Oven Sensor Reads Correct But Bakes Uneven: Why The Reading Misleads
Why this matters
An oven temperature sensor that reads dead-on is the reading that quietly derails an uneven-bake call. The tech ohms the RTD probe, sees the textbook value, confirms the control holds setpoint, and tells the customer the oven is "within spec." But the sensor only reports the temperature at one point in the cavity. Uneven baking is a heat-distribution problem, and a perfect sensor reading is fully compatible with one side burning and the other staying pale. Trusting the sensor reading to close an uneven-bake complaint is a callback waiting to happen. This tree explains why the good reading misleads and how to diagnose distribution instead of setpoint.
Symptom presentation
The customer reports cookies done on one side and raw on the other, cakes that lean, or food that browns on top but not bottom. The oven holds its displayed setpoint, the control calibration checks out, and the sensor measures correctly: an RTD probe reads about 1080 to 1100 ohms at room temperature (roughly 1000 ohms at 32 F for a 1000-ohm platinum RTD, climbing about 2 ohms per degree F). By the meter the temperature is right, so the uneven result looks like a customer-technique issue.
Quick checks
Confirm the complaint is real and not loading. Center one rack, place an oven-safe thermometer or a thermocouple grid, preheat fully, and let the cavity stabilize for 20 minutes; ovens cycle widely before they settle. Confirm both bake and broil elements glow evenly along their full length with no cold or arced sections. Confirm the convection fan (if equipped) runs and the motor is not seized. Confirm the door seals fully and the gasket is intact; a leaking door corner bakes unevenly on that side. Verify rack position matches the food and that the customer is not blocking airflow with foil on the floor of the cavity.
Then read the sensor as a baseline. Correct resistance and correct setpoint hold confirm the control loop. They do not confirm even heat across the cavity.
Why the good reading misleads
The RTD measures temperature at its single mounting point, usually the upper rear of the cavity. The control loop drives the elements to make that one point match setpoint. Everything else about heat distribution is invisible to that loop:
- Element degradation along its length. A bake element with a cold section or a partially shorted coil still lets the control reach setpoint at the sensor while delivering uneven radiant heat below.
- Convection failure. A stalled or slow convection fan, or a failed convection element, leaves the cavity stratified. The sensor at the top reads correct while the bottom runs cold.
- Door and gasket leaks. A gasket gap at one corner bleeds heat from that quadrant. The sensor compensates by driving more total heat, masking the leak while the leaky side stays cooler.
- Air-baffle or fan-shroud damage. Bent baffles or a missing diffuser skew airflow so one region overheats.
- Sensor location bias. The probe sits in the hottest or a non-representative zone; it reads its spot accurately while the working rack runs different.
- Element timing/relay sticking. A broil relay that occasionally hangs adds top heat the control did not intend, browning the top while the sensor still tracks.
The sensor is honest about its one point. Uneven bake is about the other twenty points it never sees.
Isolation tree
Map the heat, do not trust the single number.
Run a multi-point temperature survey. Place thermocouples or oven thermometers at front, rear, left, right, and center on the working rack. Preheat and stabilize.
- All points within roughly 25 F of each other and of setpoint. Distribution is acceptable; the complaint is technique, rack choice, or pan material. Educate and document.
- One zone consistently runs hot or cold by more than roughly 25 to 35 F. A distribution fault is real. Continue.
Front/rear or top/bottom split. Suspect convection or element pattern:
- Convection oven, fan slow or stopped. Test the convection motor and element; a stalled fan stratifies.
- No convection, bottom cold. Inspect the bake element for cold sections; check that hidden/lower elements energize.
Left/right split. Suspect a door gasket leak on the cold side or a bent baffle steering air. Inspect the gasket corners and the rear baffle.
Top browns, bottom pale. Suspect a broil relay sticking on or excess radiant top heat. Watch element behavior across a full cycle.
Confirming diagnosis
The confirm is a logged temperature map, not a sensor ohm value. With the survey thermocouples in place, run a full preheat and a steady-state hold and record each location every minute through several control cycles. A healthy cavity holds the working rack within a tight band around setpoint after stabilization. A documented gradient that tracks to a specific element zone, a stalled fan, or a leaking gasket corner is the diagnosis. Re-running the map after the corrective action and showing the gradient collapse is what closes the call and prevents the callback.
If the map is even and tight, the appliance meets specification and the fault is use-related. Show the customer the logged numbers rather than the single sensor reading.
Remediation
- Bad element zone: Replace the bake, broil, or convection element. Confirm even glow and re-map.
- Convection fault: Replace the convection motor or element; clear any obstruction in the fan shroud.
- Gasket/door leak: Replace the door gasket; adjust or replace a sprung hinge so the door seals flat.
- Baffle/airflow: Straighten or replace bent baffles and diffusers to restore designed airflow.
- Sticking relay: Replace the control relay board if a broil or bake relay hangs on.
References
- UL 858, Standard for Household Electric Ranges
- DOE, 10 CFR Part 430 Subpart B, Appendix I, uniform test method for conventional ovens
- AHAM OV-1, Performance of Household Electric Ovens
- ASTM F1496, Standard Test Method for Performance of Convection Ovens (commercial reference for distribution mapping methodology)