Analysis of integrating electrometers with a time-pulse signal

Conclusions

• 1. The dependence of the measuring error on the signal being measured has been characterized for three methods of converting the current being measured into a time interval.

  For methods 1 and 2 (Figs. 1 and 2) this dependence is complicated and its effect cannot be eliminated by reasonably simple means. For method 3 (see Fig. 3) this dependence is proportional so that the measuring error can be compensated.

• 2. For higher requirements on the measuring accuracy it is advisable to use methods 2 and 3 because they permit operation with comparison thresholds approximately equal to zero, which improves the measuring accuracy.

• 3. If high accuracy is required simultaneously with linearity of the conversion function, it is advisable to use method 3.

From the results of experiments on these methods of forming a time interval a relation was obtained between the relative error σ of a digital conversion and the input current. The converter using method 3 had the best accuracy, especially in the low-current region, which agrees well with the theoretical conclusions obtained above.