Porównywanie częstotliwości wzorcowych w środowisku LabVIEW

Generator częstotliwości wzorcowej jest niezbędnym elementem składowym wielu współczesnych przyrządów pomiarowych i innych urządzeń. Jego podstawowym parametrem jest częstotliwość generowanego sygnału wzorcowego oraz błąd względem wartości znamionowej, który wyznacza sie metoda fazowa poprzez porównanie ze wzorcem wyższego rzędu. W referacie przedstawiono zasadę metody fazowej porównywania częstotliwości wzorcowych, odpowiedni układ pomiarowy oraz aplikacje zrealizowana w środowisku LabVIEW przetwarzająca dane pomiarowe. Zaprezentowano rezultaty badan symulacyjnych oraz wyniki pomiarów w układzie rzeczywistym.

---

A precision frequency standard comparison method, measurement system and data processing based on LabVIEW environment are presented. Because the frequency of a standard oscillator changes due to aging, in order to remove the aging offset and return the oscillator to its nominal operating frequency calibration procedure should be used. The linear phase comparison method is widely used in the frequency standard comparison of high precision. Because this method has very high measurement resolution, it can distinguish and process a very small phase difference variation. However, the application of this method are only limited by the fact that the both frequencies of comparison signals must be the same. In this case, uncertainty of time measurement and duration of time measurement, will influence the comparison precision of the frequency standard. For higher resolution and convenient use, a method based on linear phase comparison was used. To determine the time differences between the reference signal and the signal of compared oscillator, typical digital time interval meter can be used, e.g. HP 53131A Universal Counter. The software of the system was developed in the LabVIEW programming environment. In order to reduce noise, smooth measurement data and find the mathematical relationship between signals were used different curve fitting methods implemented in LabVIEW, such as the Least Square (LS), Least Absolute Residual (LAR), or Bisquare (BS) fitting method. The main advantage of the proposed solution is that it can be easily implemented with general purpose measurements instruments. The algorithm was tested using data recorded from real frequency sources.