NTC thermistor

Discussion in 'The Projects Forum' started by matlark, Dec 8, 2011.

  1. matlark

    Thread Starter New Member

    Dec 6, 2008
    8
    0
    I am attempting to make a cheap/simple temp circuit with a voltage divider. I have a 30k resistor and a 30k thermistor with 5VDC at Vref. I have been attempting to just calibrate it with an oven and a mercury thermometer. I have been running up the temp in the oven and then note the voltage drop across the thermistor and log the voltage drop with the mercury therm. Doing this multiple times i would think my voltage drop at a given known temp should be similar but it is not. P/N is 192-303ket-a01 Any thoughts? thanks
     
  2. crutschow

    Expert

    Mar 14, 2008
    13,022
    3,236
    It should be the same and repeatable for any given temperature. How much variation are you seeing.

    What type of resistor are you using? Is the resistor in the oven? How stable is the 5V supply?
     
  3. matlark

    Thread Starter New Member

    Dec 6, 2008
    8
    0
    The supply is a stable lab supply, the resistor is an off the shelf 5 or 10%, the resistor is in the oven with the thermistor but the temps are not that high ( around 40-70 C). For one run I might have 2.45 Volt drop at 55 C and then 2 V for a another run.
     
  4. matlark

    Thread Starter New Member

    Dec 6, 2008
    8
    0
    The voltage drop is very gradual as the temp climbs, it is not erratic at all for each individual test run.
     
  5. crutschow

    Expert

    Mar 14, 2008
    13,022
    3,236
    Try it with only the thermistor in the oven.
     
  6. joeyd999

    AAC Fanatic!

    Jun 6, 2011
    2,689
    2,750
    You will need to measure the temperature of the thermistor itself, not the air in the 'box'. Get some thermally conductive epoxy and glue your thermistor to the temperature probe. Or, glue both to a hunk of aluminum or brass, very close together.

    Even better, get an "interchangeable" thermistor. One who's resistance vs. temp specs are known and published. Here's one:

    http://search.digikey.com/scripts/D...h_go&lang=en&site=us&keywords=PS502J2&x=0&y=0

    The nice thing about it, you don't need to calibrate it!

    Additionally, you will discover that the 30K resistor is not ideal for the temperature range you stated. Use excel and plot the expected Vout vs. Temp for your 30K value. Then adjust the 30K value till the range of interest falls on the most steep and "linear" part of the curve.
     
  7. matlark

    Thread Starter New Member

    Dec 6, 2008
    8
    0
    Ill try those options. thanks
     
  8. tracecom

    AAC Fanatic!

    Apr 16, 2010
    3,869
    1,393
    The best bet is to find a datasheet for the thermistor in question, which should provide a table of resistances at a large number of temperatures above and below the base temperature at which the resistance is 30kΩ. I have searched for a datasheet for the thermistor you are using and have not found one. The links at Digi-key and/or Mouser don't seem to connect with the right datasheet.

    Without a datasheet, you can take a step back and measure the resistance of the thermistor at the temperatures in which you are interested. Then, do the math to determine what the voltage should be from your voltage divider under the assumption that your resistor stays constant.

    Or, you could use a more common thermistor such as the attached, for which a datasheet and extended online support is readily available.

    ETA: I hadn't read post 6 very well when I made my post, and so I repeated some of what was stated there. I don't mean to imply that the thermistor that I recommended is any better that the one joeyd999 recommended.
     
    Last edited: Dec 9, 2011
Loading...