Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Wire temperature control

  1. May 17, 2015 #1
    Hey,
    I would like to know how I can determine and control the temperature of a wire at a specific point of temperature.
    I want to control the temperature from 100-300 celcius.
    This can been achieved from the ampere load of the wire?
    Or please send me your opinion of how it can done in a different way
     
  2. jcsd
  3. May 17, 2015 #2

    Svein

    User Avatar
    Science Advisor

    That is what every supplier of electric panel heaters are concerned with. They usually do not bother with the temperature on the actual wire, but on a point close to it.
    They control the heat in one of several ways:
    • Switching between different wires
    • Switching between wires in single, serial or parallel
    • Pulsing the power through the wire using a triac
    Measuring the heat can be done in several ways, the simplest are:
    • Using a bi-metal switch
    • Using an NTC resistor
    The trick is to electrically isolate the measurement device from the wire...
     
  4. May 17, 2015 #3
     
  5. May 17, 2015 #4
    Thanks for your reply
    I want to increase the temperature of a wire (nickelium or kanthal) to a specific temperature from 100c to 300c.
    My guess was to measure the resistanse of the wire as it increases proportionally to the temperature of the wire.
    In that way I may manage to control temperature.
    I wiil use a MOSFET that will change the voltage of the output so as to heat the wire.Everything will be handled from a MCU.
    I imagine I need and algorithm?
     
  6. May 17, 2015 #5
    I think your biggest problem will be the current needed to heat up the wire. Now, if you wanted to smoothly regulate that current to achieve any temperature in that 100-300C range, you will need a component that does so by varying its resistance (I.e. a transistor). However, now you have the problem that this current-regulating component is essentially a resistor in series with your wire, and thus it will consume a good chunk of power itself. Which results in heat that will likely fry the transistor.

    I think Svein's suggestion of pulsing the current is the only way of achieving the whole temperature range.
     
  7. May 17, 2015 #6
    Please can you explain in different words Svein's post?
     
  8. May 17, 2015 #7
    Or you Svein?
    Thanks!
     
  9. May 17, 2015 #8
    Well, as I said, you likely won't be able to control the current in the wire without majorly overheating the current-controling component as well.

    However, what you can *can* do is to switch on and off the current very fast with a transistor. Because you are essentially switching between infinite resistance in the transistor (so, no current and thus no heat up in the transistor) to zero resistance (no internal heat-up either because of zero resistance), you can control the temperature of the wire by rapidly switching on and off the current. Longer pauses between switched-on times: low heat. Short pauses with long switched-on times: high heat.
     
  10. May 17, 2015 #9
    Thanks it is good start,I will keep you updated.
    Thanks
     
  11. May 17, 2015 #10

    boneh3ad

    User Avatar
    Science Advisor
    Gold Member

    This is precisely how a constant-temperature hot-wire anemometer works. You set up a Wheatstone bridge where one side has the wire and the other has a resistor that "balances" the wire. Depending on the resistance ratios between the two sides, more or less current is passed through the wire until the two sides are back in equilibrium. This is all in a feedback loop so that a constant temperature is maintained. You can adjust the temperature by adjusting the resistor balancing the wire. Normally, the setup is used to correlate a voltage to a velocity of air moving across (and cooling) the wire, but you wouldn't really have to do that portion of it if all you need is a wire at a controllable, constant temperature.
     
  12. May 17, 2015 #11
    Hmm, I don't think I see how you would use the Wheatstone Bridge for his purpose of drastically changing the current through the wire though. The bridge has no direct connection between the two arms, instead there's a voltmeter. So you can detect deviations from the operating point (which correlates with the wind blowing over it), but you can't actively change the current of the other arm.
     
  13. May 17, 2015 #12

    boneh3ad

    User Avatar
    Science Advisor
    Gold Member

    At least in a hot-wire anemometer, the voltage across the bridge is fed into an op-amp. When the two sides are not in balance (i.e. the wire is not the temperature you want for whatever reason), the op-amp amplifies the difference and feeds the current back into the bridge (there is a drawing of the simplest of such circuits here). If the wire and the other side are in balance, the voltage difference going to the op-amp is zero and the circuit operates at steady state.
     
  14. May 18, 2015 #13

    Svein

    User Avatar
    Science Advisor

    The trouble with using the current as a measure for the temperature is that the resistance of a wire changes with the temperature, the time elapsed and the air flow around the wire.

    Now, if you are going to heat the wire using AC, I suggest you look at IC's designed for that purpose. One example is https://www.fairchildsemi.com/datasheets/MO/MOC3041M.pdf . An application note is here: http://www.nxp.com/documents/application_note/AN467.pdf.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook