A 10K space sensor is wired perfectly, but the software point has 10K Type II selected and the sensor is actually a Type III part. What does the point look like on the graphic?

It reads a live, plausible temperature that's a few degrees off — worst far from 77 °F (25.0 °C), exactly right at it — and it never alarms.. Both curves pass through 10 kΩ at 77 °F (25.0 °C) — that shared point is what the "10K" name promises — and they diverge everywhere else. A Type III part on a Type II config reads about 3 °F (1.7 °C) high when the truth is near 40 °F (4.4 °C), less near the crossover, more further from it. Nothing rails, nothing alarms, the trend looks alive — which is why this failure survives for years. Contrast the loud faults: an open pins the cold rail, a short pins hot, and anyone can see those. Nobody hunts a 3° lie.

You disconnect a 10K Type II duct sensor and your meter reads 20.0 kΩ. Per the Type II curve table, about what temperature is that sensor sitting at? Enter the answer in °F.

50 °F. An NTC's resistance rises as it cools, so 20 kΩ — double the 10 kΩ defining point at 77 °F (25.0 °C) — means well below 77. On the Type II curve it lands at 50 °F (10.0 °C). The direction check is worth automating in your head: above 10 kΩ always means below 77 °F on a 10K sensor, and vice versa. The thermistor-calculator's R/T table is this exact lookup, both directions, for every common curve.

A chilled-water plant needs supply and return pipe sensors for an energy (BTU) measurement, where the answer rides on a small ΔT — maybe 10 °F (5.6 °C) — between the two. Which sensor choice fits best?

A matched pair of platinum RTDs — pair-matching and long-term stability keep a small ΔT honest for years.. Energy is flow × ΔT, so a single degree of disagreement between the pair is a 10% error on a 10 °F (5.6 °C) ΔT — forever, on a number that may feed a bill. That's the RTD's home turf: matched platinum pairs hold their difference accurately, and platinum drifts far less over years than a thermistor. Thermistors own space and duct sensing, where their big signal and low cost win and a degree of absolute error is invisible. Same-curve agreement (d) doesn't help — two sensors can share a curve and still each drift.

A Pt100 was designed as a 3-wire landing on a 150 ft (46 m) run, but the third conductor got damaged and the input is temporarily reconfigured as 2-wire. Nothing else changed. What happens to the reading?

It reads several degrees high, steadily: with the compensation gone, the copper leads' resistance counts as element resistance, and on a Pt100 every 0.22 Ω reads as another 1 °F.. The third wire exists so the input can measure a lead-only loop and subtract it. Take it away and the subtraction stops: 150 ft of 18 AWG pair is about 1.9 Ω of copper in series, which a 2-wire Pt100 input reads as roughly 9 °F (5 °C) of extra temperature — constant, plausible, and wrong. Resistance in series always adds, so the error is always high, never low or noisy. This is also why the fault is invisible on a thermistor: the same 1.9 Ω against a 10K element's ≈240 Ω/°F (≈440 Ω/°C) is thousandths of a degree.

A 2-wire Pt100 run adds 1.3 Ω of total lead resistance (both conductors). A Pt100 moves about 0.22 Ω per °F. About how many °F high does the point read? Enter the answer in °F.

6 °F. The input can't tell lead ohms from element ohms, so the whole 1.3 Ω reads as temperature: 1.3 / 0.22 ≈ 6 °F high (in SI: 1.3 Ω ÷ 0.39 Ω/°C ≈ 3.3 °C). That's the entire lead-resistance problem in one division — small ohms over a small ohms-per-degree is a big error. The fixes, in order of bluntness: a 3-wire landing that cancels the leads, a 4-wire landing that removes them from the measurement entirely, or a Pt1000, whose tenfold base resistance shrinks the same 1.3 Ω to about 0.6 °F (0.3 °C).

A zone sensor reads 80 °F (26.7 °C) but the room feels cold, and you want a verdict before anyone swaps parts. Which sequence is the verification workflow?

