In article , Sylvia Else wrote:
Steve Barker wrote:

Well when you get right down to it, why does it need circuitry at all?
It's a motor and a heat element.

Marketing. I saw a washing machine that was advertised as having
"crystal control". Presumably, it timed its wash cycles to one part in
10^9.

Crystals aren't that good unless especially good and in
temperatrure-controlled ovens. I would hope for one part in 10^6 to one
part in 10^7 or so.

Meanwhile, crystal oscillator circuits are not expensive to make. Even
modules of such are easily available and cost maybe a couple bucks or so,
maybe closer to a buck in quantities of tens or hundreds of thousands.
Cheaper crystal oscillator circuits may be had for something like 50-60
cents each in quantities of tens of thousands.

The alternative for reliably having a complete cycle's timing down to
seconds out of an hour is "power line time base", costing a fraction of a
buck, probably a smaller fraction, less than cheaper crystal oscillator
options.

Cheaper still requires oscillators regulated by component tolerances -
and oscillators tend to have at least two components affecting frequency,
and finest tolerance of ones easily available at premium prices is 1%,
meaning low chance of achieving timing highly reliably better than 2% in
either direction unless either crystal or power line time base is used.

Mechanical timers use "synchronous motors", which gives "power line time
base", good enough to make electric clocks with. Many digital electric
clocks use electronic means to make use of "power line frequency time
base".

- Don Klipstein )