Fire and Ice - Temperature Measurement

Temperature measurement is very important in industry. In the chemical and environmental fields, temperature sensors are mostly used for control. Thermocouples are often preferred due to their low price. The questions raised in this article are: How does it work? Why does the analog voltmeter display zero? What function do the compensation cable and the cold end compensation box have? Can a thermocouple be directly connected to a standard digital multimeter? What errors can be avoided? Let's answer these questions. Historical review Le Chatelier was the first in the world to use thermocouples for temperature measurement in 1887. It is made of platinum-platinum-rhodium (Pt-PtRh) alloy and is very suitable for accurately measuring temperatures from 0 to 1300°C. Thermoelectric effect: physical explanation In every metal, when the electrons have a low binding force to the nucleus, they can move freely in the metal. Metals that are good conductors provide an average of one electron to transport the charge. This means that the number of electrons that transport the charge is the same as the average number of atoms. The atomic density per unit size is different for each metal, so the concentration of free electrons must be different. In metals, for thermal equilibrium, that is, the temperature is the same everywhere, the negative charge, and the electrons are evenly distributed. If the temperature is different, T1>T2, then the electrons diffuse from the hot area to the cold area. This behavior is similar to that of a gas, so it can be expressed as an electron gas. Gases expand when heated and contract when cooled. (Figure 1) Like a gas, the electrons evenly fill the resulting space.

If two metals with different charge concentrations are brought into contact, such as silver (high concentration) and copper (low concentration), the electrons will diffuse from the high concentration to the low concentration until they are evenly distributed. (Figure 2) The result is a diffusion voltage or contact voltage because the metal with the lower electron count will charge relative to the other metal.

Electron diffusion can be explained by kinetic energy. For thermal equilibrium, there is no pressure difference between the two electron gases, and no electromotive force caused by the contact voltage field is generated.

In principle, the contact voltage can be measured. However, as soon as the voltmeter is connected, additional contact voltages will be generated at the junction of different metals. These can be compensated for the contact voltage with thermocouples. It will be demonstrated later using circuit version 1.

This rule includes: as long as the temperature of all components is the same, the algebraic sum of the closed circuit voltages generated by heat is zero. Different temperatures will cause voltages generated by heat.

Thermocouple temperature range:

Thermocouples allow the measurement of practically relevant temperatures ranging from -270°C (gold iron-nickel chromium) to +2800°C (tungsten rhenium 3-tungsten rhenium 25). Thermocouples made of non-metallic materials allow the measurement of higher temperatures. The

HM8112-3 benchtop multimeter of Germany's Huimei Company can measure temperature using thermocouples, platinum sensors and thermistors, as shown below:

HM8112-3 Digital Multimeter measures DC, AC, frequency and
temperature (using thermocouples, platinum sensors and thermistors)

Thermo-electromotive force voltage scale:

Thermo-electromotive force voltage scale can determine the contact voltage generated by different metal combinations. In order to identify unknown metals, platinum metal is usually used at the front end of the thermocouple and the metal to be measured is at the back end. The junction temperature is 100°C and the cold end temperature is 0°C. The measured thermo-electromotive force voltage determines the type of metal to be measured. The Kxp value given by the thermo-electromotive force scale is the type of those combined metals. It uses platinum to express the temperature range in mV/100K. In fact, platinum is rarely used due to price reasons. The voltage Kab of the combined metals A and B has the following relationship:

Kab=Kap-Kbp

Table 1. Temperature differences given by the thermo-electromotive force scale between 0°C and 100°C at the junction

Example:

Iron-Constantan: Kkonst-Kfe=-3.47-1.87=5.34mV/100K. (According to DIN standard, it should be 5.5mV/100K)

Thermovoltage Vt=Kt(Tm-Tc)

Here Kt: metal thermoelectric potential constant
Tm: measuring point temperature
Tc: reference junction temperature

Example:

Copper-Constantan thermocouple sensitivity 0.05mV/K, thermovoltage 16mV, if the reference junction temperature is 0℃, then the measuring point temperature Tm is:

Tm-Tc=Vt/Kt
Tm-Tc=16mV/0.05mV/K=320K
Tc=0℃
Tm=320℃

Thermocouple types and identification methods

Thermocouples are distinguished by identification symbols, such as "Type J", "Type K". Table 2 lists the thermocouple types according to standard DIN IEC 584:

Table 2. Thermocouple Types

There are also the old DIN 437/10/13/14, the British BS 4937 and the French NF C42-324. Each standard has a different color; according to the old DIN 43710 standard, the K-type thermocouple is red/green. The British standard is brown/blue and the French standard is yellow/purple. Using the color to identify the thermocouple is very useful, especially when replacing a defective thermocouple. Of course, it is not enough to identify it by color alone, but in fact its magnetic properties must also be measured, such as iron, nickel (chromium-free) and nickel silicate. This can eliminate several possibilities.

