The resistivity of platinum is the inherent opposition the metal offers to electrical current flow, measured at approximately 1.06 × 10⁻⁷ Ω·m at 20°C, which increases in a highly predictable, linear fashion as temperature rises. Unlike copper, where we fight to keep resistivity low for power transmission, platinum’s specific resistivity profile is the exact reason we use it for precision temperature measurement. When you buy a PT100 or PT1000 sensor, you aren't just buying a resistor; you are buying a highly calibrated chunk of platinum wire or thin-film deposition whose resistance shifts by exactly 0.385 Ω per degree Celsius.

What Platinum's Resistivity Actually Changes in a Circuit

In a standard DC power circuit, resistivity dictates voltage drop and heat generation. But in instrumentation circuits, the resistivity of platinum changes the entire topology of the measurement system. Because the absolute change in resistance is so small—fractions of an ohm per degree—you cannot simply drop a platinum sensor into a basic 5V voltage divider and read it with an Arduino's 10-bit ADC. The noise floor and ADC quantization steps will swallow your signal.

Instead, platinum's resistivity forces us to adopt specific circuit architectures: a constant-current excitation source paired with a high-impedance differential amplifier (or a dedicated sigma-delta ADC like the MAX31865). It also dictates that we abandon standard 2-wire multimeter measurements in favor of 3-wire or 4-wire Kelvin sensing topologies to strip out the resistance of the copper lead wires connecting the sensor to the board.

The Math: A Worked Numeric Example

Let’s calculate the exact resistance of a standard PT100 sensor (alpha = 0.00385) at 150°C. Most hobbyist tutorials will hand you the linear approximation formula:

R(T) = R₀ × [1 + (α × T)]

Plugging in our numbers: R(150) = 100 × [1 + (0.00385 × 150)] = 157.75 Ω.

Warning: The Linear Trap
If you use the linear approximation at 150°C, your math says 157.75 Ω. But according to the IEC 60751 standard and the NIST ITS-90 reference tables, platinum's resistivity curve flattens slightly at higher temperatures. The real Callendar-Van Dusen (CVD) equation uses coefficients A (3.9083 × 10⁻³) and B (-5.775 × 10⁻⁷). Running the full CVD math yields 157.325 Ω. That 0.425 Ω difference translates to a 1.1°C measurement error. If you are tuning a PID controller for an industrial oven, relying on linear math will cause steady-state oscillation.

Where You Meet This in Practice

You will encounter platinum resistivity sensors anywhere temperature stability is non-negotiable and thermistors are too non-linear. Common bench and jobsite encounters include:

  • High-End 3D Printers: Upgraded hotends (like the E3D Revo or Prusa MK3S+) use PT1000 sensors to push past the 300°C limit where standard NTC thermistors become dangerously inaccurate.
  • HVAC Chillers and Boilers: Immersion-style RTDs monitor supply and return water temperatures to calculate delta-T for energy efficiency logging.
  • Laboratory Incubators: Class A (or Class AA) wire-wound platinum sensors maintain biological samples at exactly 37.0°C with tolerances tighter than ±0.1°C.
  • Automotive Exhaust Testing: Platinum thin-film sensors survive the harsh, high-vibration environment of dyno testing where thermocouples might suffer from EMI noise.

Bench War Story: When 2-Wire Wiring Ruined a $500 Sensor Setup

A few years ago, I was automating a composite curing oven for a local fabrication shop. The spec called for tight thermal profiling, so I installed a high-accuracy Class A PT100 sensor. The oven controller was mounted 50 feet away, so I ran 50 feet of standard 22 AWG copper wire to connect it, using a cheap 2-wire-to-Analog transmitter module to save on wiring time.

The Numbers: 22 AWG copper wire has a resistance of roughly 16.14 mΩ per foot at room temperature. A 50-foot run out and a 50-foot run back equals 100 feet of total loop length. 100 ft × 0.01614 Ω/ft = 1.614 Ω of lead resistance. A standard PT100 shifts by about 0.385 Ω per °C.

