De-embedding is covered in detail in Calibration and Measurement Enhancements but since the generation of files for some de-embedding exercises is so closely tied to adapter removal; it will be briefly discussed now. De-embedding is the removal of the effects of a network from a set of data. This network could represent an adapter or fixture among other things. To perform the de-embedding, the parameters of this network must be known. While there are many methods of deriving these parameters (including simulation), measurement in some way is often preferred. Because of the complex and incompatible media that may be involved, techniques using multiple calibrations (in different connectors or different media) or techniques using a pair of adapters/fixtures back-to-back are sometimes employed.

For the two port VNA, there are four types of extraction techniques available and are shown in the NETWORK EXTRACTION dialog box shown below in Figure: NETWORK EXTRACTION Dialog Box - 2-Port VNAs.

Two full 2-Port calibrations are performed; one each with the adapter/fixture attached to one port then the other. A single S2P file describing the adapter/fixture is generated. This is directly the method of adapter removal except the parameter file is generated explicitly rather than the calibration being directly modified.

A two tier calibration, some times called 1-Port de-embedding or the Bauer-Penfield technique. Here a one port cal is performed and then another one port cal is performed with the adapter/fixture in place. This is very similar to type A except a thru connect is not required and the far-end match is determined differently. A single S2P file is generated.

This is the network extraction method available in earlier generations of Anritsu VNAs where full 2 port calibrations are performed at the outer plane (often coaxial or waveguide) and at the inner plane (often a fixtured environment). Two S2P files are generated in this case.

This simplified method is used when standards at the inner plane are difficult to create (as in a complicated fixture structure). Two adapter/fixture “halves” are connected back-to-back and the combination measured using a single outer cal. Assuming the interconnect between the two halves is well-matched and the two halves are identical, S-parameters can be extracted. The match at the inner plane is assumed good (since no other information is available) so this technique is best used when the overall adapter/fixture return loss is quite high. Two S2P files are generated.