When an oven drifts off setpoint

“It runs cold” and “it runs hot” are the same complaint. Both mean the oven and the control no longer agree about what the cavity temperature is. There are four places that disagreement can come from, and they are worth checking in this order — cheapest and most likely first.

1. The door and the gasket

Before anything electrical: close the door and look at the gap along the top edge, then along the hinge side. A cavity that bleeds heat forces the control to fire longer and more often to hold setpoint, and the owner experiences that as an oven that will not settle.

Then check the gasket itself. A gasket that has hardened, flattened along the bottom run or pulled out of a corner clip does the same thing. Both of these are visible in ten seconds and neither requires a meter.

If the door will not close square at all, stop here. That is a hinge and spring job, and no amount of sensor work will fix it.

2. The oven temperature sensor

The sensor is a resistance device. It changes resistance as it heats, and the control reads that resistance to decide when to fire. The important property is that a failed sensor rarely fails to nothing — it drifts. It keeps reporting a plausible number that is simply wrong, which is exactly why the oven still works and works badly.

Measure it cold at room temperature and compare against the specification for that sensor. If it is out of range cold, it is worse hot. This is a short check and it settles the question either way, which is why it comes before anything involving a board.

3. Igniter current, on a gas oven

On a gas oven the igniter is also the safety interlock. It has to draw enough current for long enough to open the gas valve. As it ages its resistance rises and its current draw falls, and there is a window where it still glows bright orange — looks perfect — but no longer pulls enough to hold the valve open reliably.

That produces an oven that lights sometimes, or lights and then goes out on a long bake, or takes several minutes longer than it used to. The visual check tells you nothing here. The amp reading tells you everything.

4. The control board

Last, not first. Boards do fail, and around here they mostly fail after a storm-season power event rather than from age alone — this region gets roughly 68 thunderstorm days a year and the equipment has usually sat through a decade of them.

But a board is the most expensive part in the sequence, and it is the one most often replaced by people who never checked the three cheaper things above it. If the door seals, the sensor reads in spec cold, and the igniter pulls its rated current, then the board has earned its turn as the suspect. Not before.

What the owner can check first

Two of these four are visible without a meter, and it is worth doing them before the appointment because they change what we bring.

Close the door on a strip of paper and pull it out at several points around the seal, top, bottom and both sides. Where it slides out freely the door is not sealing there. Then run a hand along the gasket: it should be resilient rather than stiff, and it should sit evenly in its channel the whole way round.

Then use an oven thermometer, sitting on the centre rack, and let a preheated oven settle for at least twenty minutes before reading it. A single reading taken the moment the preheat tone sounds is not evidence of anything — the cavity walls are still catching up with the air. What matters is the settled figure and how far it wanders over the next half hour.

The point of the order

Every step above rules something out with a measurement, not a guess. That is what makes it a survey rather than a parts swap — and it is why the diagnostic fee is flat rather than open-ended. The sequence takes the time it takes, and it is the same sequence every time.