How we measure an oven that bakes badly

The oven is not accurate any more is the most common complaint we take, and it is the one most often misdiagnosed — by owners, and by technicians in a hurry. The reason is that a single thermometer reading answers a question nobody asked.

Three numbers, not one

An oven has a set point, an average temperature, and a swing. They are independent.

Number	What it is	What a bad value does to your food
Set point	What you asked for	Nothing by itself
Average	The mean cavity temperature over a full cycle	Everything cooks fast or slow, evenly
Swing	How far it rises and falls around that mean	Pale tops, burnt bases, inconsistent results

Every thermostatically controlled oven cycles. The element or burner comes on, the cavity overshoots the set point, the heat is cut, the cavity falls below, and the cycle repeats. A healthy swing is modest and symmetrical. A wide swing produces exactly the complaint people describe as inaccurate even when the average is dead on the set point — which is why a thermometer glanced at once reads correct while the baking is plainly wrong.

What we actually do

We put an independent probe in the cavity, at rack height, away from the walls and away from the sensor, and we log a full preheat plus at least two complete cycling periods. That takes time and it is the reason our diagnostic visit is not five minutes. It is also the only method that separates three faults that present identically.

Sensor drift. The average sits consistently above or below the set point, and the swing is normal. In a built-in of this class the sensor is usually a 1,000-ohm platinum RTD: cold, at room temperature, a healthy one measures a little over 1,000 ohms, and resistance climbs with temperature. A drifted sensor often reads plausibly cold and lies badly hot, which is why a cold resistance check alone is not proof.

Weak or intermittent heat. The average sits low and the recovery after each cut-off is slow. That is an element with a failing section, a gas valve not opening fully, or a relay contact that has pitted so the element is not getting full voltage every time it is called.

Air path. The average and the swing both look reasonable and the food is still wrong. That points at convection — a fan not turning at speed, a baffle fouled, a blocked cavity vent — or at an installation without the clearance the appliance needs.

Calibration is a last step, not a first one

Most ovens allow a calibration offset. Applying one to an oven with a wide swing, a weak element or a blocked vent hides the symptom and guarantees the customer comes back. We only adjust an offset once the swing is normal and the fault is a genuine, stable difference between the sensor’s opinion and reality.

What you can do before we arrive

Two things make the visit shorter and both cost nothing.

1. Note whether the problem is everything cooks wrong the same way or results are inconsistent between batches. The first suggests average, the second suggests swing, and that halves the search before we open the door.
2. Note whether it changed suddenly or drifted over months. Sudden points at a component; gradual points at drift or at something being blocked slowly.

What we will not do

We will not swap a sensor on suspicion, and we will not quote a board because the oven is behaving oddly. Both are common and both are guesses. The log either supports the part or it does not, and if it does not we keep looking. The diagnostic fee is flat at $80 and credited toward the repair, so there is nothing gained on our side by shortening the measurement.