever remember there is a difference.. Gets me every time... sorry about the ambiguity folks.
Haha thanks - it's an easy one to misspeak/mistype
ya i have one of this on my 3d printer cooling my hot end works great
Does anyone know how cold this can get with one of the sparkfun heatsinks? (Maybe the Heatsink and Fan 70mm PRT-10686)? Someone mentioned that -20C was achieved with water cooling but was wondering what it can do with something air cooled.
This is a neat device, but I have to ask, what is the reason for the $15 price? I purchased one here and then, while looking for heatsinks for it on Amazon, I found it priced for only $5.00! That is a huge difference!
Any one know how this would hold up in a very moist environment. say for example the inside of a pot lid?
could these be placed along the insides of a 5 gallon beverage cooler? How many would I need to chill the water 10 degrees?
How long does this last. I talked to spark fun and they don't have any info on it. Does anybody know how long this can run before falling with the proper heat sinkage.
These devices generally have a MT-BF (mean time between failure) of about 10000 hrs from what I've read.
As to it falling, if you put it on a flat stationary surface the device shouldn't fall over. I suggest glue if you have further gravitational concerns.
These are solid state devices. In general if it's solid state, as long as you keep within the stats in the datasheet, it should last much longer than any moving parts of your creation.
Is there anyone here that has experience using this device for both heating and cooling to maintain a steady temperature. My son and I are attempting to design an incubator/cooler to make a specific type of Swedish cultured milk. The bacteria needs to be at a steady 24˚C (+- 1˚C). We live in San Diego, where air conditioning is rare, and the ambient household temperatures often surpass 24˚C – hence the cooling. We are working on using a Beagle Board Black as the controller. Any pointers and/or suggestions.
Is it at all possible to regulate/control the temperature of the heat side on this thing?
Yes, by regulating the power going into the device, by either changing the voltage, or turning it on and off (PWM).
That'll do. Weird follow-up question: can I break the ceramic plates so as to make it have this shape ^ without turning it into junk?
From what I know about the construction of these (it's a sandwich of numerous semiconductor junctions between conductors on the top and bottom, but note that I'm not an expert), you may be able to cut these and still have them function, but it wouldn't be at all easy or guaranteed. You might want to look at multiple smaller TECs instead of trying to cut a larger one down to size.
Ok, so I bought one of these and it got really hot and I was just using a fan to cool it down. It probably wasn't enough. What side gets hot and which side gets cold?
It depends the polarity.
wow the benefits of thermal recovery from tech like this could make many hot systems with ready cooling also produce suplimental electricity for them self.
HOW DO WE GET THAT MOSFET POWER TO BE HIGHER FROM THE EXAMPLE CODE GIVEN ABOVE?? PLEASE HELP
you are not allowed to write in ALL_CAPS unless you are in serious danger. are you alright?
Hey Does anybody know the exact data sheet for this thermal generator?
You can power one of these with an old ATX PSU's 5 volt rail. Just use the metal case of the PSU for the heat sink. I'm in the process of making a video that demonstrates this, though my TEC came from a minifridge.
I think you can make a simpler TEC generator with wires of dissimilar metals - nichrome & copper for instance. Twist segments together at both ends and wire in series. One half of each pair is exposed to heat, the other to cool air. I haven't done this, but practical generators for small heat generators like candles and oil lamps have been used to power small appliances. In the old Soviet days, I saw photos of folks (in Siberia, where the cold half was easy to find) gathered around oil/kerosene lamps that powered radios. Also, the Plutonium-powered electric sources in the Curiosity Mars lander run off the heat generated by simple radioactive decay.
Im working on a SNES portable and I want to use this keep the heatsink cold enough to touch but, I'm confused about which side should touch the heatsink and if I should put a fan on the side that is pushing out heat or will this be enough on it's own?
I am confused by the Tmax. What does it mean exactly? Maximum temperature difference that we can achieve using this TEC? Maximum safe temperature difference it can be exposed to, either powered or unpowered?
Instead of having a huge array of thermocouples, couldn't there just be one fractal thermocouple? Like fractal antennas in cell phones?
Connected a heatsink and 12V PC Power supply(20A), It only consumes 3 Amp. that's only 38 watt. Any idea how to get it use more, like 7 amp or 60 watts ?
YEA I HAVE THE SAME PROBLEM!!
I'm guessing these are only uni-directional, right? You can't reverse the voltage and get the heat to flow in the other direction, can you?
Quote from the research links: "Dubbed the “peltier effect”, it was later clarified by Emil Lenz who observed that by passing a current through a bismuth-antimony junction, water could be frozen and that ice could be melted if the current were to be reversed."
To answer your question, yes, the peltier plate will heat side A and cool side B and when you reverse the voltage will cool side A and heat side B. Note: To get ice, you need a heatsink on the hot side.
Oh, and if you put a multi-meter across the TEC the resistance is more of the order of 2.5ohm, which ties in with the stated values...ie: 15.4V/7A= 2.2ohm. So I'm not sure where the stated 1.7ohm comes from?? But it's questionable...
