Oven Temp Inaccurate Decision Tree
Why this matters
"Oven runs hot" or "oven runs cold" calls are deceptive. The customer is comparing actual baking results to a stored expectation, not to a calibrated measurement. Some calls are real component failures and some are user-side issues (recipe, altitude, rack position, preheat assumed complete when it is not). The decision tree below puts a calibrated thermocouple in the oven on minute one, splits real faults from perception faults, then walks the real fault to thermistor, igniter, gas valve, control board, or element in priority order. Reaching the right answer fast keeps a control board misdiagnosis, the most expensive part in the range, from torching the margin.
The decision flow at a glance:
Oven temperature inaccurate?
|
+-- 1. User offset set (~10%)? --> reset calibration
| to zero, re-test
|
+-- 2. Bake element worn? -------> cold spots / 15-25
| ohm: replace
|
+-- 3. Thermistor drifted? ------> 100+ ohm off at
| room temp: replace
|
+-- 4. Igniter weak (gas)? ------> current <3.2A:
| replace igniter
|
+-- 5. Valve coils? -------------> glow good, no open:
| ohm coils
|
+-- 6. Convection issue? --------> fan wobble / seized
| element: inspect
|
+-- 7. Control board? -----------> commands wrong
| duty: diag mode
|
+-- 8. Door seal? ---------------> paper-strip drag
| test: replace
| gasket
Symptom presentation
The customer reports cakes burning, casseroles undercooking, or general "the oven doesn't bake right anymore." Probe questions:
- Is it consistently off in one direction, or intermittent?
- Did it happen suddenly or gradually?
- Bake only, broil only, or both?
- Convection vs conventional difference?
- Did anything precede it - power outage, range bumped during cleaning, self-clean cycle run?
Sudden-onset bilateral inaccuracy after a self-clean cycle is the leading pattern for a blown thermal fuse or a heat-damaged thermistor mount. Gradual drift over months is usually a thermistor aging or a worn-out bake element.
Quick checks (under 15 minutes)
- Pull the racks and place a calibrated thermocouple at center-of-cavity (12-channel data logger if you have it, single-probe NIST-traceable if not). Avoid IR thermometers - they read the cavity walls, not the air the food sits in.
- Preheat to 350F and let the oven cycle 3 full cycles after the preheat tone. Record peak and trough.
- Compare to the OEM-published cycling band. Most domestic ovens specify plus or minus 25F around setpoint averaged across a cycle - that is the spec, not "stays at 350F." A 30-40F swing is normal; a 60F swing or a setpoint-average miss of more than 15F is a real fault.
- Visual element check on electric ovens. Bake element should glow cherry-red uniformly across its length within 5 minutes. Cold spots = open element segment.
- Visual igniter check on gas ovens. Should glow bright orange within 30 seconds, then gas valve should open and flame should establish across the burner within another 30 seconds.
If the calibrated test shows the oven is actually within spec, the conversation shifts: this is a recipe/altitude/expectation issue and you can offer the customer a calibration offset (most domestic ovens support plus or minus 35F user offset via the keypad) to match their target.
Isolation tree
Branch 1 - User offset already set (~10%). Customer or previous tech adjusted the offset and forgot. Enter the calibration menu (procedure varies by OEM, see the tech sheet). Reset to zero, re-test.
Branch 2 - Worn or open bake element (electric, ~30%). Element shows cold spots, hot spots, or visible break. Ohm cold - typical 15-25 ohms across the terminals. Open or substantially out-of-spec = replace. Verify the broil element separately; broil-only inaccuracy can mask a bad bake element if the unit runs both during convection.
Branch 3 - Drifted thermistor (~25%). Modern domestic ovens use a thermistor (typically 1100 ohms at 75F, rising with temperature on most platforms). Pull the thermistor, ohm cold, compare to OEM table. A drift of 100+ ohms at room temp is enough to throw cavity temp 30F at setpoint. Replace the thermistor; no calibration adjustment compensates a drifted sensor.
Branch 4 - Failed igniter (gas, ~20%). Hot-surface igniter draws current proportional to its resistance; as it ages, resistance climbs, current drops, the safety gas valve does not see enough current to open and gas does not flow - or flows late. Igniter resistance below 80 ohms is suspect; below 40 ohms is good; above 200 ohms is failed. Measure current draw with a clamp meter at the igniter circuit - 3.2A minimum to open the safety valve on most domestic gas ovens.