In fact, as mentioned above, it is difficult to accurately choose the type of thermocouple. When choosing a thermocouple, many parameters need to be taken into account (temperature at the measuring point, ambient temperature, humidity, electromagnetic fields, etc.). Condustrie-MET AG (CH8260 Wetzikon, Switzerland) specializes in the design and manufacture of temperature sensors.

Compensation cables

Compensation cables are cheaper than thermocouples (especially platinum and rhenium), they are flexible and have low resistance. Compensation cables are preferred when long distance measurements are required.

Figure 3 K-type thermocouple, connector, and compensation cable

Two types of plugs are more common (see Figure 3). The green plug (female) is the standard plug for R and S type thermocouples and is mostly used in fixed installations. Handheld instruments use a yellow plug (female) and are K type thermocouples. The metal rod is a thermocouple with a stainless steel metal seal. The thermocouple wire is made of magnesium oxide. Bending and pressing will not cause a short circuit. These metal rods are mainly used in industry, when the measuring point is difficult to reach, their bending properties can be used, or metal rods can be used when vibrations are encountered. The metal rods must be connected to special crimping ends or connected with compensation cables.

Standard circuits, reference points and compensation boxes

Figure 4. Thermocouple temperature measurement circuit

There are two standard circuits. The reference points in Figure 4 are point 2 and point 3. This is very important for absolute measurements. It must be kept at a constant known temperature (such as an ice bottle). For long-term measurements, an ice bottle is not enough. It must be replaced by something called a compensation box. The compensation box holds the reference junction and the compensation coefficient NTC for the influence of the ambient temperature at the reference point. The reference point is mostly an isothermal block that is both an electrical insulator and a good thermal conductor.

The HM8112-3 digital multimeter from Germany's Huimei Company uses thermocouples, platinum sensors and thermistors to measure temperature.

The second connection method is shown in Figure 5.

Figure 5. Thermocouple (reference junction)

This circuit often appears in the literature. Here the reference point is not the copper wire, but point 2. For absolute measurement point 2 must be kept at a constant known temperature or must be calibrated. The copper wire points must be kept at the same temperature to eliminate thermal voltages.

Advantages of thermocouples:

● Economical
● Long-term stability
● Small and low thermal capacitance
● Fast
● Active sensor
● Large temperature range (0 to 2800°C)
● Rugged

Disadvantages of thermocouples:

● Since thermocouples consist of a pair of dissimilar metals, care must be taken to prevent unwanted thermocouples in a closed circuit.
● The voltage generated is very small, and a very low drift amplifier is required to identify a resolution of 0.1 degrees.
● Depending on the type of medium, total nonlinearity must be compensated.
● The signal is very low, for example 7 to 75 μV/degree
● The reference point must be kept at a constant known temperature or the output must be calibrated.

Errors often encountered in practice:

Thermocouples are two different metals welded together. Bad welding points will cause measurement errors. Sharp bends will break the junction or short-circuit. Another error is polarity error.

Detection and repair of faults:

If the measuring junction is hot and the voltmeter still displays zero, there is most likely a break in the circuit. A bad solder joint can be found with an ohmmeter. Resistance values ​​of more than 1K indicate the fault. Any short circuit in the circuit does not mean that the output is zero. On the contrary, a new thermocouple will produce erroneous results at the short-circuit point. This can be verified by heating the measuring point: if the voltage does not change or is not noticeable, there is most likely a short circuit somewhere.

If the thermocouple is connected with the wrong polarity, it will not cause e.g. -150 degrees but +150 degrees! Depending on the type of measuring instrument, any value can be displayed. The test method is the same: heat the measuring point, if the voltage drops when the temperature increases, it indicates wrong polarity.

So-called diffusion errors are caused by the diffusion of particles from the surrounding environment into the junction. Most of them are caused by strong mechanical stress (such as bending and vibration) or high temperatures (>1000°C) applied to the thermocouple. Such bad thermocouples are called poisoned. This error is difficult to detect when the temperature indication is still correct. Diffusion errors occur when the measured value drifts slowly. When replacing a poisoned thermocouple, it is best to also replace all compensation cables, connectors, etc.

Proof: The sum of the thermal voltages in a closed circuit is zero:

Further explanation: When a circuit is made up of different metals, the different temperatures will produce thermal voltages.

Proof:

The sum of all voltages M:

-V1+V2+V3=0, therefore: V1=V2+V3

V1(Iron-Constantan): KFe-CuNi=KFe-KCuNi:

1.87-(-3.47)=5.34mV/100K

V2(Copper-Constantan): KCu-CuNi=KCu-KCuNi:

0.72-(-3.47)=4.19mV/100K

V3(Iron-Copper): KFe-Cu=KFe-KCu:

1.87-(0.72)=1.15mV/100K

Therefore V1=V2+V3:

5.34mV/100K=4.19mV/100K+1.15mV/100K

Calculation proves that when the voltmeter is directly connected to the thermocouple, the voltage measured is zero, and the temperature of all components in the circuit is the same. Thermal voltage can only be generated if the measurement and reference junction temperatures differ.