a resistance temperature detector (RTD)
/ R-T-D; Pt100 = P-T-one-hundred /
An RTD is the calm, dependable cousin of the thermistor. It also measures temperature by changing its resistance, but instead of a sensitive, wildly curved ceramic, it uses a coil or film of pure metal — almost always platinum. Metals change resistance with temperature gently and very predictably, like a well-mannered ruler that expands the same amount every degree. You give up some sensitivity in exchange for accuracy, stability, and a response that is nearly a straight line.
The most common RTD is the Pt100: pure platinum with exactly 100.00 Ω at 0 °C, rising by about 0.385 Ω for every degree Celsius (a temperature coefficient near 0.00385 per °C). So at 100 °C a Pt100 reads about 138.5 Ω. There is also the Pt1000 (1000 Ω at 0 °C) which is ten times more sensitive and easier to wire. Because the resistance change is small, two practical problems appear: the resistance of the long copper wires to the sensor adds into your reading, and the small sense current can self-heat the element. The classic cures are 3-wire and 4-wire connections, which let the circuit subtract out the lead resistance so it does not masquerade as temperature.
Why this matters: when you need accuracy and long-term stability — industrial process control, lab references, food and pharma — the RTD is the workhorse. It is more linear and far more stable over years than a thermistor, and a Pt100 is traceable to international standards. The honest tradeoffs: it is more expensive, less sensitive, slower to respond because of its mass, and it genuinely needs 3- or 4-wire wiring and a small, well-controlled excitation current to reach its rated accuracy.
A Pt100 fed with a steady 1 mA current sits at 100.0 Ω at 0 °C, giving 100 mV across it. Warm it to 50 °C and it rises to about 119.4 Ω, so the voltage climbs to about 119.4 mV — a clean, almost-linear 0.385 mV per degree that an instrumentation amplifier can boost and read.
Platinum changes resistance gently and predictably — accuracy over sensitivity.
Wire resistance is the classic RTD trap: in a simple 2-wire hookup the leads add directly to the 100 Ω, faking degrees of error. Use 3-wire or 4-wire sensing for any real accuracy, and keep the excitation current small to avoid self-heating.