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

Sensistor question

  1. Mar 16, 2010 #1
    I'd like to create a circuit that fits the profile attached.

    It must be passive. I've thoguht about different combinations of sensistors/thermistors but can't think of anything that exhibits this third order type profile.


    Anyone? Bueller? Bueller?
     

    Attached Files:

  2. jcsd
  3. Mar 16, 2010 #2
    You don't give the resistance or temperature range. You also don't give the resistance or temperature of the inflection point. Furthermore your curve shows a positive temperature characteristic but thermistors normally have a negative tempereature characteristic.

    I'm not saying it can't be done, just give us some more information.
     
  4. Mar 16, 2010 #3

    berkeman

    User Avatar

    Staff: Mentor

    How much circuitry are you willing to use? A 50 cent uC with built-in temperature sensor and an on-chip calibration table in flash, driving an external FET could do it nicely....
     
  5. Mar 16, 2010 #4
    cant have uC, I have no possibility of a voltage source. This is a box that im shoving current into, (5uA to 270uA). I can only use "passive" circuitry.

    Thermistors and/or sensistors are fine.

    As far as the range of temperatures, doesnt matter. Let say -45C to +85C and room is at 25C. Resistance...100 to 3000?

    However, I dont really care about exact values, I just want some sort of a layout of components. You dont have to give me any values, I can do the details myself.

    Thanks!
     
  6. Mar 16, 2010 #5

    berkeman

    User Avatar

    Staff: Mentor

    What is the application? If it's for a school project, you will need to do the bulk of the work. We can provides hints (like names of components, names of circuit topologies, etc.), but you should do the legwork.

    Do you have a circuit simulator like SPICE available for trying things out?
     
  7. Mar 16, 2010 #6
    I'm more than willing to do the bulk of the work, i can make all the calculations and everything else. The only thing I'm stuck on is an appropriate physical layout for said components.

    I'm not asking anyone to design the circuit for me and even a 'hint' would be fine. However this isn't an easy thing to achieve without some sort of active inputting, if its even possible at all.

    I can make it behave exponentially, i can make it behave linearly, i can vary the slope, the shift, but I cant think of any way to achieve this third order profile.

    Also, its not homework im trying to redesign a temp compensation network. I can go with a more basic design, but this would be ideal.
     
  8. Mar 16, 2010 #7

    berkeman

    User Avatar

    Staff: Mentor

    Sounds like a good attitude. Yes, that point of inflection is a bit of a bear. You need something to fundamentally change at that temperature. I can think of one thing that's passive and senses temperature and is kind of like a relay, and has a fundamental discontinuity at a set temperture.... Can you think what it is from those hints? It won't do the curve, but it could possibly be used in combination with other components...
     
  9. Mar 16, 2010 #8
    are you thinking of a thermostat? resettable thermal switch?

    You're right, i need something to 'click' and another part of the network to take over.
     
    Last edited: Mar 16, 2010
  10. Mar 16, 2010 #9
    I once needed a specialized voltage temperature curve and used a network similar to a butterworth filter only using thermistors in place of capacitors. It would take a lot of work but I think you could develop a curve close to what you want with such a network.
     
  11. Mar 16, 2010 #10
    interesting. ill play around with that... thanks
     
  12. Mar 16, 2010 #11

    berkeman

    User Avatar

    Staff: Mentor

    Very good. It will be interesting to see what you come up with. Challenging problem!
     
  13. Mar 19, 2010 #12
    SL20T-101-40 :-)

    40C is good enough for me
     
  14. Mar 19, 2010 #13

    berkeman

    User Avatar

    Staff: Mentor

    Interesting part. Looks a lot more vibrataion-proof than a standard mechanical thermostat. Good find.
     
  15. Mar 21, 2010 #14
    good enough
     

    Attached Files:

  16. Mar 21, 2010 #15

    berkeman

    User Avatar

    Staff: Mentor

    Awesome!
     
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook