How to Test an Oven Temperature Sensor and Its Harness

Why this matters

An oven that runs cool, runs hot, or throws an open or shorted sensor code gets a new sensor more often than it needs one. The sensor is cheap and easy, so it is ordered on the complaint. But the control does not read the sensor; it reads the sensor plus every connector and inch of wire between the probe and the board, and the part of that path that fails most is the connector near the back of the oven, where heat and the range being pushed against the wall do their work. Reading the same circuit at two points, then wiggling what sits between them, tells you which half is bad. A sensor swapped in front of a bad connector reads fine in the shop and fails in the kitchen.

Power, the range, and heat first

De-energize at the range. A freestanding electric or dual-fuel range runs on 240 V: unplug it or turn off its breaker, and prove absence of voltage at the terminal block or the board's line terminals with a tester proved on a known live source before and after, per NFPA 70E Article 120 (120.5 in the 2021 edition; check the current edition for its number) and 29 CFR 1910.333(b)(2). A gas range still has a 120 V cord for the control and igniters; unplug it the same way. Every resistance reading on this card is taken power off.

Pulling the range out. Slide it straight out on a protected floor and disengage the anti-tip bracket as you go. On a gas range, the flexible connector bends as the range moves: watch it does not kink, and after the range goes back, leak-test the connector joints you may have disturbed with a noncorrosive solution, never a flame, per 'Gas Range Connection and Leak Check Standard'. The anti-tip bracket must be re-engaged before you leave; a range without it can tip forward onto a child standing on the open door.

Heat. If the oven ran recently, the probe and the cavity walls burn on contact. Let it cool with the door open before reaching in, and before taking a room-temperature reading, which is meaningless on a warm probe.

What each reading includes

 BOARD CONNECTOR (sensor pins)
        |
        |  harness, along the back of the range
        |  (pinch point: range pushed to the wall)
        |
 INLINE CONNECTOR at rear wall   <- wiggle zone
        |
        |  sensor pigtail, through the cavity wall
        |
 PROBE inside the oven cavity

 Reading A: at the inline connector, sensor side
           = probe + pigtail
 Reading B: at the board connector, unplugged
           = probe + pigtail + connector + harness

The sensor on most current ovens is a platinum resistance temperature detector (RTD): its resistance rises as it heats. 'NTC Thermistors and Oven RTD Sensors Explained' covers why, and which sensor types run the other way. For this card, two consequences matter.

Extra resistance reads as extra heat. Anything that adds ohms in series, a corroded terminal or a strand-broken conductor, tells the control the oven is hotter than it is, so the control shuts the element or the gas valve off early and the oven runs cool. On a platinum sensor of the common family that reads near 1,000 ohms at freezing, resistance rises roughly 2 ohms per degree F, so a connector adding 20 ohms costs about 10 F. That is how a sensor that tests perfect produces a real complaint.

Lost resistance reads as cold. A short across the leads, or a lead touching chassis where the control references it, reads low and makes the control overheat the oven or flag a shorted-sensor fault.

The readings, in order

1. Get the chart from the tech sheet. Sensor resistance against temperature, usually a table. Note the kitchen temperature with a thermometer away from the range. Many current ovens use a sensor reading somewhere around 1,080 to 1,100 ohms at room temperature, but that is typical only; the chart for that model governs, and some ovens use an entirely different sensor type. Without the chart you have no pass or fail, only a number.

2. Reading A: at the sensor's own connector. Unplug the inline connector at the rear wall and read the sensor side. This is the probe and its pigtail alone. Within a few ohms of the chart passes. OL is an open sensor; near zero is a shorted one. A good Reading A clears the sensor of a constant fault, not of an intermittent one, which step 4 hunts.

3. Reading B: from the board connector. Reconnect the inline connector, unplug the sensor connector at the board, and read the two sensor pins on the harness side. Now the reading includes every connection between board and probe. Reading B should equal Reading A plus a few ohms at most. A difference of tens of ohms is a bad connection in the path, and you have already measured how many degrees it is costing.

4. Wiggle while you watch Reading B. Leave the meter on the board connector and flex the harness along the back of the range, then the inline connector, then the pigtail where it enters the cavity wall. A steady reading passes. A jump, a drift, or a flicker to OL is the fault, and the point you were flexing is where it lives. Intermittent sensor codes that appear only on long bakes are usually this: the harness expands with heat and a cracked conductor separates.

