Thermoelectric Device Heat Sink

Discussion in 'The Projects Forum' started by electech5, Jun 3, 2014.

  1. electech5

    Thread Starter New Member

    May 14, 2014
    11
    0
    Hello,

    I am designing a Peltier Device using Tellurex Thermo-eletric module I am using it to control the temperature inside a sealed wooded box I will build. I am having trouble determining what size of heat sinks to use with the device. Do anyone know how to figure out what size in required?

    I understand how to build the device and how the current in the device causes a warm side and cold side. My design is to have a heat sink on both sides of the device and a fan on one of the heat sinks to move the heat away from the heat sink. Would this entire unit sit inside the sealed box or would the heat sink with out the fan on it have to not be isolated from the other side?

    Thanks
     
  2. Sensacell

    Well-Known Member

    Jun 19, 2012
    1,128
    266
    If it's all in the same insulated space, it will all just get hot.

    You need to have the device go through the wall of the insulated box, with a large heatsink on the outside to dissipate the heat to the ambient environment.

    A fan inside is not strictly necessary, but will help even out the temperature gradients in the box. A fan on the external heatsink will improve it's efficiency greatly.
     
  3. wayneh

    Expert

    Sep 9, 2010
    12,094
    3,033
    The heat sink should be at least as large as the peltier. Computer CPU heat sinks with fans are a popular solution.

    The hot side will need to dissipate all the power drawn by the peltier, plus about 10% more to cover the heat being removed from the cold side. You should be able to calculate how much heat this is, and the temperature difference above ambient that you can tolerate. Your heat sink and fan will have specifications to compare to your heat dissipation needs.

    The cold side needs to absorb the amount of power required to cool your enclosure, the heat leaking in from all sources. There are heat sinks designed for still air, and you might consider this on the cold side if you'd rather not have a fan inside.
     
  4. electech5

    Thread Starter New Member

    May 14, 2014
    11
    0
    Thanks for you help, I am desiging a temperature control in combination with humidity control. The reason for the fan is to not only help with the circulation of the heat but also in front will be a damp sponge to control the moisture in the air. I will pump a small amount of water into the sponge and have the fan out to add moisture.

    I did kind of figured part of the unit would have to sit outside the box, but was hoping I was wrong. The thermoelectric device is quite skinny, any suggestions on how i have have on side inside and one outside while still maintaining an enclosed box?

    Thanks
     
  5. Metalmann

    Active Member

    Dec 8, 2012
    700
    223

    Maybe mount that Peltier on 1/8 dense plywood, good quality ApplePly, and insulate the hell out of it?:confused:
     
  6. wayneh

    Expert

    Sep 9, 2010
    12,094
    3,033
    I put mine onto aluminum stock that could be built into a larger structure. Everything but the peltier was insulated. The aluminum surface touching the peltier was sanded smooth and a small amount of thermal paste applied.
     
    Metalmann likes this.
  7. electech5

    Thread Starter New Member

    May 14, 2014
    11
    0
    wayneh,
    I don't quite understand how it could be built into a larger structure. would it not always have the issue of having to isolate one side of the device from the other?
     
  8. #12

    Expert

    Nov 30, 2010
    16,261
    6,768
    Your thermo-electric device has 2 large, flat sides. The cool side has to be in the box you want cooled and the warm side has to be exposed to free air, sometimes with a fan helping it. You can build this into a jewelry box or an outhouse. It doesn't matter, as long as one flat side is exposed to, "inside" the box and the other flat side is exposed to, "outside" the box.
     
    Metalmann likes this.
  9. wayneh

    Expert

    Sep 9, 2010
    12,094
    3,033
    A cooler AND a fan, on an outhouse? Sounds like royalty!
     
    #12 likes this.
  10. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,282
    1,235
  11. wayneh

    Expert

    Sep 9, 2010
    12,094
    3,033
    For what it's worth, they're much more efficient when operated well below the rated maximum current, e.g.. at 4A for a module rated at 12A. That means you may need several of them to get the same amount of cooling, but this makes more room for heat transfer also.
     
  12. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,282
    1,235
    How big is your box? Maybe a 6 pack cooler would work???
     
  13. #12

    Expert

    Nov 30, 2010
    16,261
    6,768
    I'm getting a little weird today. :p
     
  14. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,282
    1,235
    And a six pack.:rolleyes:
     
  15. electech5

    Thread Starter New Member

    May 14, 2014
    11
    0
    Hello,
    After your comments and doing some more research, i'm going to try to control the temperature and humidity using a DHT11 Humidity and temperature sensor. I have the Arduino Uno i will hook it into. I want to try a few different methods of testing both the temperature and humidity separately. Once the best method for both quantities are found I will combine them to see them work together.For the temperature i'm going to try to put together a sealed box with the one heat sink on the inside and one on the outside, with the Thermo-electric device. My sealed unit will be a small Tupperware that I can seal. Another test was heating up a coil to control the heat and still determining a way to try to cool it down

    The humidity I am going to try and control by putting a sponge in front of a fan and when the humidity is to low I can pump water into the sponge slowly and have the fan add mositure. another option to add moisture is to use a misting stone and have tubes run throughout the box and have a fine mist moisten the air.

    I am a little stuck on how I can remove moisture from the air, any suggestions?
     
  16. #12

    Expert

    Nov 30, 2010
    16,261
    6,768
    The dew point of the air is related to humidity. Any time the cooler surface is below the dew point, condensation will form. One way to find the numbers is with a psychrometric chart. Bottom line, lower the temperature of any spot in the container below the dew point and you can drain the water out with a tube.
     
  17. electech5

    Thread Starter New Member

    May 14, 2014
    11
    0
    The dew point currently, coming from the DHT11, is around 10 degrees with room temperature and the moisture in my house which is currently 80. I will keeping my enclosure at 23 degrees, so i don't think that it is a huge issue in the climate i am in, however just it case i should consider the possibility. Currently I have the sensor connected to the UNO and 4 LEDs each on their own outputs, triggering at high and low temperature points and high and low humidity points. My next step is to put together the peltier devices. I am going to try a few different types and size of heat sinks to see what works best.

    The peltier device operates off of 12V (http://www.hebeiltd.com.cn/peltier.datasheet/TEC1-12706.pdf) so I was thinking of either using a transistor or a relay. what would work best? I have read the transistor is better for fast switching applications and relays are more economical if speed isn't an issue. How would I determine if this application needs speed?

    Thanks
     
  18. #12

    Expert

    Nov 30, 2010
    16,261
    6,768
    For this, you need to think about the thermal time constant of the box and cooler, and the range of acceptable temperatures. In terms of using the AC power line, If your accuracy goal is +/- 1 degree, and the box requires more than .016 seconds to change by one degree, you can do this with a triac.

    Usually, that's enough explanation for people to realize that fast control is absurd for most heating and cooling applications. However, a transistor is very reliable and not noisy like a relay. I'd use a transistor if the power doesn't get huge. 3 or 4 amps is easily in the range of a mosfet. 30 or 40 amps is moving into the range where you would want to do the math before deciding.
     
  19. wayneh

    Expert

    Sep 9, 2010
    12,094
    3,033
    I used an IRF540N to switch about 4A to my 12V TEC. (It is rated to 10-12A as I recall). You might be interested in my little refrigerator project.
     
Loading...