The Outcome: The PLC read the total loop resistance, meaning it saw 1.614 Ω of phantom resistance added to the sensor. Divided by 0.385 Ω/°C, the system reported the oven was 4.2°C hotter than reality. The controller backed off the heating elements early, and the first batch of carbon fiber parts came out under-cured and structurally compromised.

What Went Wrong: I treated a platinum resistivity measurement like a standard voltage drop. To fix it, I ripped out the 2-wire transmitter and moved to a 3-wire Kelvin topology. Here is how you wire a 3-wire RTD correctly to cancel out that exact lead resistance:

  1. Identify the wires: A 3-wire PT100 has one red wire and two white (or identical colored) wires. The two identical wires are connected to the same internal platinum terminal.
  2. Connect the excitation: Wire the red lead to the ADC's positive excitation current source (IOUT+).
  3. Connect the sense lines: Wire the first white lead to the positive voltage sense terminal (VIN+), and the second white lead to the negative voltage sense terminal (VIN-).
  4. Complete the circuit: Wire the red lead's voltage sense (or a jumper from IOUT+ to VIN+) to complete the measurement loop, allowing the ADC to measure the voltage drop across the sensor while mathematically subtracting the voltage drop of the white lead wires.
  5. Verify: Measure the resistance across the two white wires at the PLC terminal block with a multimeter. It should read near 0 Ω (just the wire). If it reads open, you have a broken sensor lead.

Common Confusions: Resistivity vs. Resistance and RTDs vs. Thermocouples

When ordering parts or debugging a schematic, people frequently mix up three distinct concepts:

Resistivity vs. Resistance: Resistivity (ρ) is a fundamental material property of platinum, measured in ohm-meters. It doesn't change based on the shape of the metal. Resistance (R) is the actual ohmic value of your specific PT100 component (100 Ω at 0°C), which is determined by the resistivity of platinum multiplied by the length of the wire and divided by its cross-sectional area. You measure resistance; you exploit resistivity.

RTDs vs. Thermocouples: Both measure temperature, but the physics are entirely different. An RTD relies on the resistivity of a single metal (platinum) changing with heat. A thermocouple relies on the Seebeck effect, where a junction of two entirely different metals (like Type K's chromel and alumel) generates a tiny millivolt signal due to a thermal gradient. If your schematic specifies an RTD amplifier (like a MAX31865) and you wire a thermocouple to it, you will read garbage data. For deep-dive sensor selection, the Omega Engineering RTD learning center provides excellent cross-reference charts.

FAQ: Platinum Sensor Specs and Selection

Should I choose a PT100 or a PT1000 for my microcontroller project?

Choose a PT1000 if you are wiring directly to a microcontroller ADC or using long cable runs. Because a PT1000 has 1000 Ω base resistance, it shifts by 3.85 Ω/°C. This makes the signal 10x larger than a PT100, meaning lead wire resistance and ADC noise have 1/10th the impact on your final temperature reading. Choose a PT100 if you are integrating into legacy industrial PLCs (like Allen-Bradley or Siemens) that expect standard 4-20mA RTD transmitter loops.

Does thin-film platinum perform as well as wire-wound?

For 95% of hobbyist and commercial applications under 400°C, thin-film sensors (which look like tiny surface-mount chips encased in ceramic) are perfectly adequate and cost under $5. Wire-wound sensors (platinum wire coiled inside a ceramic mandrel) are required for extreme precision (Class AA), cryogenic temperatures, or environments exceeding 600°C where thin-film deposition can physically degrade and drift.

Why does my PT100 read 200°C when the water is clearly boiling at 100°C?

You likely have a short circuit in your 3-wire or 4-wire cable. If the excitation current wire shorts to one of the sense wires near the connector, the ADC measures the voltage drop across the short rather than the platinum element, resulting in a massive, pegged-high reading. Disconnect the sensor and check for continuity between the red and white wires at the probe tip; it should be exactly 100 Ω (at 0°C) or roughly 138 Ω (at room temp), not 0 Ω.