A Generalization of the TSD Network-Analyzer Calibration Procedure, Covering n-Port Scattering-Parameter Measurements, Affected by Leakage Errors
Ross A. Speciale
Abstract
Ross A. Speciale
Abstract
The basic philosophy of the through-short-delay (TSD) calibration procedure for two-port automated network analyzers has been extended to n-port scattering-parameter measurements, while also accounting for the errors due to possible signal leakage between all port pairs. The system errors are represented by the scattering response of a 2n-port virtual error network, having n ports connected to the device under test and n ports connected to an ideal error-free multiport network analyzer. The (2n)/sup 2/ T-parameters of the error network are explicitly expressed in blocks of n/sup 2/ at a time, as matricial functions of the 3n/sup 2/ S-parameters of three n-port standards, sequentially replacing the device under test during system calibration. The possibility has also been investigated of correcting the errors due to repeatable measurement-port mismatch changes, typical of switching scattering-parameter test sets. This capability has been introduced and tested in the classical two-port TSD calibration algorithm, by means of a minor modification and data postprocessing, applied after the removal of conventional errors.
OpenAlex reports 134 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The basic philosophy of the through-short-delay (TSD) calibration procedure for two-port automated network analyzers has been extended to n-port scattering-parameter measurements, while also accounting for the errors due to possible signal leakage between all port pairs. The system errors are represented by the scattering response of a 2n-port virtual error network, having n ports connected to the device under test and n ports connected to an ideal error-free multiport network analyzer. The (2n)/sup 2/ T-parameters of the error network are explicitly expressed in blocks of n/sup 2/ at a time, as matricial functions of the 3n/sup 2/ S-parameters of three n-port standards, sequentially replacing the device under test during system calibration. The possibility has also been investigated of correcting the errors due to repeatable measurement-port mismatch changes, typical of switching scattering-parameter test sets. This capability has been introduced and tested in the classical two-port TSD calibration algorithm, by means of a minor modification and data postprocessing, applied after the removal of conventional errors.
Key concepts: Scattering parameters, Network analyzer (electrical), Calibration, Device under test, Port (circuit theory), Scattering, Observational error, Leakage (economics)