Thermistors

A thermistor's resistance changes with temperature. Most thermistors installed by WERC are either manufactured by Alpha (model 13A5001-C3) or YSI (model 44033) and temperature is inversely related to resistance. To convert: Temperature (Degrees C) = 1 / [a + b * ln (R) + c * ln (R) ^ 2] - 273.15 where, R = Thermistor resistance in Ohms a, b, and c are Steinhart-Hart equation coefficients. For Alpha model 13A5001-C3:

 a = 1.28e-3
 b = 2.36e-4
 c = 9.06e-8

For YSI model 44033:

 a = 1.47e-3
 b = 2.39e-4
 c = 1.0002e-7

To measure these values with a CR10X data logger you might use the following short program (assumes a YSI 44033 thermistor). For this example we'll assume the wiring looks like this:

 CR10X SE1 --> Thermistor +, 15.0 kOhm Resistor to CR10X E1
 CR10X AG --> Thermistor -
 CR10X E1 --> 15.0 kOhm Resistor to CR10X SE1
 One note about the thermistors, the two leads of a thermistor are interchangeable.  So either could go into the measurement channel and either could be put into analog ground.

Program:

 1:  Excite-Delay (SE) (P4)
  1: 1        Reps
  2: 1        2.5 mV Slow Range
  3: 1        SE Channel
  4: 1        Excite all reps w/Exchan 1
  5: 4        Delay (0.01 sec units)
  6: 2500     mV Excitation
  7: 1        Loc [ thermRES  ]
  8: 0.0004   Multiplier
  9: 0        Offset
 2:  BR Transform Rf[X/(1-X)] (P59)
  1: 1        Reps
  2: 1        Loc [ thermRES  ]
  3: 15.0     Multiplier (Rf)
 3:  Z=X*F (P37)
  1: 1        X Loc [ thermRES  ]
  2: 1000     F
  3: 1        Z Loc [ thermRES  ]
 4:  Steinhart-Hart Equation (P200)
  1: 1        Reps
  2: 1        Source Loc (R)(Ohms) [ thermRES  ]
  3: 2        Destination Loc (Deg C) [ thermTEMP ]
  4: 1.47     A
  5: -3       x 10^n
  6: 2.39     B
  7: -4       x 10^n
  8: 1.0002   C
  9: -7       x 10^n

A note about the program. In the first program step the thermistor is excited with 2500 mV of DC electricity. after a delay of 0.04 seconds the voltage drop across the thermistor is recorded. In this circuit the thermistor is placed in series with the resistor, Rf, going between E1 and SE1. Because we know the total excitation voltage (2500 mV) and the resistance, Rf (15.0 kOhms) we can use the P59 program instruction to calculate the resistance of the thermistor based on the voltage measured by SE1. Next we convert that resistance from k-Ohms to Ohms for use in the Steinhart-Hart equation.

If you are working with an older data logger the program instruction P200 may not be in its program library (the program won't compile when you send this program to the data logger). You can gain access to this program instruction by updating the operating system (OS) in the data logger by visiting the Campbell Scientific software download site. Use DevConfig, part of the Loggernet package to send the OS to the data logger.

Manufacturing Instructions

It's a pain to embed images otherwise there would be quite a few here. Instead, look here:
http://people.iarc.uaf.edu/~bbusey/pictures/Thermistor_String_Construction/
for some more recent pictures and text on making thermistor strings.
The following text is from this WERC document:
A WORKPLAN FOR THE STUDY OF PHYSICAL, BIOLOGICAL AND CHEMICAL IMPLICATIONS OF MID-WINTER PUMPING OF TUNDRA LAKES, NORTH SLOPE, ALASKA June 2004
By Larry D. Hinzman, Johnny Mendez, Michael R. Lilly, Douglas L. Kane, Quinton Costello and Derek Dan Miller

3.2.2 Thermistor construction

The following steps describe the procedure for constructing thermistor probes: Lay 24-gauge 12-conductor cable in a location where it can be stretched to its full length. Designate one end of the cable as the top and the other as the bottom. The top end will be wired to the data logger; the bottom end will have the pair of thermistors soldered to it.

