TEC Cell Help

Discussion in 'The Projects Forum' started by shadowvolts, Jul 20, 2012.

  1. shadowvolts

    Thread Starter New Member

    Jun 22, 2012
    6
    0
    Hi all,

    I am a chemistry student and I am building a temperature controlled heat cell for an experiment.

    I am using a peltier element. I want to control the current based on a variable voltage. The voltage is computer controlled from 0-10v And I use a thermistor to get the temperature of the aluminum cell. This is all controlled through labview. I need a circuit that can switch the way the current is running through The peltier and also drive different amounts of currents based on the variable voltage. I am also able to output a voltage to switch relays.

    The items I have at my disposal.
    - DC power supply with 12v 5v and a variable supply 0-30v
    -TIP120 darlington transistors
    -5vDc coil rated relays
    -lm741cn op amps.
    Peltier elements rated at 12v
    -mje 2955t Pnp transistors.
    Resistors I have plenty of. <1Ω-1MΩ

    Thanks
     
  2. MrChips

    Moderator

    Oct 2, 2009
    12,449
    3,364
    You cannot control current and voltage at the same time.
    Choose current or voltage and the TEC will determine the other based on Ohm's Law.
     
  3. wayneh

    Expert

    Sep 9, 2010
    12,151
    3,058
    You might like my calorimeter project, just look in the completed projects area.

    A TEC consumes a LOT of current, and you haven't mentioned the capacity of your power supply. You may need to limit the TEC to what your supply can handle. Running at the max isn't a good idea.
     
  4. shadowvolts

    Thread Starter New Member

    Jun 22, 2012
    6
    0
    MrChips- the voltage will not change going through the peltier only the current, I just need a way to drive the current
    Wayneh- I have looked at the project before but Its a bit over my head. The power supply limits at 3A. Also I have a tight budget and time constraints and am limited to what I have.
     
  5. MrChips

    Moderator

    Oct 2, 2009
    12,449
    3,364
    You do not understand.
    Ohm's Law states that I = V/R

    If you change V, I will change.
    If you want to control I you have to adjust V.
     
  6. shadowvolts

    Thread Starter New Member

    Jun 22, 2012
    6
    0
    Okay I did not understand your first statement. I have a variable voltage that I want to use change the current by using a darlington. The power supply allows for the circuit to pull the current it needs.
     
  7. wayneh

    Expert

    Sep 9, 2010
    12,151
    3,058
    MrChips is correct. My TEC1209 was something like a 1.2Ω equivalent. At 12V, it would draw its max current of 10A, just like a resistor.

    But it is less efficient and in risk of damage at the max current, and needs heat sinking, fans etc. to prevent the hot side from overheating. I chose to run mine from my 5V supply, drawing maybe 4A (all by itself, no extra control, just like a resistor). I achieved temperature control within 0.1°C using simple on/off thermostat control using a MOSFET as a switch. The next, more advanced approach is on/off control using PWM to vary the on time proportion.
     
  8. wayneh

    Expert

    Sep 9, 2010
    12,151
    3,058
    Up to the limit for the power supply. And I'd not plan on running continuously at more than, say, 80% of the rating.

    So let's say you tell your power supply to set at 3V. Your TEC attached directly to the supply will draw a bit less than 3A (double check this) and everything should be fine. You'll feel a mild cooling on one side and obvious heating on the other. There's about 10X more heat delivered to the hot side than is removed from the cold side, the balance coming from your power supply.
     
  9. MrChips

    Moderator

    Oct 2, 2009
    12,449
    3,364
    If you want to raise the temperature above ambient you can use a heater element.
    A TEC is usually used for cooling though you can use it for both heating and cooling.

    If you are going to use it for cooling with a supply of 5A@12V, that's 60W. The hot side of the TEC will get very hot and will destroy the TEC if the heat is not efficiently removed. It is common practice to remove significant amount of heat with cold water plumbing.
     
  10. shadowvolts

    Thread Starter New Member

    Jun 22, 2012
    6
    0
    The Tec does max out around 2.4A. The actual element is sandwiched between a cooling system and the block that needs heated/cooled. In the circuit is there a better way to switch the way the current is running other than relays?
     
  11. MrChips

    Moderator

    Oct 2, 2009
    12,449
    3,364
    What do you mean by switch the current?
    Are you heating or cooling?
    In general, no relays are used. The TEC is driven by a high power bi-polar amplifier or a bi-polar PWM digital output.
     
  12. wayneh

    Expert

    Sep 9, 2010
    12,151
    3,058
    If you really want both heating and cooling from the same element, look into an H-bridge. They're commonly used to control the direction a DC motor turns. I'd steer away from relays, personally, although you probably could get something to work.

    A lower-tech way to both heat and cool, though, would be to just use 2 elements and control them separately.
     
  13. shadowvolts

    Thread Starter New Member

    Jun 22, 2012
    6
    0
    So I am both heating and cooling. Could I make an h bridge with the TIP120 and the MJE2955? The goal is to keep a constant temperature. By using a variable voltage not the same as the circuit could I change the amount of current going into the tec by hooking the variable voltage to the transistors? With respect to which way I want to drive the load of course
     
  14. wayneh

    Expert

    Sep 9, 2010
    12,151
    3,058
    I'm still wondering why/if you really need BOTH heating and cooling. As long as you're not operating right at ambient, you likely only need one or the other. In my calorimeter, for instance, the TEC cools at about 10X the rate at which the chamber will return to ambient conditions on its own. So I can hold at, say, 10°F below ambient with a fairly low duty cycle on the TEC. I don't ever need to apply heat - it'll warm up on its own.

    Even if you need both functions, maybe you could just manually switch from one mode to the other (reversing the leads on the TEC0 depending on the situation? Do you REALLY need fast switching from one mode to the other?

    There are H-bridge experts here that can help you if you really need this. And I'm not one of them!
     
  15. shadowvolts

    Thread Starter New Member

    Jun 22, 2012
    6
    0
    sorry for the misunderstanding.

    I want to be able to keep constant temperatures in a varying range. so say i wanted to collect data at every 5 degrees from 15C to 80C and at each point have a accurate temperature.

    As stated before i measure the temperature with a thermistor and that is controlled in labview. I then, based on the difference between the set temperature and current temp, ou put a variable voltage(-10V to +10V) to control the circuit.

    I have attached a picture of a circuit that seems to be what i need but with less current. could i modify this circuit to draw up to 2.4A across both directions of the peltier(http://www.alltronics.com/cgi-bin/category.cgi?item=04U004)?

    I would use a LM741CN instead of the OP2117 and MJE2955/MJE3055 vs the TIP42/TIP41
    circuit found at...
    http://www.edaboard.com/thread150294.html
     
    • c.jpg
      c.jpg
      File size:
      22.7 KB
      Views:
      21
    Last edited: Aug 1, 2012
  16. wayneh

    Expert

    Sep 9, 2010
    12,151
    3,058
    Increasing the current is going to present a problem: The power transistor needs about 1/10th of the current under control to be applied to the base of the transistor.

    There are several options for solving this. 1) Use a power op-amp capable of supplying that current. 2) Use a darlington arrangement on the base. This uses another smaller transistor to increase the current to the base of the power transistor. 3) Use MOSFETs instead of BJTs.

    The problem with (3) is that the MOSFET will not be linear. You'll need to control its gate voltage within a narrower range. This may not be a problem, and it's probably the approach I would take first. Don't forget heat sinks.
     
Loading...