Please, oh please try and find a data sheet for these TECs. It is very important to understand the drop off in efficiency at lower voltages, ie: 12V. The efficiency at the quoted max values is 59.6% of input power being 'converted' to cooling. If you repeat the calc with the same efficiency at 12V, it shows that you should get 25.8W of cooling. HOWEVER, this assumes the same maximum efficiency. In actual fact the efficiency could be much lower, giving more like 10-15W of cooling. You can see the large margin for error when trying to design a system using these TECs. They are very good units, but we really do need a data sheet. So puuhhhrettyy please, ask your suppliers, do some digging, and try and find us some efficiency curves, etc :-) Thanks.
So, I'm thinking of buying one of those for building a small beer fridge, that could keep 12 cans cold (~4C). I calculated it would take ~285kJ to be removed, and if this unit transfers heat at 62.2W, it would take ~1h16min. Accounting for the aluminium can's mass, and not-perfect isolation, I can double this number, and call it 2.5h. Would this actually work, or am I missing something?
P.S. I assume beer is about as dense as water, and heat capacity is the same, and room temp is 20C.
I know this is an old post, but for you hackers doing an alcohol related project, Drinking Alcohol (ethyl alcohol) actually has a specific heat of 2.46 kJ/(kg * K) , where water (H20) has a specific heat of about 4.18 kJ/(kg * K). What this means is that the higher the alcohol content, the less energy it takes to heat (or cool) said volume of liquid.
Also, According to Engineeringtoolbox.com The density of ethyl alcohol is 785.1 Kg/m3, where as water's density is generally defined as ~1000 kg/m3.
So to find the actual specific heat of the beverage, with known alcohol content, use this formula.
C = (2.46 * (%Alc)) + (4.18 * (1 - %Alc)) Where C = specific heat of liquid beverage %Alc = Percent of alcohol per volume in beverage (ex: If your beverage is 5% alchohol by volume, you would use the number .05 in your calculation)
I am not going to poke at the mass change atm.
I'll Post again at the end of the semester (End of May 2015) with the rest of the calculations necessary to accurately calculate the cooling with this specific cooler once I finish thermodynamics. Good luck! -JT-
Ben Heck from the Ben Heck Show did a cool project with two of these. An semi-instant can cooler. Check it out. http://www.element-14.com/community/community/experts/benheck
It isn't quite true that nothing will happen without a heat sink. I've seen the leads melt off peltiers before. It was an accident...
So yes, use a heat sink. I don't know exactly where the semiconductor alloy melts- pure Bismuth Telluride melts at over 500 C, but this has some other stuff like antimony in it, and I don't know how much. In any case, if you want to generate power from one, you probably need to know when the leads are going to melt off, and when the Bismuth Telluride melts. And you still need to dump the heat into a heatsink on the cold side. And these are potted with something, but if it is silicone, it will take some heat.
The junctions in TECs are of Fe strips soldered on the opposite ends of the semiconductor cubes and are held in place by the ceramic plates.
The Thermal expansion becomes a major factor when you try to maintain large temperature differences. The Junctions of cold side shrink and the junctions of hot side expand. The unit will suffer a mechanical failure way before T(hot) reaches the melting point of the semiconductor.
Since it's a series of diodes by definition (mostly 127), a single failed junction renders the TEC unit useless.
I connected my TEC to a watercooling and got the temperature down to -36,9 *C (-34,42 fahrenheit), the element of the watercooling was about 19 Celsius.
I have a 140 Watt (at full load) CPU that I would like to cool with this. Since this device disipate about 100W, is it possible to stck 2 of the to get above the required 140 watts of dissipation? This would ofcourse be heat-sinked with a typical CPU sink and fan which would also have to be rated at around 200 watts of dissipation. My goal is to get a CPU running at below ambient temperature. (but not too much as I don't want condensation forming.) Does this idea seem possible?
Can I really use it to chill something, like a cup of soda?
Like this one http://www.thinkgeek.com/computing/accessories/96b3/ ?
Yep. You would need to figure out how to use it with 5v, but that could be a use for it.
One of the cooler applications for using temperature delta to power something the Ecofan: http://www.caframo.com/hearth/hearth_products_woodstove_ecofanairplus802.php
I'm tempted to make something like this over the winter but pipe to an external radiator.
I would have called it a Heat Pump as it isn't just for cooling, can be used for heating (as you state in the description) :D
They are commonly referred to as TECs or peltiers. I'm just sticking with that convention.
Neat! If you use one of these, make sure you put a heat sink of some sort on the hot side... Peltier devices are pretty inefficient and most of the 60 watts this dissipates is waste heat. Water cooling or putting a fan on the heat sink is even better.
The power generation aspect is also worth investigating. I believe that most if not all of the nuclear power sources used in satellites have (in simple terms) a hot isotope on one face and a heat sink facing deep space on the other.
Since a particular face can be switched from heating to cooling by switching polarity you can connect one to an 'H' bridge and maintain a precise temperature on that face by toggling polarity to heat or cool as needed.
That's interesting. I didn't know satellites used nuclear power sources.
Radioisotope thermoelectric generators
If I'm reading the specifications (and the linked guide) correctly, isn't it dissipating more like 100W as heat, and also transferring about 60W worth of temperature from the "cold" side to the "hot" side?
This is what makes thermoelectric coolers so impractical for many cooling applications and why they're typically coupled with a phase-change refrigeration system (which, despite its old-fashioned-ness is much more efficient) in high-performance CPU cooling.
The consumption is something around 107.8W. (15.4V x 7A) at peak of course...