General
In this Chapter the most common sources of errors and measures for their correction are presented. The list does not claim to be complete. The details relate only to the tem-perature sensors and their leads. Any instruments connected for processing the signals will only be included if they provide feedback about the operation of the temperature sensor.
Quick checks of the thermocouples (TC) and resistance thermometers (RTD) and their measurement circuits in the installed condition
• Required test instruments:
Portable multimeter with mV and Ω ranges, insulation tester with 60...100 V = volt-age; all measurements are made at room temperature.
• At room temperature the continuity and insulation are tested; use “knocking“ to detect wire breaks.
• A TC, under certain circumstances, can probably be considered to be acceptable if R < 20 Ω (wire > 0.5 mm (0.020”) Ø); the value is a function of the wire diameter and length Rins≥ 100 MΩ (for an insulated TC).
• A RTD is also probably acceptable if R ≈ 110Ω (for Pt100) Rins≥ 100 MΩ.
Heating a TC or RTD, e.g., with a gas flame to approx. 200...400 °C (392...752 °F) (without a controlled temperature) will provide information regarding breaks, reversed polarity (for a TC), too low insulation resistance, etc.
Testing in the installed condition
• Additional instruments required:
mV-source, resistance decade or a commercially available Pt100-simulator
• TC: Disconnect connection leads; use the mV-source to inject voltages into the mea-suring circuit and check indication. Test determines whether the TC or the connected measuring circuit is in error.
• RTD: Disconnect connection leads at thermometer; connect the resistance decade and simulate the measurement resistance and check indication. Test determines whether the RTD or the connected measuring circuit is in error.
• Additional tests were described in the previous section.
If the TC or RTD has a exchangeable measuring inset, replace the inset with a test measuring inset with known values. Test determines whether the temperatures sensor or another component in the measuring circuit is the cause of the measurement error.
Error Table for Thermocouples and Resistance Thermometers
– Install galvanically isolated transmitter – Maintain a distance of least 0.5m (20”)
between signal and power leads when installed in parallel
– Use electrostatic shielding by installing a grounded foil/screen
– Use twist lead (pairs) to eliminate mag-netic coupling
– Cross signal and interfering power leads in right angel
b) Ground loops – Only one ground point in measuring circuit or measuring system “floating“
(not grounded) – affected by an
inter-fering heat source
– Select installation site so the medium can transfer its temperature undisturbed to the temperature sensor and eliminate the influence of an interfering heat source
b) Incorrect installation – insertion length too
short
– poor heat coupling
– too high heat loss e.g. through exten-sion tube
– Insertion length of thermal element should be at least + 5 x d (liquids) up to 20 x d (gases)
(d = thermowell outside diameter) – Assure good heat contact, especially for
surface measurements, by using appro-priate contact surfaces and/or heat con-ductive materials (e.g. heat concon-ductive paste, grind surface)
– Reduce effect by suitable insulation c) Thermowell too thick – Use the smallest technically capable
thermowell; response time is proportional to the first power of the cross section or volume of the tempera-tur sensor, dependent on the heat trans-fer coefficient and air gaps in the as-sembly. Fill the latter with contact materials (oil, grease if possible) d) Deposits on the
thermowell (it has a thermally insulating effect)
– During inspection, remove
– If possible, select a different thermowell, or another installation location
Continuation: Error Table for Thermocouples and Resistance Thermometers
Error Probable or
possible cause
Corrective measures
Break in the temperature sensor
a) Vibrations – Stronger springs for measuring inset – Shorten insertion length
– Move measuring location (if possible) – Specially designed measuring inset
and thermowell
b) Thermal shock – Select a temperature change resistant sensor design
Very corroded, abraded or eroded thermowell
– Incorrect thermowell material selected
– Analyze defective thermowell and select a more suitable material;
provide supplementary surface protection (e.g. armoring or eccentrically drilled thermowell, impact rod)
Error Table Specifically for Thermocouples
– Instrument with a low input or internal resistance, high lead resistance
– Adjust leads
– Select an instrument with a higher input resistance
– Reference junction temperature or reference junction simulation must be maintained constant resistance (acts as a shunt path, decreases
– Exchange thermocouple measuring inset, then seal against moisture Large deviations of
– Check thermocouple and leads, exchange if necessary cables or their polarity is reversed
– Check if the correct compensating cable has the correct polarity – If a compensating cable is used:
Temperature of connection terminals max. 200 °C [392 °F).
Same temperature of connection terminals at > 100 °C [212 °F]
Indication changes over the course of time
– Chemical effects on the thermocouple especially at higher temperatures
– Exchange defective thermometer, possibly by a suitable thermocouple (e.g. Green rot in Type K
→replace with Type N)
– The measuring location wanders with the “healthy“ material into cooler regions, possibly insert thermocouple deeper, install air purge (O2-addition) – Thermal aging of the
thermocouple
– Select larger wire size in order to slow down the aging process
– Generally an aged thermocouple indicates lower temperatures than a new one.
– Check critical measuring locations regularly
– Regular recalibrations
Continuation: Error Table Specifically for Thermocouples
Error Probable or
possible cause
Corrective measures
Indicating instrument shows room tempera-ture (reference junction in instrument)
– Lead break – Check continuity
Negative temperature indication
– Incorrect polarity at thermocouple
– Reverse thermocouple polarity
Indication in error by 20...25 °C (68...77 °F)
– Thermocouple Type L linearized as Type J or reverse
– Correct linearization
Indication even though temperature sensor disconnected
– Pick up on the compen-sation cable due to electromagnetic noise – Parasitic galvanic
voltage (adjacent meter location) due to moisture in the compensation cable
– Dry compensation cable
Error Table Specifically for Resistance Thermometers
– Install larger wire size cables – Compensate leads
– Use sensor head transmitters – Convert to 3- or 4-wire circuits – Reduce connection lead lengths – Self heating by
measur-ing current too high
– Use a smaller measuring current (recommended 1 mA) are subjected to large a temperature change
– Convert to 3-wire circuit, which essentially eliminates the ambient temperature effects
– Convert to a 4-wire circuit
(Connection lead resistance effects resistance, acts as a shunt path for the measured signal
– Rins approx. 0.1 MΩ in parallel with 100 Ω gives an error of the same magnitude as Tolerance Class B Recommended:
Rins at 20 °C ( 68 °F): ≥ 100 MΩ Rins at 500 °C (930 °F): ≥ 2 MΩ (Minimum requirements per EN 60751) – Exchange defective thermometer Deviations of the
tem-perature indication from the values in the tables (parasitic and galvanic EMF’s)
– Poor lead material, contamination, over the course of time
– Thermal aging (Drift of the measuring resistor)
– Select suitable high temperature design – Recalibrate regularly
– Exchange if necessary