[time-nuts] Capacitive temperature sensing

Bruce Griffiths bruce.griffiths at xtra.co.nz
Fri Aug 22 22:15:14 UTC 2008


Magnus Danielson wrote:
> My proposal to use capacitive sensing rather than conductive sensing 
> would handle the electrode oxide issue. It is meant as a means to go 
> around the sensing issue with parts at hand and only some new electronic 
> design of very simple form, not the means to supercalibrate something.
>
> I guess this only shows that time-nuts are time-nuts...
>
> Cheers,
> Magnus
>
>   
Hej Magnus

A capacitive sensing AC bridge can be very sensitive, one only has to 
look at the work of RV Jones at the university of Glasgow in the 50's, 
60's and 70's. He and his collaborators used capacitive sensors to 
detect (among other things) the extrusion of a micrometer shaft as it 
was clamped to realise just how sensitive it can be.
They found it possible to detect length changes of less than 1E-11m with 
a suitably designed sensor.

Even more sensitive capacitance bridge displacement sensors have since 
been constructed. Off course the critical bridge components have to be 
maintained at a reasonably constant temperature. Since the most critical 
component the balancing capacitor in the other bridge arm can be very 
small it can easily be located in the same controlled temperature 
environment as the sensor itself. One of the major contributors to 
instability will be creep in the glass capillary and bulb dimensions. 
Even with a relatively crude guard ringed coaxial sensor electrode 
surrounding a capillary tube is capable of submicron sensitivity without 
undue effort.

If one uses a mercury in glass thermometer with say 0.1C resolution with 
the 0.1C graduation say 0.5mm apart then 1micron change in mercury 
column length is equivalent to a temperature change of 200uK. The 
performance will be determined largely by mechanical instabilities not 
the bridge sensitivity.

Bruce






More information about the Time-nuts_lists.febo.com mailing list