Thus, by cutting the printed recorder out of the system and using 
only the tape punch, the sampling rate may be increased to 9-10 
channels per second. By using a faster tape punch (soon to be 
available), the sampling may be increased to approximately 17 
channels per second. If the recorder is not the limiting factor, the 
sampling rate could be increased to 50 channels per second (limited 
by present input scanner). Procurement of a magnetic tape re- 
corder would be necessary in this case for recording purposes. 
Finally, the digital voltmeter has a maximum possible conversion 
rate of 1000 readings per second. Also, the number of inputs to be 
scanned and the time between cycles or logging times can both be 
varied easily. 
OVER-ALL ACCURACY AND PRECISION 
As described in Appendix B, the thermistors are calibrated* 
in a tank in which the water can be maintained at an indicated temp- 
erature with a high degree of precision (see below). The accuracy 
of the temperature setting is determined by the accuracy of a. 
platinum thermometer which has been compared with a recently 
calibrated secondary temperature standard. The result is shown 
in fig. 10, curve A, which includes the accuracy of the temperature 
bath as well as the bridge measurement. 
The linearization of the thermistor-bridge combination as 
described in Appendix B is not perfect, although it is so good that 
corrections are not applied. The small inaccuracies that do occur 
are expressed in one form in the appendix. In practice (because it 
is easier to do so), the null and the sensitivity are calibrated at 
9°C and 18°C respectively; the relation between temperature and 
voltage is taken as a straight line between these points. For this 
case, the inaccuracies are given in figure 10, curve B. 
The error in accuracy of the temperature indicating system 
is the algebraic sum of the individual errors in accuracy; it is 
given in figure 10, curve C. Here the errors in accuracy of the 
digital data system are not included. The accuracy of the digital 
*Reference 3, page 55 
19 
