MEASUREMENT AND CONTROL OF TEMPERATURE 



measurement current through the thermometer wire will heat it unless it is 

 very small, the temperature measured will in fact be that of the heated wire 

 less the cooling effect of its surroundings, and it may well heat its surround- 

 ings appreciably. This current must be kept to an absolute minimum. 

 A typical thermometer has a resistance of the order 100 Q, with an order 

 of 0-4 n/degree C change. Readings to 0-01 Q (0-025°C) should be quite 

 straightforward, with the precaution that a voltage of 1 V across the element 

 will cause it to dissipate 10 mW — a fine coil could easily rise TC when 

 passing this current — and the actual temperature in the element will be 

 different according to the thermal capacity and rate of fluid flow past it. 



Figure 29.5 A.c. bridge circuit of a resistance thermometer (RT) 



The resistance in series with the element must be at least of 1 kD, or larger, 

 in order to limit the current in the element to a negligible heat effect (i.e. less 

 than a mW). The support for the very fine wire must also be minimal in 

 size — mica sheet is suggested. Calibration against a mercury thermometer 

 in a gas needs an extremely long time for the mercury to equilibrate; it is 

 possible to calibrate in a bath of pure mineral oil provided: (1) it is ascertained 

 that immersion has no effect on the resistance of the element and (2) that 

 it is cleaned with chloroform and pure alcohol afterwards. 



Conclusion — The resistance thermometer is an excellent accurate method 

 for measuring gas temperatures and for following quite rapid fluctuations 

 of temperature in them. The application of a seff-balancing bridge would 

 enable it to be used with recording gear, though it is rarely used as a control 

 element. 



The thermistor — This is a resistance thermometer of a very different kind 

 which merits special treatment. It consists of a minute bead of heavy- 

 metal oxide which has a large negative temperature coeflScient of resistance 

 and offers an extremely convenient method of measuring temperature to 

 quite fine limits. It has a small thermal capacity and can be obtained 

 sealed in a very thin glass capsule so that, unlike the platinum thermometer, 

 it can be immersed in most liquids: it is therefore not subject to contamina- 

 tion effects, it is reasonably stable against aging and is certainly no less 

 robust than a mercury thermometer or other glass instrument probe. A 

 whole range of types and resistances is available (2 Q. to 100 kQ at 20°C). 

 The resistance change is logarithmic and it approximately halves its resistance 

 for every 20° rise in temperature; a combination of very high rate of 



390 



