Dryers Self-Adjusting PID With No Retained Memory

I put a thermocouple in my dryer duct to monitor and trigger an alert when the dryer was done, as the beeper on the dryer can’t be heard unless you’re in the utility room in the basement.

On medium heat the temperature rises to 125°F in about 25 minutes. Then it drops to 111°F. Up to 124. Down to 112. Up 123. Down 113

Until it figures out the PID (I’m guessing) and maintains 119-120°F until done. However, it doesn’t figure this out until about 5 minutes before the clothes are dry.

Next cycle = no PID memory. Start all over. [FYI - Samsung]

On the side, under Optional Tags there should be a “Rant” option, maybe an “OT” too

Can you put in some sort of moisture sensor?

I have a Samsung and it has a pop-up on my phone and or TV when its done

It runs its cycle on a moisture sensor, dries until the sensor is appeased and then goes into cool-down for 2 minutes. This seems like a PID for the gas control valve.

The other thing I would like to do is add an electronic ignition spark controller to light it, as it has a flat lightning bolt shaped piece of steel that gets red hot to light the gas, I figure a spark control would use a lot less electricity and light the gas faster. It does turn the gas and steel heater on together and blow unburned gas out the duct until the plate heats and ignites the gas, then continues heating the plate constantly. Just borrow an ignition control from a gas stove.

PS - no online appliances in my house except a thermostat

what is the character of this temperature swinging? Is it a sine curve, or does it decay toward each local extremum?

What is the gas valve doing? is it variable or is it doing bang-bang control?

I would guess bang bang, as proportional valves cost more, and typically larger foot print.

KenMoore is right. A proportional valve would cost more than the price of a dryer.

It follows sort of a sine wave like monitoring any self tuning PID would make.

Nice, symmetrical sine wave like this [updated to add decaying amplitude]?

Or asymmetrical cycles, like this?

Wait a minute, if PV to the feedback control is a moisture sensor, then maybe that PV is affected by both temperature and how wet the clothes are.

So when the clothes are wet (and cold?), the “off” and “on” moisture triggers occur at 125°F and 111°F, but as the clothes lose moisture (and get warmer?) the temperature gap between the “off” and “on” moisture triggers drops.

That fits the data, especially the part where the final trigger temperature gap is small when the clothes are dry (no more moisture coming out), but I am struggling to get my head around the physics. Whatever it is, I don’t think it’s a PID control.

From what I know of classic Whirlpool style electric dryers, the temperature fluctuations are intentional and an important part of drying efficiently. The heater element is cycled on and off by a limit with the desired deadband. If that limit fails and the temperature does not cycle, the clothes take much longer to dry, even though it still actually gets hot.

1 Like

interesting. Google AI has some explanation (below). Given that, temperature upper limit (and maybe the lower limit) is a function of the moisture upper limit. Also when the clothes have more moisture in them, they might be able to heated to a higher temperature without risk of shrinkage.

Asymmetrical, but reversed. Longer to heat and cooling is quicker

I follows the same when on timed dry and not dampness controlled. Plus a friends (newer) electric dryer gets up to temperature and almost immediately maintains a constant temperature with very little fluctuation.

AI frequently makes up answers to validate whatever hypothesis you have, then you feel better about using AI in the future.

Ask all the lawyers that have been sanctioned for filing AI “hallucinations” that gave them non-existent court case transcripts and records (even one making up a court that didn’t exist) Turns out every time a court case is referenced in a court filing the court has someone contact the other court and verify it

fair point about AI. I had asked it for sources and it gave me nothing useful.