Suppose that we want to cool down our old Pentium 4 423 @ 1.4G. This had cooling needs reaching up to 72 Watts.
From the first characteristic curve we find that we will draw about 6 amperes (6.05A) and that a temperature difference of about 5 degrees.
Then we go to the second diagram. Following the line of the 6.05A, we see that for 5 degrees we need to have supply 10.8 Volts to the Peltier.
Cooling a peltierNEVER POWER A PELTIER WITHOUT A COOLING METHOD
A Peltier thermo element mounted on the surface of a heat-sink with a cooling fan
That is a rule if you want your TEC to go on working. The Peltiers generate a lot of heat on the hot surface, far more than the heat they dissipate. And that is because the TEC itself draws a lot of current that generates itself heat due to I2xR heat losses.
The most common cooling method for TECs is the air cooling. The hot plate of the TEC is mounted (with heat transfer paste always) on the surface of a heat-sink that carries a cooling fan. The heat sink must be calculated in a way that it will be able to draw all this heat power that comes of the TEC's hot plate.
As you will find out later on this article, air cooling is sometimes inefficient. You should consider building a project with water cooling if you would like to have maximum cooling results from a TEC, as a heat sink to handle the heat power generated under heavy conditions would be almost impossible to be built.
How to calculate the power that the TEC cooler must have
It is sometimes difficult for people to follow mathematical forms and types, but it is always necessary to do for the sake of efficiency.
At first, let's give some symbols to our terms:
TAMB = Ambient temperature
QC = Heat load (in watts) that we need to cool
TH = Peltier hot plate temperature
QΗ = Total heat sink load (watts)
C/W = Heat sink efficiency (Degrees C/watts)
PE = Peltier electrical power input (watts)
V = Voltage that the Peltier is powered with (Volts)
I = Current that the Peltier draws from our power source (Amperes)
Now to connect all them together!
TH = TAMB + (C/W)(QH)
QH = PE + QC
PE = V x I
And what have we come up with? Suppose that we have a Peltier that is powered with 10.8 Volts, draws current 6.05 Amperes and pumps 72 Watts of power. This system is placed in a box that has inside 25 degrees. What can we do with the above formulas is to calculate the appropriate heatsink power needed in order to cool efficiently our peltier.
The QH parameter is the one that we are currently interested at. This is the total thermal load that needs to be sinked from the heat sink attached to the TEC's hot plate. The formula is:
QH = PE + QC
PE is the electrical power drawn by the TEC, that would be VxI:
PE = 10.8 x 6.05 = 65.34 Watt
QH = PE + QC => QH = 65.34 + 72 = 137.34 Watt
So, we need to place a heat-sink to the hot plate of the Peltier that will be able to sink 137.34 Watts of thermal power.
How to calculate the temperature of the hot plate of the TEC during operation
Another very important parameter that needs to be taken into account is the temperature of the hot plate of the TEC that needs to be cooled. This parameter is the TH. Let's calculate it for our P4 cooler!
TH = TAMB + (C/W)(QH) => TH = 25 + 0.55 x 137.34 = 100.5°C A typical heatsink efficiency is about 0.55 C/W
Now you understand the reason for this calculation. The hot plate will reach up to 100 degrees during operation. The TEC must be cooled efficiently otherwise it will be destroyed. The temperature will be able to boil water!
Advantages and disadvantages of Peltiers
Advantages of Peltiers
Peltiers are a cheap and easy way to achieve sub-ambient temperatures for any reason. They are solid state and minimal in dimensions, and this gives them also the advantage of portability, something that other systems are incapable of. Also, they are solid state with no moving parts inside. This makes them extremely reliable and needs of no maintenance at all.
Disadvantages of Peltiers
TECs comes with a series of disadvantages that makes them sometimes completely inappropriate for cooling parts. TECs are extremely inefficient. They draw a lot of current to operate and generate too much heat. If a part is cooled with normal heat sink methods and a Peltier is added to boost this operation, then the heat sink that needs to be mounted on the TEC sometimes must be double in power than the original heat sink that was attached to the part! This makes also the TECs a little bit of ecology enemies.
Considerations
Cooling with TECs is not always an easy task. If a CPU with temperature load of 100W needs to be cooled, then mounting a Peltier on it and try to cool it is not always the right thing to do. Peltiers should be used combined with water cooling. This makes the system even more reliable and more controllable.
When you apply a voltage to a TEC and get a temperature difference you should have in mind that you will also cause a back voltage created. This is caused by the Seebeck effect and is similar to the back EMF caused by electric motors. It is possible therefore to get a negative linear load output curve.
A Peltier cooler should have controllable power supply or current flow. The power provided to the TEC should ne in conjunction to the thermal load and the ambient temperature. This would increase the efficiency of the TEC. You can use PWM to control the power provided to the TEC
Always have in mind that the more cooling you want from a peltier, the bigger the difference between the amount of heat your CPU generates and the rated cooling power of the peltier. This is because the efficiency of the TECs is getting even lower as the temperature pumping demands are increased. In such situations it is good to have as rule that "The bigger the peltier the better the result". However, bigger peltiers will require much more cooling. Sometimes air cooling will be not enough. Water cooling is the only solution if you want extreme sub ambient temperatures to be achieved.
Be aware also that underrated TECs applied to CPUs for cooling may not only cool the CPU but heat it even more. There are some small TECs 25-30 watts around in the market. Forget about them. You need brute power, even if you use just a part of it.
