This Instructable will show you how to make a environmental friendly portable air conditioner. This machine uses a Peltier Module as a cooling  mechanism and some cooling fans for blowing air.

Step 1: Here is a Picture How It Works

This is just like the everyday air conditioner that we use. The only thing unique about this that the cooling unit has no moving parts. The only moving parts here are the PC fans.
a good improvement would be to water cool the hotside of the peliter and to put the radiator outside. peltiers use alot of electricity, they use 100w to move 100w of heat roughly.
Hi,<br>I'm trying to make something very similar to this design, but I'm having some wiring problems. Peltiers are really hard to find where I live and the only one I could get my hands on has no markings and no instructions - the only thing the seller knew about it is that it's 12v. I have a 12v 3.3A DC adaptor and when I wire it all up in parallel the LED on the adaptor blinks, the peltier works, but the fans just twitch like they're trying to start but can't. When I wire it in series it all works fine, but the peltier doesn't get very hot/cold at all. Is my DC adaptor amperage too low or too high? Any suggestions would help!
<p>Hi Rmckay5,<br><br>I found some fans will not operate outside of a specific voltage range - I'm guessing they are voltage regulated/gated. </p><p>I've had success with the cheap non-led Coolermaster fans for voltages above 30-34V DC. <br><br>They run. </p>
<p>your power supply indeed has a to low current. the reason its blinking is becous it shuts down do to the lack of power / starts up for a second and then shuts down again ( this is the easiest way to say it its actually more complex) thats why your fans are twitching. if you wire it in series the current is to low for your peltier do to the fans that requir less current.</p>
something like this is usually very hard to get working, because there are some many things that can go wrong. fans heatsinks power supplies. I have a peltier and a watercooler for it, but i still cant find a 12VDC 14 amp power supply. :(
I have found that an old PC power supply does exactly what I need. If you need help on getting the power supply itself to turn on, there are some YouTube videos to help.
<p>not all PC power supplies can supply enough current on the 12V rail to run a peltier, most of them that I've used will shut down when confrunted with a high drain device such as the peltier. </p>
oh oh oh u could make the 14 amp supply from an old microwave transformer just chisel away the smaller 2000 volt windings and wrap it with some 10 gauge wire until it puts out 14 amps at 12 volts but i think that wire thickness determines the amount of amps idk im not sure and u could just buy some beefy diodes to make the voltage dc and a capicitor but on the other hand is it possible to force too power into a peltier or will it just use what it needs?
u can also buy an rv power converter they usually put out around 45 amps
<p>You can use a Computer SMPS, there is absolutely no problem with that. Go ahead. (y)</p>
ive actually already got a power supply :) my xbox 360 just dies so Im taking the power brick from that is has a 16 amp 12v rail and a 1 amp 5 volt rail
ur lucky. im gonna try my idea and if ur peltier takes 14 amps wont the 16 amp power supply fry the peltier?
<p>mr. clean. Power supplies supply power. They are like a source. Devices pull what they need from it. so 14 A module will only pull 14 A from power supply :)</p>
no 16 amps is the max current that can be drawn from the psu, 14 amps os the amount the TEC will draw, so it wont fry. btw you can get 360 PSUs on ebay for $15-$20
Ok thanks now i don't have to worry and that's a good idea i think i will get 1 of those or i could make a power supply (its gonna be a giant wallwart) and i could spend the money on a giant peltier idk.
do whatever you want :) It's all for fun anyway.
try amazon, ebay, ask.com,google.com
I already got a power supply a loooooong time ago, its an xbox 360 psu, those things are awesome!
A lot of computer power supplies will do this no problem.
You should use two (or 3) power supplies together with more peltiers.
where to get the heat sinks on web?
<p>Hi, As we know peltiers are only 15 to 20% efficient as compared to A/C refrigrant gas type, but what if Peltiers are used as Heat pump inplace of compressors in A/C units. Yes they will consume more energy, but where is the problem?? We can use free energy to supply it that power. The Renewable energy source, Yes!! The sun. We can use Solar panel + invertors + Peltier TEC and can make such an assembly which can circulate water through cold and hot side of peltier. The we can run that water through Cooling coils as is done in normal AC. The more the sunlight/heat we get more will be energy produced and more heat will our TEC pump out of the room. In Areas where there is hot temperature.... This will prove very effective!!!! and you don't have to pay any electricity bill too. Just some intial investment in Solar panel, panel invertor and TEC A/C assembly. <br>We can call it - SOLAR THERMOELECTRIC AIR CONDITIONER. :) :) :)</p><p>If you want to know how to build one!!! contact: bhimeshsharma@gmail.com</p>
