Sensitivy- Method

  • Different from most NMI‘s the sensitivity of TC‘s at BEV is measured once as a receiving test, and the corresponding value  is put into the software when AC/DC measurements with single- converters are done.
  • Circuit: The input of a TC is connected directly to a calibrator, the output connected to a nanovoltmeter (input and output low are grounded). All instruments are PC- controlled.
  • Usual measurement: After warm-up (about 1 hour) with nominal current the current is reduced by steps of 0.05 mA and after a  waiting-time (normally 60s) the voltmeter is read out and the calibrator is reduced to the next lower current. This measurement is performed from 5.05 (or 10.05) down to 1.95 mA.
  • Observation: Different from a normal behavior (see graph „A“) sometimes perturbations have been observed (graph „B“), where the output of the TC does not follow the square law (approximately). It is assumed that mechanical stress in the heater causes these disturbances.

    Investigations have been performed to check possible nonlinearities of the calibrator and the nanovoltmeter. It can happen, that the calibrator causes a disturbance at one point, when it switches its range (for example from 22 to 2.2 mA). Other perturbations have their origin in the TC‘s.

  • Measurement- variation: to investigate small perturbations the step is reduced to 0.02 mA. Sometimes the sensitivity was measured in the current- range from plus 5 mA to minus 5 mA (see graph „C“).
  • Effect: when using the converter in the current-area where a perturbation occurs, large AC/DC errors and large standard-deviations have been observed. In addition to that a large std-dev can be seen at low frequencies, also when used at nominal current, depending from the thermal time- constant of the stress in the heater.
  • „Noise“ – sheet: graph „A“ shows the normal std-dev. of the measurement- circuit; „B“ and „C“ the std-dev. of converters with perturbations as detected with the sensitivity- measurements.