Lab 5: Thermistors sensing temperatures

Background:  All resistivity (rho) is temperature dependent, which is usually a bad thing, unless you want to sense temperature, then you can go all out, using materials that change resistance rapidly as temperature changes. Such is the case with thermistors. Remember from materials science or electricity how metals and nonmetals differ in the slope of their resistance-temperature characteristics...

At any rate, thermistors--unlike mercury bulb or liquid crystals--can be involved in converting temperature directly to voltage (use voltage divider).

Requirements:
(1) Design and build an electronic temperature sensor, starting with two NTC (negative-temperature-coefficient) thermistors. You will want the thermistors to be part of a circuit that generates a temperature-sensitive voltage. In addition to connecting with your voltmeter, your temperature sensor output should feed into LabVIEW, to be displayed as a chart waveform. Use math functions in LabVIEW to convert from voltage to temperature.

Either solder or screw-terminal connect the leads of the thermistors to the wires leading to your bridge. Simply twisting the wires together will lead to loose connections...

Use a Fluke 50-series thermocouple-based temperature sensor to calibrate your temperature sensing circuit, since the Fluke 50 will be used as the standard for the temperature tests described below. You may want to calibrate your own biosensor, at the end of your index finger, too.

(2) In the first test we'll aim a hair dryer at the Fluke thermocouple. We will guarantee that the hair dryer temperature reading will be in the range 40 to 60 deg C.

Your team gets to select which hair dryer and and what speed the test will be run. (It has to be a setting that can reach 60 deg C...) We will set the distance and angle of the hair dryer from the thermocouple. We will be able to manipulate the thermocouple independently of your thermistor mount... The hair dryer will be clamped in a vise, and the thermocouple will be positioned with a gooseneck mechanical manipulator. You can maneuver your thermistor(s) with another mechanical manipulator to be near (very near) the thermocouple. You are not allowed to put tape on the thermocouple. After we turn on the hair dryer a Post-It will cover the display of the Fluke meter.

Once you say you're ready, the Fluke will be switched to the Record Max mode: we'll tell you when a 60 interval starts. At the end of the 60 seconds we will press the Hold button and that will lock into the display the maximum temperature of the past 60 seconds.

Your challenge: Estimate what the max-temperature reading is, based on the output of your thermistor circuit (and perhaps placing your "neural-network-trained" index finger near the thermocouple...) Beware that the temperature in the air stream may vary by several degrees over a few millimeters distance! After you write down your estimate, we'll uncover the display and see what the correct=thermocouple answer is. Your estimate must be within ± 1.5°C of the Fluke for you to proceed to the next test:

(3) In the second test we'll leave the Fluke probe in stirred chilled tap water. We will guarantee that the chilled water temperature reading will be in the range 5 to 15 deg C.

As before, the Fluke display will be covered and we'll announce a 60 second interval during which the Fluke will be in the Record Min mode. At the end of the 60 sec we will press the Hold key and lock in the answer.

You need to have your thermistor in the chilled water during this interval, and be monitoring your circuit's output. Again you are to make a estimate: What was the minimum water temperature during the 60 sec interval? The position of your thermistor in the stirred water won't be critical, except perhaps for its depth.

At any rate, your estimate must be within ± 1.0°C of the Fluke for you to meet requirements.

(4) We may observe that for the 60 second intervals of the 2 tests you will produce a LabVIEW waveform with temperature as a function of time.

If either of your temperature estimates is out of range, you need to come back a day later to try again. If you are correct on both temperature estimates ON THE FIRST TRY then there is no FTQ.

What resistance thermistor should you choose?
Lower resistance thermistors may be susceptible to self heating, and higher resistance thermistors may be compromised by parallel conductance from ions in the cool tap water.

Free Advice: Read the lectures notes on temperature sensors on the EN123 website.
And look at the 2001 archive site Lab 2 writeup for advice on how to construct your thermistor-system...
Below is a suggestion for arranging two NTC thermistors:
The 10v "excitation voltage" should really be somewhere between 1v to 6v, so you can use the "6v max" on the Agilent power supply in the lab.

You may want to have one mode of your circuit specialized great sensitivity over the 5-15 deg range, and another mode for sensitivity in the 40-60 deg range. You can measure the differential voltage directly with your DMM on DC volts. If the reading seems to fluctuate too much, consider placing a 30 micro-Farad capacitor across the left and right sides of the bridge...

Possible FTQs for Thermistor Lab:
You may be asked a question based on the lecture notes discussion of self-heating in thermistors. Likely we will show you the Simulink simulation mentioned in the notes, and ask you to predict something about ΔT or Vout from the circuit in the class demo, based on changes in the thermal properties of the thermistor, or the size or shape of the thermistor.

OR

Explain extemporaneously, perhaps with the aid of a diagram, the difference between diffusion and mobility of electrons with regard to the how a thermocouple works.