I&nbsp;am a professional designer and user of TEC/Peltier modules <br /> <br /> This instructable, whilst very enthusiastic, and very nicely illustrated won't work as advertised. Moving HEAT&nbsp;is not the same as creating a temperature difference. &nbsp;A Peltier cell will only create a maximum delta T of about 50 C, and pump NO&nbsp;heat, or it will pump its &quot;rated power&quot; at ZERO delta T. In truth, getting 25 C at half power, in normal ambients is doing well.<br /> <br /> Real airconditioners can cool more than their electrical input, Peltier cells can't technically, the Coefficient of performance of a real aircon is &gt;&gt; 100%, for a TEC&lt; 100%. <br />
<p>Peltier have an average efficiency of 15 to 20% while standard A/C are about 60%<br><br>the only advantage of one using a peltier A/C over a pump one i the reduction of moving part (thus long term reliability) and possible noise reduction caused by the pump, but they could have improved that since i still run an hold 1995 A/C :)</p>
Doesn't a device with COP &gt; 100% break the laws of thermodynamics? Seems like if that were true, you'd be able to harvest the difference in output to power the system, which... you obviously cannot.
<p>At least in theory it should be possible to make a heatpump power-plant. The energy is absorbed from the surroundings. The problem is to make enough heat to run a small power-plant with losses here and there, with a fairly expensive setup(compared to potential production).<br><br>A heatpump regularly has a COP above 3, for heating purposes(1 part electric=3+ parts heat output, depending on outside temperature).</p>
Its easier to point you at Wikipedia than to explain here, but suffice to say, no, thermodynamics isn't violated. ........
<p>as a slight improvement, you should put the hot side on top. Because hot air then to rise and cold air &quot;drop&quot; your hot part would not have a tendency to heat up and natural convection could improve airflow.<br><br>other then that, anyone have an idea were to get cheap heatsink?</p>
Is it my cause problem to the peltier with 12volts and 15 amps of supply? I want to produce a very cold cooler.. but the peltier is not enough of its cold how should i do it,,
<p>It is fun to read comments from some of the users. No offence but many don't know about peltier at all. I have been playing with them for past 2 months. They catch your attention when you see them at work. <br>First of all, do not connect peltier with a battery. It will overheat and fry. <br>Attach a heatsink or someother mechanism like water flow to keep its hot side cold.<br>Which one is hot side and which one is cold?<br>The side with text written is usually the side that gets cold when current is applied.<br>I am in China and its super easy to buy them. You will not be able to get them from local stores, try online shopping websites.<br>All peltiers can create electricity as well if you create temperature difference between its both sides. Warm its warm side while keep cold cold.<br>To start with them by the TEC-12706 its should be dead cheap like under 5$</p>
<p>5 for $10 at ebay ;-)</p>
<p>Any reason why there are two heatsinks at the cold side, or is it just what was available?</p><p>I have ordered a peltier element to do something similar(but to dry the air).<br>Mounting the cold side at the top might be unvise. Reason is that the cold side will get condensation that will flow downwards and might creep into the electrics.</p><p>My plan is to mount it ideways, and thinking that at least the cold fins should face downwards. Also that there should be some kind of reservoir. </p>
<p>how big is this unit...once it's built? how portable is it?</p>
<p>I hace recently purchased TEC1 12706. Provided power input to it from a 300 W SMPS. Volatge 12 V. Room temperature was ~35 degree . Heat sink and fan applied on hotter side that of CPU hitsink and fan . When I tried to measure the DC current ... using clamlmeter it showed reading to be nearly 1.6 A for peltier . I think it should consume 6 A why it is consuming less current.</p>
<p>a few notes, a radiator with water on hot side would be great addition. as your design is you are blowing both hot and cold air into a room (I admit they are blowing in different directions though) with a water setup on the back you could take the heat outside and let the air blow on it to cool it down. the addition of several items like a pump will be required.</p><p>also using 2 or 3 peltiers piggy backed will be better than 1</p><p>and last but not least is you may encounter an icing issue.</p>
hi <br>i want to know if i want to use it as heater what will i do b/c whenever it is used for cooling heating side is connected to the heating side so if we reverse the polarity than cooling side become hotter so is there heat sink is required for heating side?
attach it on the computer... 28c idle 60c full load!
Now that's a great idea!
or use thermal tape double sided, wich would be better. then u could take them apart. if u screwed up on some thing.
how expensive is one of this peltier coolers?
http://tegpower.com/ <br>go there and check them out