The same will happen if you do not have a good cooling system for the TEC...
At 26 August 2015, 3:53:21 user pev69 wrote: [reply @ pev69]
The explanation is somewhat misleading. The P- and N-type semiconductor pellets do not work as pairs or junctions, but work individually. The whole TEC could be made of only P or only N pellets, but because the electrical connections between pellets would then need to zig-zag between the cold and hot sides, they would reduce efficiency by conducting heat back to the cold side. So the only reason to use both P and N (and alternatingly) is to have the electrical connections between adjacent pellets stay on one (cold or hot) side.
Hello sir/madam
Informations are very much useful, but how to calculate the heat transfer in watts depends on the application.. Could tell me how to make a small thermo electric cooler
Kindly reply me @ mahesh.chandran@lnttechservices.com
Dear Sirs,
I would be interested to know the effciency of a Peltier module when temperature has reached the set point and switches off. Since the module is a heat transfer unit and we have two heat sinks with a high temperature diference surely the cold side will transfer the heat to the cold side. If that could be the case, I would imagine a great loss of heat.
Many thanks,
Chris
@Giorgos Lazaridis
soory i was wrong, didn t pay attention to the description (i watched it late at night), the copper was used as heat sink a peltier chip is hidden between two stainless steel plates
@Giorgos Lazaridis you can make thermocouples out of iron and copper,
to make the most efficient possible peltier you would need doped metals but for a "fonctionnal if observed with a voltmeter" iron and copper should do it, i found a vid on youtube of a guys actionning a small motor with a weird peltier made of inox and copper tubes
@ngc1300 as far as i know, simple metals won't work. Peltiers have semiconducting materials, not same as transistors or diodes. They have lots of additives to allow many free electrons to form. It won't be that easy like getting two metals and connecting them. You can start by adding additives to pure silica.
@Giorgos Lazaridis
i'd like to try at least, i'm kind of hyperactive and i love doing physics experiments and building stuff (this would be two birds one shot) and I actually would need to build one for a presentation that counts in my exams, i dont need it to be efficient, just to be able to show that it works, but the more efficient the better
if i need pure metals i can buy them there http://elementsales.com/pl_element.htm#f
i can use my schools workshop if required
the peltier design seem pretty straight forward if i knew what materials to use it could be simple,
i might go for iron and copper for the P and N materials, but i do not know what to link them with, if it matters or not
@ngc1300 you want to make a peltier???? this is the first time i hear this. i cannot help you, my knowledge is poor. But if you manage to make one, please let me know. send me an email if you do.
hi
could you please tell me what kind of preferably easily accessed cheap metals I could use to build one?
what to use as N as P and the plates between
I tried to find some but didn't succeed... every single tutorial to build a pelted module sais "buy a Pelletier module" then how to use it as a thermoelectric generator
I would need to buil one for a presentation (I m in a grade between high school and ingenniering school in France) and none of my teachers are capable of givin me any useful information (they really suck as teachers...sadly)
At 30 October 2011, 15:58:10 user Prasad wrote: [reply @ Prasad]
Very impressive this TEC module theory? Just wonder if anyone would have used this to pump water from a deep well.
ok.
my question is that if efficiency is so low(eg. <10%) then the rest of the input power is being wasted. So is it true that the wasted energy is in the form of heat(which is probably why they need cooling always)???
also if for instance we are using it to cool a pc down. As mentioned in article we would have to more than double the coolong right. So why put the TEC instead just double the cooling alone? In other words if we use Peltier chip for cooling, it will along with the CPU heating, produce more heat itself which we will have to cool down altogether. so what is the resulting benifit of using it???
please help i am confused.
Thanx
At 18 November 2010, 4:15:09 user Zach wrote: [reply @ Zach]
You should also mention that condensation (or ice) might form on the cold side which can damage semiconductors like a CPU.
Kammenos, thanks.I shall go through the material in more detail. I visited the website www.top-cool.eu. I saw the product with two water circulations around Peltier. I assumed I can send water to be cooled through one circuit and pass water through the other circuit which will pass through hotside of peltier. My knowledge is very preliminary and I shall go through the details. Thanks for the warning.
Hello Gurumurthy,
If i understand correct, you want to cool down water from 80 to 20 deg? With 8 liters per minute flow? I think that the Peltier is not capable to do this. If you use a huge scale peltier or multiple peltiers in series, then the power consumption will be completely inefficient. Peltiers may have efficiency as low as 0.5%, which is usually unacceptable.
The peltier is not a heater or a cooler. It is a heat-pump! Draws energy from one plate and transfers it to the other. In other words, regarding the rice application, the cold rice will not become colder, as the hot rice will prevent the peltier to transfer to it more energy. The peltier would work perfectly and with high efficiency, if it had to draw energy from the hot rice and transfer it to the cold, meaning that the hot rice would become cooler and the cold would become hoter.
I am very impressed with the article. I bought a Peltier chip in the year 1995 for deploying in some crazy application like: In India we take curd rice cold and Sambar rice hot. So I wanted to design a tiffinbox to keep one section cool and the other hot. The design never happened due to other work pressure. I got this article when I am recommending peltier to cool a liquid flow of 8 lpm from 80 deg C to 20 deg C. I am sure my friend will learn a lot.
Home
Projects
Experiments
Circuits
Theory
BLOG
PIC Tutorials
Time for Science
Contact
Forum
RSS
Reddit this
HOT in heaven!
NEW in heaven!
