The Truth About Thermistors

Thermistors were so named by Bell Telephone Laboratories and the name is a contraction for thermal resistors. They are resistors with a high temperature coefficient of resistance and their resistance is a function of absolute temperature. Thermistors are not new. They have been known for about 160 years, but have had very little use because they were considered to be unstable and nonreproducible. This is no longer true.

About 50 years ago, Bell Telephone Laboratories started investigating various materials and processing procedures to determine how to make stable and reproducible thermistors. They were quite successful in their investigations and actually put thermistors into large scale telephone usage about 30 years ago. The units they developed were both highly stable and closely reproducible to standard specifications. It is only in the last 20 years, however, that thermistors have been generally available for large scale industrial use.

There are so many thermistor manufacturers, how can I choose a high quality sensor?

  • The element is the key. The PreCon Raw Thermistor Chip is made of metal oxides compressed and heat treated to a coherent nonporous mass. This treatment process is what gives the thermistor its basic curve. The material is trimmed for the proper resistance at 25°C, leads are attached, and various chemical treatments are applied. The sensor is then encapsulated in a plastic cup to become a PreCon ST-R3. During the assembly process the sensors are 100% screened which assures the high interchangeability tolerance and long-term stability for which PreCon Sensors are known. PreCon utilizes a chemical compound more pliable than glass and yet more resistant to moisture than the epoxy used by other manufacturers. This process allows PreCon to warrant their sensors to be free from drift of more than 0.24°F for 5 years. The 8-foot leads attached to the sensor provide isolation from moisture which will destroy a thermistor.

    • My sensor reads accurately at room temperature but reads incorrectly at other temperatures. What is wrong?

  • Check the Characteristic Curve and Tolerance. The Characteristic Curve is the temperature/resistance curve that is established by the manufacturing process. This curve is the one most often used to describe the basic sensor characteristic. It does not describe the performance of the sensor. A major manufacturer of high accuracy thermistors catalogs four different sensors with the same characteristic curve. Each sensor becomes progressively more expensive as the sensor performance improves. To even be considered as an equal, a competing thermistor must have an identical curve. Any deviations in the curve indicate a questionable material mix and manufacturing process.

    Tolerance is the ability of the thermistor to adhere to its characteristic curve. It is tolerance that begins to separate high quality thermistors from general purpose sensors. Good tolerance characteristics of a thermistor can be obtained in several ways. PreCon obtains tight tolerance by excellence in the manufacturing process. If the manufacturing process is correct, then the thermistor tolerance will be good. Some manufacturers use a selection process to obtain sensors that will meet predetermined tolerance specifications. This method is necessary when the manufacturing quality control is poor or the manufacturing techniques are not available for producing high quality sensors. In any case, sensors obtained by the selection method are not likely to maintain their tolerance over time or through temperature cycling which is characteristic of HVAC applications. PreCon warrants its tolerance over a period of 5 years.

    •

    How can I be sure I am getting a stable sensor that will not drift with time?

  • Stability is perhaps the most important, and yet the most elusive, of the key words which are used to determine quality. Stability is the term used to describe the characteristic drift that some thermistors display. Thermistors become more stable with time, with the most change in tolerance occurring in the first 6 months following manufacture. Stability is a function of the manufacturing process. Although the process is difficult and expensive, a high quality manufacturing process will yield pre-aged, highly stable thermistors. Since it is difficult to test a thermistor for stability except under laboratory conditions, the sensor manufacturer (assembler) should be willing to guarantee stability of the sensor they are selling. PreCon thermistors are inherently stable at 0.027°C a year. A competing thermistor cannot be considered an equal unless the manufacturer will offer the same guarantee.

    • Other companies sell "Bare Bead" thermistors mounted in space wall mount enclosures. Why doesn't Kele?

  • Kele sells PreCon Thermistor Sensors. Simply stated, PreCon Sensors are designed to interface with automation systems more accurately than bare bead sensors. Most thermistor-based automation systems in use today will read room temperatures 1°F higher than the actual temperature if a bare bead sensor in a ventilated plastic enclosure is used to measure the space temperature. Heat is generated in all resistance temperature measurement devices as a result of current flow in the measurement circuit. PreCon Raw Sensors will dissipate the heat developed in the measurement circuit almost three times better (2.7 mW/°C) than a bare bead (1 mW/°C) sensor.

    • What is a heat dissipation constant?

  • Heat Dissipation Constant is the power in milliwatts required to raise a thermistor one degree centigrade above the surrounding temperature. Even if the zero power resistance/temperature curve claimed by the manufacturer is close enough to the PreCon Type III curve, so as not to cause readily apparent problems, other factors can cause significant and troublesome deviations from the actual performance of PreCon Type III Sensors. Wire size, type, and method of connection to the chip itself critically affect the ability of the sensor to dissipate the heat generated in it by measuring the flow of current through it. This is called "self-heating" and actually causes the complete sensor assembly to give readings that differ from the characteristic temperature curve. The wire PreCon uses is designed to limit the effects of self-heating, reverse temperature transmission, and moisture migration. The encapsulation material used to coat the thermistor provides a low mass and high conductivity for good heat transfer characteristics. The PreCon ST-R3 Raw Sensor has a Dissipation Constant in still air at 25°C (77°F) of 2.7 mW/°C. Most "similar" bead sensors on the market will self-heat at almost 3 times the rate of a PreCon ST-R3 Type III Sensor. A "similar" raw bead thermistor sensor may read correctly with an ohmmeter but exhibit a temperature indication error of one degree Fahrenheit when applied to a computer input. An equal raw bead sensor must possess the same Dissipation Constant to read as accurately as a PreCon Type III Sensor under the same conditions.

    • How to determine your sensor's heating error:

    1. With the sensor connected to the computer, measure and record the voltage drop across the sensor.(E)
    2. Record the temperature reading at the sensor at the time of the voltage reading.
    3. Using the thermistor temperature/resistance chart, record the sensor resistance at the temperature measured above.(R)
    4. Record the heat dissipation constant for your sensor. [Typical Bead Sensors (K) = 1 mW/°C]
    5. Calculate the power generated in your sensor:
      P = E2/ R
    6. Calculate the temperature reading error:
      Error = Power Generated (P) / Sensor Dissipation Constant (K)

    Example:

  • Bead Sensor
    (Heat Dissipation Constant = 1 mW/°C)
    Voltage Measured = 2.5V
    Room Temperature = 25°C (77°F)
    Sensor Resistance at 25°C (77°F) =10,000 ohms
    Power generated = E2 / R = 2.52 / 10,000 = 0.625 milliwatts
    Heating error = P/K = 0.625 mW / 1 mW/°C = 0.625°C
    In this example the bead sensor would read 0.625°C high or in °F, 1.8 x 0.625 = 1.13°F high.