In article ,
Jon Elson wrote:

N_Cook wrote:

Probably 1977 , is about right, from product numbers
C26-P0606
2-77-2P
Does whatever fills the tube change its character over the decades,
expanding for any given temp?. Leaking I could understand causing
problem, but that would be one-off.
Does not seem to be any wear on the pivotted rocking plate inside , that
transfers bellows motion to microswitch plunger. Other pivot point moves
via cam action of the dial up knob, unlikely that is worn.
Over time the setpoint on the dial has had to go up and up and now it is
effectively off the top of thse scale and microswitch is hardly on.
There is adjustment inside for the switching point of the microswitch,
which at half the stem is probably correct. If the rocking plate bent
wiht the pressures then I'd expect to see too low a set point developing.
I'll take some temp / bellows deflection data just out of interest, but
also confirm it has reasonable linearity.
The original setting of the pivot seems to be by choice of the
deformations in the pivot/rocker plate, no way of setting at initial
assembly other than go/no-go testing of each one.
"Sensor" tube is well buried and next to impossible to replace with new,
entire or partial because impossible to make a join in the capilliary.
It is a leak. Probably an insanely tiny one, if it is only showing up after
almost 40 years. The fill is usally a Freon appropriate for the temperature
range.

Almost undoubtedly alcohol. If it were freon, it would all boil away no
matter how small the leak, because it would be pressurized.

Alcohol, because it's non-corrosive (while water is). All it needs to do
is convey the volume change from the bulb without friction, to make the
bellows expand.

Isaac