The types of calibrations are defined by what ports are involved and what level of correction is accomplished (see Table: MS464xB Series VNA Calibration Types).

This is the most commonly used and most complete calibration involving four ports. All sixteen S-parameters (S11, S12, S21, S22, S13, S23, S33, S31, S32, S14, S24, S34, S41, S42, S43, and S44,) are fully corrected.

This is the most commonly used and most complete calibration involving two ports. All four S-parameters (S11, S12, S21, and S22) are fully corrected.

A single reflection parameter is fully corrected in this case (S11 or S22). Both ports can be covered but only reflection measurements will be corrected. This calibration type is useful for reflection-only measurements including the possibility of doing two reflection-only measurements at the same time.

In this case, reflection measurements on one port are corrected and one transmission path is partially corrected (load match is not). Here forward means S11 and S21 are covered while reverse means S12 and S22 are covered. This technique may be used when speed is at a premium, only two S-parameters are needed and either the accuracy requirements on the transmission parameter are low or the DUT is very lossy (@ >10-20 dB insertion loss).

This calibration is essentially a normalization and partially corrects one parameter (although two can be covered within the cal menus). Only the frequency response, or tracking slope, of the parameter is corrected so directivity and match behaviors are not taken into account. This technique is valuable when accuracy requirements are not at a premium and a quick measurement is all that is needed.

Each of these calibrations has an associated error model that describes what is being corrected. The error coefficients used fall into several categories that roughly describe the physical effect that they are responsible for correcting. To establish a context for these error terms, consider the usual model where all of the VNA/setup errors are lumped into error boxes (that act like S-parameters) between a perfect VNA and the DUT reference planes (see Figure: Classic One- and Two-Port Error Models).

Two slightly different error models are used: one where each port is considered to be driving separately (so one can clearly delineate source match from load match) and one where both ports are present and no driving distinction is made (requiring some preprocessing to take care of source match-load match differences).