5. Each lead to chassis. Meter on its highest resistance range, one lead on a sensor pin at the board connector, the other on bare chassis metal, then the other pin. Both must read OL. Any reading is a lead rubbed through to the cabinet, or moisture in the probe, and the probe goes if it is the probe.

6. Repair with parts rated for the location. A bad terminal at the rear wall gets the maker's repair pigtail or high-temperature terminals rated for that spot, crimped, never ordinary wire nuts or vinyl tape, which fail in the heat behind an oven. Re-route the harness clear of the edge or bracket that damaged it.

7. Recheck B, then restore. Board connector unplugged, Reading B should now sit within a few ohms of Reading A. Reconnect, restore power, and run the finish test.

Worked example: 25 F cool and two fault codes

A homeowner reported the oven baking slow for months, then twice throwing the open-sensor code during a long holiday roast. The tech sheet chart gave 1,085 ohms at 70 F and about 2 ohms per degree near room temperature. The kitchen read 72 F, so the expected reading was about 1,089 ohms.

• Range unplugged, absence of voltage proved at the terminal block, anti-tip bracket released and the range slid out.
• Reading A, at the inline connector on the sensor side: 1,088 ohms. The sensor matched the chart within 1 ohm.
• Reading B, at the board connector: 1,141 ohms. That is 53 ohms more than Reading A.
• At about 2 ohms per degree, 53 ohms is roughly 26 F the control believed the oven had that it did not, which matched "runs about 25 F cool."
• Wiggle test: steady while the harness along the back was flexed, then jumping between about 1,140 ohms and OL when the inline connector was squeezed.
• Each pin to chassis: OL.

Opened up, one terminal in the inline connector was darkened and loose on its crimp, sitting against the back panel where the range met the wall. The tech installed the maker's sensor repair pigtail with high-temperature terminals and routed the harness under its retaining clip, away from the panel edge.

Reading B after the repair: 1,090 ohms, 2 ohms above Reading A. Range pushed back, anti-tip bracket re-engaged and checked by tilting the range gently forward against it, power restored.

The sensor in the van stayed in the van. Had it gone in, the new sensor would have read 1,089 ohms on its own, the oven would have kept running about 25 F cool behind the same connector, and the next open-sensor code would have been a callback on a part already replaced.

Where the answer changes

Reading A is off the chart and B equals A. The sensor is the part. Replace it with the listed sensor, then still read B, because a heat-damaged pigtail sometimes means the connector took the same heat.

Both readings are good and the oven still runs a little off. If the oven is within about 15 F at the finish test and every reading matches, the remaining difference is calibration, and the control's offset procedure belongs to 'Oven Bake Element and Temperature Calibration'. Never use the offset to cover a harness fault: setting a 25 F offset on the oven in the example would have hidden the complaint until the connector failed open.

Readings good, no heat at all. The sensor is not stopping the oven. Go to the element or the igniter and valve side.

How to verify you got this right

Put an oven thermometer on the center of the middle rack. Set bake to 350 F, wait for the preheat signal, then wait another 20 minutes so the oven has cycled. Read the thermometer through the window at the top and bottom of two heat cycles and average them; the oven swings above and below set, and one reading taken at a peak or a trough misleads. A common shop default is an average within 10 F of the set temperature.

• Reading A and Reading B both on the ticket, with the chart value and the kitchen temperature beside them.
• Reading B within a few ohms of A, and steady under the wiggle test after the repair.
• Both leads OL to chassis.
• Thermometer average within the shop default of set, display showing no fault code through a full preheat and two cycles.
• Anti-tip bracket engaged, and on a gas range the connector joints bubble-free.

References

• See related: NTC Thermistors and Oven RTD Sensors Explained; Oven Bake Element and Temperature Calibration
• See related: Gas Range Connection and Leak Check Standard; The Back-Probe vs Disconnect Test Decision
• NFPA 70E, Standard for Electrical Safety in the Workplace, Article 120, current edition
• 29 CFR 1910.333, Selection and use of work practices (OSHA general industry)
• Manufacturer tech sheet sensor chart and installation instructions for the anti-tip bracket