| Popis: |
The IEEE has many industry standards. One of these is standard number 1057 which addresses the test of Waveform Recorders in its different performance related metrics. One of those metrics, specifically, deals with the estimation of aperture uncertainty (jitter). Here we propose that the same test that is used to estimate the jitter standard deviation, the one proposed in clause 12.2.2 (IEEE 1057–2017 version), can also be used provide an estimate of additive noise standard deviation. The main advantage is the reduced test time since the data points that are used for this new estimation are the same and thus renders unnecessary to carry out other tests specific only to random noise estimation. Here we present the new estimator which uses the mean square error of the sampled voltages relative to the least-squares error fitted sinewave of the two data sets acquired with two sinusoidal stimulus signals with different frequencies. Together with this proposal, the statistical properties of the estimator are studied, and an analytical expression is given for the estimator standard deviation as a function of the stimulus signal frequencies and amplitude, digitizer quantization step and number of acquired samples. This is invaluable in choosing the test parameters that lead to the shortest test duration by optimizing the number of samples to acquire in order to achieve a desired bound on the estimated value variance. |