Branch 5 - Bad bake gas valve coils (~10%). Igniter glows correctly, current is in spec, but the bake valve does not open or opens late. Ohm the bake coil per the OEM service manual. Bake-only inaccuracy with proper igniter behavior points here.
Branch 6 - Convection fan or convection element issue (~5%). Convection inaccuracy only - the cavity reads correct but baked goods burn one side. Convection fan motor wobble or seized convection element. Visual inspection from inside the cavity with the fan cover off.
Branch 7 - Control board (~5%). Everything else in spec, board logs the setpoint but commands the wrong duty cycle. Verify only after thermistor and elements are confirmed good. Diagnostic mode shows commanded vs measured temp.
Branch 8 - Door seal blown (~5%). Casing-mounted gasket torn or compressed. Cavity loses heat through the door, element runs at high duty cycle, edges of food overcook while center undercooks. Inspect with the door closed and a paper strip caught between gasket and door frame - the strip should drag firmly on removal across the full perimeter.
Confirming diagnosis
- Element: ohm test at the element terminals with leads disconnected. Visual cold spot during preheat confirms.
- Thermistor: ohm at room temperature and at boiling (212F) in a beaker of distilled water - the OEM service manual lists the two-point check.
- Igniter: current draw with a clamp meter at the igniter lead during ignition cycle. Resistance test as a secondary confirmation.
- Gas valve coils: ohm the bake and broil coils separately. Spec is typically 1.0-2.5K ohms - confirm against the tech sheet.
- Control board: entering service mode and reading the actual commanded duty cycle against the measured thermistor input. A board issuing incorrect commands with correct thermistor input is the only correct condemnation criterion.
- Door seal: paper strip drag test plus IR thermometer scan of the door perimeter during heat - a leaking seal shows a clear plume of warm air on the IR.
Remediation - repair vs replace ROI
Temperature faults are the cheapest oven complaint to cure. Seven of the eight branches are a low-cost part and about an hour on site, which means almost every one of these calls should end in a repair.
| Branch | Relative part cost | Call it |
|---|---|---|
| 1 user offset | None | Reset, re-test, bill the diagnostic. Show the customer the menu so they can trim it themselves |
| 2 bake element | Low | Always repair. Ohm the broil element while you are in there and quote both if either is marginal |
| 3 thermistor | Lowest part on the machine | Always repair. Never try to offset around a drifted sensor |
| 4 igniter | Low to moderate | Always repair on a gas oven that is otherwise sound |
| 5 gas valve | Moderate | Repair. The valve is a safety component, so no partial fixes and no cleaning it up and hoping |
| 6 convection fan or element | Low to moderate | Repair |
| 7 control board | Highest part on the machine | Judgment call, see below |
| 8 door gasket | Low | Always repair, and check the hinges while the door is off |
The board is the only real decision. On a wall oven or a range past roughly a decade, a control board can run a meaningful fraction of a replacement unit, and boards rarely fail alone on an oven that has been cooking hot for years. Before you order one, confirm the thermistor and elements are genuinely in spec, and price the board against the installed cost of a comparable new appliance. Past about half, put both options in front of the customer.
Things that move the line:
- A built-in wall oven is not a swap. Cutout sizes have changed over the years and a replacement often brings cabinet modification with it. Repair tolerance on a built-in runs well past what you would accept on a freestanding range.
- Hinges and a sagging door turn up alongside Branch 8. A gasket fitted to a door that no longer closes square fails again in months. Quote them together or explain why you are not doing the gasket alone.
- Parts support and warranty. Check availability before quoting a board, and run the serial-based warranty lookup first. Elements and boards often carry longer coverage than the base term.
- The in-spec oven. If the calibrated test says the oven is fine, do not sell anything. Set the offset to their target, show them the probe reading against their own thermometer, and document that the unit met the cycling band. That call ends with a satisfied customer and no parts on the truck.
References
- UL 858 Standard for Household Electric Ranges
- NFPA 54 National Fuel Gas Code (gas range connection and service)
- NEC 422.31 (disconnect requirements for appliances)
- DOE 10 CFR 430 Subpart B Appendix I (conventional oven test procedure)
- GE Appliances Service Bulletin (oven temperature calibration offset procedure for Profile and Cafe series)