Testing a Temperature Sensor in the Field: Multimeter Checks for RTDs, Thermocouples, and Thermistors

When a temperature reading looks wrong, the first reaction is often to blame the sensor and swap it. Sometimes that's right. Often it isn't, and the replacement shows the same fault because the real problem was a loose terminal, a wrong setting, or a damaged cable. A basic multimeter, used correctly, can tell you in a few minutes whether the sensor itself is healthy or whether you should be looking elsewhere.

This guide gives maintenance teams a practical set of checks for the three most common sensor types, with the reference values you need to interpret the results.

Before You Start

A few ground rules keep the tests safe and meaningful.

• Isolate the sensor. Switch off the equipment and disconnect the sensor from the controller or transmitter before measuring. Testing a sensor while it's still connected to a live instrument gives false results and can damage the meter or the input.
• Test at the sensor first, then at the cable end. Measure at the probe head, then at the far end of the cable. If the reading is good at the head and bad at the far end, the cable is the problem.
• Know the temperature at the sensor. Resistance and voltage values only mean something when compared with the actual temperature of the sensing tip. Use a reference thermometer, or test at a known point such as an ice bath.
• Use a decent meter. A meter with good low-resistance accuracy matters for RTDs, where one ohm is a couple of degrees. Hold probe contact steady and keep fingers off the contact points.
• Mind self-heating. Meters apply a small current when measuring resistance. For tiny or low-mass sensors, take a quick reading and don't leave the meter connected for long.

Testing RTDs (PT100 and PT1000)

A platinum RTD is a resistor whose value rises almost linearly with temperature. The standard PT100 reads 100.00 ohms at 0°C and changes by roughly 0.385 ohms per degree. A PT1000 follows the same curve at ten times the resistance.

Reference values for a PT100 (approximate):

For a PT1000, multiply each value by ten (for example, about 1,000 ohms at 0°C and 1,385 ohms at 100°C).

Steps:

1. Set the meter to resistance and measure across the two leads of the same color for a 3-wire RTD, or the two element leads for a 2-wire RTD. For a 3-wire sensor, the two leads of one color should read close to zero ohms with each other, and the odd-colored lead should read the element resistance to each of them.
2. Compare the reading with the table at the sensor's actual temperature.
3. Check for an open circuit (meter shows "OL" or infinite), which means a broken element or lead.
4. Check for a short (reading near zero or far too low), which suggests damaged insulation or a failed element.
5. Check insulation to ground. Measure between a lead and the sheath or housing; it should read open (very high resistance). A low reading points to moisture ingress or insulation damage. For a rigorous test, an insulation resistance tester is appropriate, used according to the manufacturer's recommended test voltage, since a standard multimeter only gives a rough indication.

Interpreting results: a reading within a couple of ohms of the expected value at the known temperature is healthy. A steady offset that matches several degrees suggests drift or lead resistance, and an erratic or jumping reading points to a loose connection or damaged element.

Testing Thermocouples

A thermocouple produces a small voltage rather than a resistance, and its two wires have low resistance. Two simple tests help.

Continuity and Resistance Check

• Disconnect the thermocouple and measure resistance across its two leads.
• A healthy thermocouple reads a low resistance, typically a few ohms for short, thick wires, rising with length and for thin wires. The exact value depends on wire gauge and length.
• An open circuit means a broken junction or wire. A reading far higher than expected can mean a partially broken or corroded junction.

Voltage Check

• Set the meter to millivolts DC and connect across the thermocouple leads.
• Apply heat to the tip, with a lighter, warm water, or a hand, and watch the voltage change. A healthy thermocouple produces a rising voltage as the tip warms.
• Compare the measured voltage against the thermocouple's published table. Remember that the table assumes a 0°C reference junction, so in practice you add the voltage equivalent of the meter-end (room) temperature, as covered in our article on cold junction compensation.

Reference values for a Type K thermocouple (approximate, 0°C reference):

If the tip is warm, the voltage rises. If polarity is reversed on the meter, the voltage reads negative and the display will move the wrong way when heated, a clue that the connection or meter lead orientation needs checking.

Common faults to spot:

• Open circuit: broken wire or junction
• No voltage change when heated: open or failed junction
• Voltage far from table at a known temperature: contamination, drift, or wrong type
• Erratic readings: loose connection, damaged insulation, or intermittent junction
• Low insulation resistance to the sheath: moisture or damaged insulation, particularly in mineral insulated probes where moisture at an open end is a known risk

Testing NTC Thermistors

An NTC thermistor's resistance falls as temperature rises, and the relationship is strongly nonlinear.

Steps:

1. Measure resistance across the two leads with the meter on a suitable range.
2. Compare against the thermistor's nominal value at 25°C. For example, a 10 kΩ NTC should read close to 10 kΩ at 25°C, within its stated tolerance.
3. Warm the sensor in your hand or in warm water and confirm that resistance falls. Cool it and confirm that it rises.
4. For a more precise check, look up the expected resistance at the actual temperature in the datasheet's resistance-temperature table, then compare.

Rough guide for a typical 10 kΩ NTC with Beta of about 3950:

Exact values vary with the specific part, so use the manufacturer's datasheet for any serious comparison. A tolerance of ±1%, ±3%, or ±5% at 25°C accounts for part-to-part variation, as explained in our article on tolerance grades.

Common faults:

• Open circuit: broken lead or element
• Very low resistance: short or contamination
• No change with temperature: failed element or damaged connection
• Reading wanders when you touch it: poor contact or a partly broken lead

Is It the Sensor, the Cable, or the Instrument?

A good diagnostic sequence narrows down the cause quickly.

1. Test the sensor at the probe head. If the resistance or voltage matches the expected value, the sensor element is probably fine.
2. Test at the far end of the cable. If the value matches at the head but not here, the cable or a junction is faulty.
3. Inspect terminals and connections. Loose, corroded, or wet terminals cause many intermittent faults.
4. Check the wire type for thermocouples. Using wrong extension wire creates hidden errors, as covered in our guide to thermocouple extension wire and compensating cable.
5. Check the instrument configuration. Confirm the input type (PT100, PT1000, Type K, and so on), wiring mode, and scaling match the sensor.
6. Substitute a known value. Connect a precision resistor, a decade box, or a thermocouple simulator in place of the sensor. If the instrument displays the correct temperature, the instrument is fine and the sensor or wiring is at fault.
7. If a transmitter is installed, check its output and configuration separately, as described in our article on head-mounted temperature transmitters and 4–20 mA signals.

A Simple Ice-Bath Check

For a quick accuracy check on any of these sensors, place the tip in a mixture of crushed ice and water, stirred, and allowed to settle. This holds close to 0°C. A healthy PT100 should read close to 100 ohms, a thermocouple should read close to 0 mV (referenced to the meter-end temperature), and a 10 kΩ NTC should read close to its 0°C datasheet value. It's a low-cost way to catch a sensor that's drifted or been damaged.

Limits of a Multimeter

A multimeter confirms basic health but isn't a calibration tool. It can't tell you a sensor is accurate to a fraction of a degree. For critical measurements, periodic calibration against a reference standard is the right approach, and multimeter checks complement rather than replace it.

Keeping Plants Running With Quick Checks

A few minutes with a meter can save an unnecessary sensor replacement and, more importantly, point to the real fault faster. Zentriad supplies NTC thermistor, thermocouple, and RTD sensors for industrial applications, and our team can provide datasheets with resistance-temperature or voltage-temperature tables to support your maintenance checks.

Need datasheets or replacement sensors for your equipment? Contact Zentriad's team with your sensor type and application details.