Disconnect the sensor at the input → meter the ohms → convert ohms to temperature with the R/T table for the sensor's type → compare against a reference thermometer at the sensor.. Order matters. Disconnect first: landed, the meter reads the sensor in parallel with the input's own circuit, and the number lies. The ohms are the sensor's honest testimony; the table (the thermistor-calculator carries it for every common curve) turns them into a temperature; the reference thermometer closes the loop. Swapping first (b) can replace a good sensor to fix a config problem, and an offset (d) that papers over a curve mismatch fixes one temperature while the error keeps moving everywhere else.

An OAT point reads −38 °F (−38.9 °C) on a mild fall morning — the classic open-circuit signature. At the controller you disconnect the pair and the meter reads OL (open). What's the fastest way to split "bad sensor" from "bad run"?

Meter again across the sensor at the sensor head: plausible ohms there means the run or a termination is broken; OL at the head convicts the sensor.. The display signature told you what (open circuit → cold rail); metering at both ends tells you where. The circuit is sensor + run + terminations in series, so one measurement at each end splits it: healthy ohms at the head with OL back at the panel means the break is between them — and loose terminations and damaged splices fail far more often than elements do. A curve change (c) can't fix an open — no curve maps infinite ohms to a mild morning.

An outdoor-air sensor was replaced a while back. Since then the OAT point tracks the weather smoothly — but it runs a few degrees high in cold weather and dead-on in mild weather. Here's the point config next to the new part's label. What's wrong? point: OA-T (universal input, resistance mode) sensor label: 10K NTC thermistor — curve: Type III point config: sensor type = 10K Type II observed: 35.2 °F on a morning the airport reports 32 °F; read dead-on through the mild weeks after the swap

The configured curve doesn't match the part: Type II is selected for a Type III sensor. Both read 10 kΩ at 77 °F, so mild weather looked perfect — but near freezing the mismatch reads ≈3 °F high. Select the matching curve.. The mild-weather clue is the tell: near the 77 °F (25.0 °C) crossover the two curves agree, so the mismatch hid until the weather left the crossover — the signature of a curve problem, not a hardware one. The offset (d) is the tempting wrong fix: the error varies with temperature (zero at 77 °F, ≈3 °F near freezing, ≈5 °F at 0 °F / −17.8 °C), so a constant offset just relocates the lie. And (c) is the RTD failure mode — a 10K thermistor shrugs off a couple of ohms of leads. Match the config to the label; the wire can't do it for you.

You meter a known-working space sensor at about 4.7 kΩ with the room near 77 °F (25.0 °C). The job standard says "10K sensors everywhere." What's the most likely story?

It's a shunted vendor convention — a 10K element with a parallel resistor (the JCI 10K + 8.7 kΩ pattern), which reads roughly half of 10K at room temperature. The name is not the curve.. Several vendors flatten the 10K curve with a parallel shunt — the JCI convention (8.7 kΩ) reads ≈4.7 kΩ at 77 °F (25.0 °C), the TAC Type 5 (11 kΩ) ≈5.2 kΩ — and both still get called "10K sensors." A real short reads near 0 Ω, not 4.7 kΩ; leads add ohms (and only a couple); and a Pt1000 sits near 1.1 kΩ at that temperature. The action item: the input must have the vendor's shunted curve selected, not bare Type II or III — and the calculator's Identify mode ranks every curve when the label is gone.

Hold a 10K thermistor in your fist and its resistance falls as it warms; do the same with an RTD and its resistance rises.

True. True — and it's a free identification trick. NTC means negative temperature coefficient: resistance falls as temperature rises, steeply. RTDs run the other way — platinum and Balco both rise with temperature, and nearly linearly. With an unlabeled sensor on the bench, warming it by hand and watching which way the ohms walk sorts the family in ten seconds; the resistance neighborhood (10-ish kΩ vs ≈100 / ≈1,000 Ω at room temperature) then narrows the type before the R/T table settles it.

Practice · Temperature Sensors

Temperature Sensors Quiz Signals

Read the Temperature Sensors lesson first, then test yourself — ten questions on the curve behind every ohm: Type II vs Type III and the wrong-pick symptom, the shunted vendor conventions, RTDs and the lead resistance that reads as temperature, and the meter-and-table workflow that decides sensor vs input vs configuration. Each reveal links back to the section it came from. For the wire-side fundamentals — landings, shields, loop power — the Field Wiring & Sensors drill is the companion set.

← All practice