Removal of time-varying errors in network-analyser measurements: Signal normalisation and test results

Research output: Contribution to journalArticlepeer-review

12 Scopus citations

Abstract

A new approach to network-analyser measurements has been developed, which uses a dynamic-error suppression technique to remove time-varying component-drift errors (such as amplifier- and filter-response changes) and physical-device errors (such as response changes due to cable flexure). This dynamic-error suppression technique is combined with a conventional static-error suppression technique, such as 'short/open/load/thru' (SOLT) or 'thru/reflection/line' (TRL), to remove all types of errors. This procedure is referred to as the accurate real-time total-error-suppression technique (ARTTEST). Special signal flow models and normalisation equations are developed for the suppression of both static and time-varying errors in the transmitter, receiver, reference, and link circuitry. In order to test the prototype system, time-varying reflection and transmission errors are introduced in the various paths in the ARTTEST network analyser. Based on these tests, it is found that the ARTTEST network analyser's dynamic-error suppression technique provides an improvement, when compared with conventional network analyser measurements, of greater than 50 dB for reflection and greater than 40 dB for transmission measurements. Corresponding improvements of greater than a factor of 300 for reflection phase and greater than a factor of 100 improvement for measurements of transmission phase are also achieved.

Original languageEnglish (US)
Pages (from-to)31-36
Number of pages6
JournalIEE Proceedings: Science, Measurement and Technology
Volume149
Issue number1
DOIs
StatePublished - Jan 2002

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Removal of time-varying errors in network-analyser measurements: Signal normalisation and test results'. Together they form a unique fingerprint.

Cite this