The prices on these can be from a few dollars to a few hundred. You have to choose your sources well.<br><br>Personally, I find that eBay has the best prices. I just bought two modules and each cost about $4, free shipping.<br><br>http://cgi.ebay.ca/Cooler-Peltiers-TEC1-12706-Thermoelectric-Fine-12V-60W-/280705809456?pt=LH_DefaultDomain_0&amp;hash=item415b5ebc30
http://tegpower.com/ there they have them from 25$ units up to 2000.00$
well one good place is to check out this link http://tegpower.com <br>go there is yr looking 4 thermal electric generators. yr looking at 85% heat conversion or soo <br> <br>
What awesome presentation. A pleasure to read. Well done.
If you follow the instructions you will fry your peltier. I just fried mine. A 7AH cell in good working order will overpower it within about 30 seconds. Melted the solder right out of the side before I knew what happened. (And yes I had a substantial heat sync). You need something to limit the current if you are going to use a battery.
depends on the voltage you supply the peltier with- but I would agree generally it would be better to limit the current somehow. But you can measure the resistance of the peltier when it's at room temperature, and calculate the expected amperage at whatever voltage battery you use (V=IR).
A peltier doesn't work that way. It's like a diode- it's made of semiconducting materials, therefore different rules apply. First, here's a quick comparison of voltage (or &quot;potential difference&quot;) versus current for resistors, diodes and filament lamps:&nbsp;<br> <br> <a href="http://www.bbc.co.uk/schools/gcsebitesize/science/add_edexcel/controlling_current/resistancerev2.shtml" rel="nofollow">http://www.bbc.co.uk/schools/gcsebitesize/science/add_edexcel/controlling_current/resistancerev2.shtml</a>&nbsp;<br> <br> In a graph of voltage vs current, the derivative (i.e. the slope) of the line at any point is the resistance. When the graph is steep, the resistance is high. This comes from V=IR (in this case, it makes more sense as V / I = R). Now, lets take a look at a more detailed generic voltage (V_D) vs current (I_D) graph for a diode:&nbsp;<br> <br> <a href="http://tinyurl.com/diode-graph-jpg" rel="nofollow">http://tinyurl.com/diode-graph-jpg</a><br> <br> As you likely already know, when you connect a peltier to a power source, one side gets hot and one gets cold. If you apply a negative voltage (i.e. switch positive and negative), the sides essentially switch. You can see this in the graph- notice what the current does when you go further left, giving a negative voltage. These two regions are labeled in the graph as the Forward and Reverse regions.&nbsp;<br> <br> As you can see from this graph, a peltier's electrical resistance varies exponentially as the voltage changes. That's why you can't approach the problem this way (unless you have some expensive lab equipment like a curve tracer that could model your peltier for you).&nbsp;<br> <br> So how do you avoid frying your peltier? You don't limit the current (directly, at least). Instead, you limit the voltage by putting a resistor between the battery and the peltier. You need to use the correct resistor- you can either do the math or find a calculator online to determine what resistor you need in order to make the voltage whatever your peltier is rated for. You want to get as close to the right voltage without going past it... look at the graph again- the correct voltage for your peltier is labeled as the &quot;knee voltage.&quot; What happens if you are just below the target voltage? Your peltier will simply not get very hot/cold. But what happens if you are just above the target voltage? Your peltier will melt your face off.&nbsp;
Robcull, I've reread what you said and you're not wrong. But I don't think you're really right, either. <br> <br>Peltier modules will, without heatsinking, be melt away by my methodology. A, say, 5A peltier module will quickly melt with 5A, but it's rated as such that if it's well heatsunk, it can tolerate 5A. <br> <br>I know at the place I worked last semester we simply had a 48V DC, applied to massive peltier modules. The current was limited by the resistance of the modules. There was some talk about in the future matching it more adjustable by chopping the DC voltage. <br> <br>Long story short, I'm not a fan of using a ballast resistor, because it wastes energy. A peltier module is a semiconducting material, but it is not quite like a diode. It does drop resistance a little as it heats up, because it its easier to jump the band gap, but it can still be treated as a relatively constant resistance unit, at least in my experience. <br> <br>I am very very open to being corrected, and admit that I only have a bit of experience with peltier modules. Am I missing something huge? Please correct me!! But I don't think your explanation was sufficient.
I'm sorry, I should have said that you need to measure it's AC resistance with a special meter. I'm not sure why I said it so simply before. I'm going to read your comment now and double check what you're talking about.

About This Instructable


215 favorites


Bio: Join me as I build fun and random weekend projects!
More by ASCAS: Rig A Car Horn To Your Alarmclock! Build A Smartphone Projector With A Shoebox HealthBand: A Remotely Monitored Health Status Bracelet
Add instructable to: