An RTD (Resistance Temperature Detector) is a precision temperature sensor that calculates heat by measuring the predictable change in electrical resistance of a pure metal—usually platinum—as its temperature changes. Unlike sensors that generate their own voltage, an RTD is a passive component. In a real circuit, it changes a thermal variable into a highly linear resistance variable, which means your measurement circuit must supply a known excitation current (usually 1mA or less to prevent self-heating) to read the resulting voltage drop across the sensor.
The Core Physics: How Platinum Tracks Heat
The most common RTD element is made of platinum wire or thin-film platinum deposited on a ceramic substrate. Platinum is chosen because its resistance-to-temperature relationship is exceptionally linear and repeatable over a massive range, from -200°C to +850°C. The defining metric for an RTD is its alpha coefficient ($\alpha$), which describes the average change in resistance per degree Celsius between 0°C and 100°C.
For high-precision laboratory work, the relationship isn't perfectly linear across extreme temperatures. Engineers use the Callendar-Van Dusen equation to calculate exact resistance at sub-zero temperatures, accounting for the slight curve in the platinum response. But for 95% of bench and industrial work, the linear approximation holds up beautifully.
PT100 vs. PT1000: The Numbers That Matter
When you ask 'what is an RTD', you will immediately run into two main variants: the PT100 and the PT1000. The 'PT' stands for Platinum, and the number is the nominal resistance at 0°C.
| Feature | PT100 | PT1000 | 10k NTC Thermistor (For Comparison) |
|---|---|---|---|
| Resistance at 0°C | 100 Ω | 1000 Ω | ~27,219 Ω |
| Resistance at 25°C | 109.73 Ω | 1097.3 Ω | 10,000 Ω |
| Sensitivity (approx at 25°C) | 0.385 Ω/°C | 3.85 Ω/°C | -400 Ω/°C (Non-linear) |
| Linearity | Excellent | Excellent | Poor (Exponential) |
| Lead Wire Impact | High (Requires 3/4-wire) | Low (2-wire often OK) | Negligible |
Worked Numeric Example: Calculating PT100 Resistance
Let's calculate the expected resistance of a standard PT100 sensor sitting in a water bath at 75°C. Using the linear approximation formula $R = R_0(1 + \alpha T)$:
- $R_0$ = 100 Ω
- $\alpha$ = 0.00385
- $T$ = 75°C
$R = 100 \times (1 + (0.00385 \times 75))$
$R = 100 \times (1 + 0.28875)$
$R = 100 \times 1.28875 = $ 128.875 Ω
If your multimeter reads 128.88 Ω across the sensor leads, your bath is exactly 75°C.
Where You Meet This in Practice
You won't often wire a raw PT100 directly to a microcontroller like an ESP32 or Arduino without an amplifier (like the MAX31865 breakout board). In the field, RTDs are the backbone of heavy industry and precision HVAC:
- Injection Molding & Extruders: Platinum RTDs embedded in heater bands monitor barrel temperatures. The high linearity ensures the PID controller doesn't overshoot and burn the plastic.
- HVAC Chillers and Boilers: Immersion-style RTDs measure supply and return water temperatures to calculate thermal energy transfer (BTUs).
- Food & Pharma Cold Chain: Walk-in freezers and autoclaves rely on RTDs because they do not suffer from the 'drift' that thermocouples experience over months of continuous use.
In these installations, the raw RTD signal is usually converted locally by a temperature transmitter into a 4-20mA current loop or a digital signal (HART/Modbus) to survive the noisy electrical environment of a factory floor.
Scenario Walkthrough: The 3-Wire Compensation Failure
To understand why RTD wiring topology matters, let's look at a real-world installation failure.
The Setup: An installer is wiring a PT100 sensor inside a commercial walk-in freezer to a PLC located 50 feet away in the mechanical room. To save time and money, they use a standard 2-wire configuration with 24 AWG copper control wire. The freezer is set to maintain 0°C (32°F).
The Numbers: According to the NEC Chapter 9 tables, uncoated 24 AWG copper wire has a resistance of roughly 25.67 ohms per 1,000 feet at 75°C.
The run is 50 feet out, and 50 feet back, creating a 100-foot loop.
Lead resistance = $100 \text{ ft} \times (25.67 / 1000) = $ 2.567 Ω.
The Outcome: At 0°C, the PT100 sensor itself is exactly 100.00 Ω. However, the PLC measures the total circuit resistance: $100.00 Ω + 2.567 Ω = 102.567 Ω$.
Using our alpha coefficient (0.00385 Ω/°C), that extra 2.567 Ω looks like a temperature increase of roughly 6.6°C. The PLC thinks the freezer is at 6.6°C (44°F). Satisfied that the space is 'cold enough', the PLC shuts off the compressor. The food spoils.
What Went Wrong: The installer failed to use a 3-wire or 4-wire configuration. Think of the wire as a toll road; the longer the road, the more tolls (resistance) you pay before reaching the sensor. A 3-wire setup uses a third wire to measure the 'toll' of the lead wires and mathematically subtracts it via a Wheatstone bridge circuit inside the transmitter, isolating the true sensor resistance.
What People Commonly Confuse RTDs With
When diagnosing a temperature circuit, mixing up sensor types will instantly fry your input card or give you wildly inverted readings.
- Thermocouples (Type K, J, T): Thermocouples generate their own tiny millivolt signal via the Seebeck effect at the junction of two dissimilar metals. RTDs do not generate voltage; they require an external excitation current. If you wire an RTD to a thermocouple input card, the card will read a flatline or throw an open-circuit fault.
- Thermistors (NTC/PTC): Thermistors are made of sintered metal oxides, not pure metals. They have massive sensitivity (high resistance change per degree) but terrible linearity and a very narrow operating range (usually -40°C to 150°C). An RTD is the choice when you need wide-range accuracy; a thermistor is the choice for cheap, narrow-range consumer appliances like coffee makers.
FAQ: RTD Bench Testing and Wiring
How do I test a PT100 with a standard multimeter?
Set your multimeter to the lowest ohms range (usually 200Ω or 400Ω). Disconnect the sensor from the transmitter to avoid reading parallel circuit resistance. Place your probes across the sensor leads. At room temperature (approx. 22°C / 72°F), a healthy PT100 should read between 108.4 Ω and 108.6 Ω. If it reads 'OL' (open) or near 0.0 Ω (shorted), the element or internal wiring has failed.
What is the difference between 2-wire, 3-wire, and 4-wire RTDs?
- 2-Wire: Cheapest, but includes lead resistance in the measurement. Only acceptable for short runs or when using high-resistance sensors like PT1000.
- 3-Wire: The industrial standard. Uses a third wire to compensate for lead resistance, assuming all three wires are the same length and gauge.
- 4-Wire: The laboratory standard. Uses two wires to supply current and two separate wires to measure voltage, completely eliminating lead resistance from the equation.
Can I use an RTD directly with an Arduino or ESP32?
No. Microcontrollers can only read voltage via their ADC (Analog-to-Digital Converter), and they cannot supply the highly stable, low-noise constant current required to measure an RTD accurately. You must use a dedicated RTD-to-digital converter IC, such as the MAX31865, which handles the excitation current, analog-to-digital conversion, and fault detection, outputting the data via SPI.