Strip approximately 6 inches of insulation off the top end of the cable to expose the bundle of wires. At the bottom of the string, strip approximately 1 ½-inch of the cable. Select two wires for each thermistor, and strip these wires about ¼ inch.

After each pair of wires is stripped, record their colors and the intended depth of the associated thermistor string in form F-001 and F-002 (Appendix B). Record the primary wire color first, followed by the color of the stripe, or secondary color. Label the ends of the wire pairs (at the top end of the cable) with the depth of the associated probe. Cut pieces of shrink tubing (about half the length of the exposed copper wires) and fit loosely over the exposed wire ends where the probe will be located. Bend each wire pair such that they are slightly apart, forming an angle with one another of approximately 30 to 45 degrees.

Take a thermistor and attach it to a digital voltage multi-meter. Set the DMM to read electrical resistance, let the reading stabilize for about a minute and record the resistance read by the meter in form F-001. On each thermistor, bend the two exposed wires in the same manner as above, and then tightly twist one thermistor wire around one of the exposed wires on the cable. Repeat with the other thermistor and cable wires. There may be some excess loose wire from both sets, which will be snipped off later. Use a hot soldering iron to solder the wires together, but first clean the tip on a wet sponge. Keep the iron close to the thermistor for as little time as possible so as not to damage the thermistor with the heat from the soldering iron. Gently pull on the wires to make sure that a good connection has been formed. Snip off the loose ends of the wires.

Straighten the wires so that the shrink tubing can be pulled up over the top of the connection. A heat gun can be used to shrink the tubing, which increases the strength of the connection. Test each thermistor by attaching a digital voltage multi-meter (DMM) to the two free wires at the top of the string corresponding to that thermistor. Set the DMM to read electrical resistance with the appropriate number of significant figures and check the resistance of the thermistor. Record the reading in form F-001. Verify that the resistance decreases in response to being warmed by the touch of your fingers. (Note that the thermistor may still be hot from the heat gun and the resistance may still be decreasing when the circuit is first tested.) A decrease in resistance with the addition of heat indicates that the thermistor is functioning. If this is not the case, there may be a break in the circuit, the thermistor may be defective, the DMM may be set up incorrectly, or the incorrect wires may have been used. Verify that the thermistors all have similar resistance values.

Cut a piece of shrink tubing to fit over the entire thermistor probe, and place the tubing over the probe flush against the cable. The tubing should be slightly longer than the probe so as not to expose the tip of the thermistor. To waterproof each thermistor probe, fill both ends of the shrink tubing with silicone. Allow the silicone to harden overnight. After the silicone hardens, use a heat gun to shrink the tubing.

To further waterproof and strengthen the probes, cover both thermistors with a third, larger heat shrink tubing. As before, apply silicone to both ends of the tubing, and let air dry for a few hours. Then shrink tubing with a heat gun.

3.2.3 Thermistor calibration

The following steps describe the procedure for calibration of thermistor strings: Select the thermistor string to be calibrated. Prepare an ice bath by filling a cooler with crushed or small-cube ice and then fill remaining volume with water. Use form F-001 (see Appendix B) to record calibration information including, time, date, name of the person performing the calibration, intended depth of each probe, wire color used for each probe, temperature of the water bath, and resistance of the probe at that temperature. Immerse all of the thermistor string except for the wires at the top of the string into the water/ice bath. Agitate the water bath continuously to make sure the water doesn’t thermally stratify.

Let the thermistor string equilibrate with the bath for approximately 30 minutes. Using a high precision thermometer, measure the temperature of the water bath and record in the form. Using a digital volt/ohm meter, measure the resistance of each thermistor probe by attaching the leads to the two wires that correspond to that probe. Record each resistance in form F-001 (see Appendix B). Using the temperature and associated resistance measured during the calibration procedure, calculate the appropriate linear offset for equation 1.