Study Of Calibration Techniques For Vector Network Analyzer

The vector network analyzer(VNA), widely used in the aerospace, satellite communications and radar monitoring, play an important role in the filed of modern high-precision and intelligent microwave measurement. The measurement accuracy of the VNA is affected by the non-ideal hardware features, test environment variability and device under test(DUT) connected nonrepeatability. Using higher performance module and higher precision calibration standards can improve the the measurement accuracy, but increases the technical difficulty and the cost of the VNA. The calibration and error correction, which is the key techniques of vector network analyzer, is one of the indispensable steps for high precision measurement. Research on calibration techniques for VNA, not only improve the measurement accuracy, but also can improve the vector network analyzer measurement speed. The author’s works and creative achievements can be summarized as follows:1. The hardware structure, measurement principle and system error model of two port VNA are introduced. The calibration technique based on 12-term error model and 8-term error model is studied for VNA with three measurement channels and four measurement channels respectively. Without considering crosstalk errors, based on Transmission parameters and SOLT calibration process, a simplified formula for error correction is derived.2. A calibration method based on 8-term error model of three-channel vector network analyzer is proposed. On account of crosstalk errors and switching errors, 8-term error model for vector network analyzer not able to achieve calibration accuracy of 12-term error model. With the inconsistent reference planes for the forward and reverse measurement, the plane-compensation errors are described to improve 8 error model, and the corresponding error correction formula is derived.3. Based on 4n-term error model, a calibration technique for n-port vector network analyzer with n+1 measurement channels is presented. The error model with the switch offset error term is closer to the actual situation. Using the matrix operation and the concept of general SOLT connection, the formalae can be simplified and obtained for n-port vector network analyzer.4. Based on general signal flow graph and general two-port concept, a calibration technique of the n-port vector network analyzer is explored. To describe 3n2-term error model for n-port VNA, the coefficients in signal paths are replaced by the complex square matrixes and the nodes correspondingly become complex vectors. Although the crosstalk error term is considered, the calculation procedure is simplified and measurement efficiency is enhanced. Finally, S-parameters of n-port device can be solved by a unified formula.5. Without considering crosstalk errors, based on Transmission matrix and GSOLT calibration process, a simplified formula for error correction is derived. This method has advantages of rapid measurement speed in measurement of multiport devices.6. Based on OSL calibration, a technique for second-order correction of the system error parameters of a calibrated two-port VNA is presented. The determination of the complex-valued residual directivity, source match, and reflection tracking by one single reflection measurement employing a high precision airline terminated by a short. A complete set of second-order corrected error model parameters is determined resulting in an accuracy enhancement.