News | March 7, 2003

Technical Feature: Designing and Characterizing TRL Fixture Calibration Standards for Device Modeling

Source: California Eastern Laboratories

By Julien Fleury and Olivier Bernard
California Eastern Laboratories

To measure and characterize devices, California Eastern Laboratories (CEL) uses different fixtures and calibration methods. One is the through-reflect-line, or TRL, which can be used to calibrate the HP8510 network analyzer.

Like other calibration methods, the TRL introduces a 12-term error correction vector for each frequency point. To calculate these terms, standards for which S-parameters are known must be measured. These standards include a short order open a through and a delay line with known electrical delay. The TRL calibration corrects phase and magnitude errors introduced by the sliding of reference planes and the insertion loss of cables, fixtures and connectors. Precision calibration standards should be used because each defect in a standard will be transformed into calibration errors.

This application note explains how to characterize and build a proficient TRL calibration kit for the HP8510 Network Analyzer. The first part explains how to characterize a TRL calibration kit. The second part describes how to improve the calibration kit using the measurements made on it. First, the TRL is described in more detail.

Calibration of a network analyzer
In any measurement, systematic error sources change the response of a device under test (DUT). Annuling the effects of cables, connectors, mismatched lines and couplers requires a calibration that provides correction term vectors for each frequency point. With multiple frequency points, this vector becomes a matrix. Two popular correction vectors are the eight-term and 12-term correction vectors.

The 12-term error correction
Twelve-term error correction permits the correction of six terms in forward and six in reverse mode. These six terms are:

  • ED, the directivity term
  • ER, the reflection tracking term
  • EL, the load match term
  • ET, the transmission tracking term
  • ES, the source match term
  • EX, the isolation term

These are seen in the flow graph (Figure 1), which represents the forward error terms. This correction is the most efficient, is available on high quality network analyzers and it corrects all systematic errors provided by the network analyzer. The eight-term calibration is often used in less expensive network analyzers. It is simpler but does not correct all the systematic errors, such as the errors introduced when switching to an alternate measurement direction.

Using a system of equations, it is possible to find 10 of the 12 error terms. The two missing terms are the forward and reverse isolation terms which can not be extracted from that set of equations. If the user wants to calibrate these parameters, isolation calibration with two loads (as in the OSLT calibration) is necessary...

Click here to download the complete paper in pdf format.