Concepts and features
R&S
®
ZNA
190
User Manual 1178.6462.02 ─ 20
The multiple port assignment method considerably extends the range of applications of
the calibration units. However, the method entails some loss of convenience because
you have to reconnect the calibration unit between the different calibration stages
(assignments). It can also cause a loss of accuracy because only a subset of all possi-
ble through connections is measured.
To calibrate n test ports with an m-port calibration unit (m<n), the calibration unit has to
be reconnected at least n/m times. Each of these (re-)connections is described by its
"port assignment", i.e. the mapping of calibration unit ports to test ports. Then for each
assignment an automatic calibration is performed. Finally the analyzer combines the
calibration data and calculates the required n-port error terms.
This is possible if and only if:
●
the port assignments "cover" the calibrated test ports
●
the overlap between assignments allows a(n ordered) "chain of measured through
connections" between any (ordered) pair of test ports requiring a Through mea-
surement
For Full n-Port calibrations, the R&S
ZNA applies the "reduced through" logic to calcu-
late the correction terms for those test port pairs that are not covered by a single
assignment and hence cannot be measured directly (see
Chapter 4.5.1.11, "Full n-Port
calibration with reduced number of Through connections"
the calibration type, a "minimal" valid and complete solution can be described as fol-
lows:
Calibration
type
Minimal solution
Default solution (minimal)
Full One Port
Each calibrated test port must appear in exactly one
port assignment.
Subdivide the n test ports into
groups of m ports with increasing
port numbers. Create a separate
port assignment for each group.
One Path Two
Port
●
The node port must be included in all port
assignments
●
Each of the n-1 other calibrated test ports must
appear in exactly one port assignment
Additional condition:
The calibration unit port
assigned to the node port must be the same in all
assignments. This minimizes the number of port
reconnections between the calibration stages.
Assign the node port to port 1 of
the calibration unit. Subdivide the
remaining n–1 test ports into
groups of m–1 ports. Create a
separate port assignment for each
group. Leave port 1 of the calibra-
tion unit connected to the node
port and connect the ports of each
of the port groups to the remain-
ing m–1 ports of the calibration
unit.
Full n-Port
●
the two generic conditions stated above (with
"unordered Throughs")
●
the number of assignments must be as small
as possible
Additional condition:
The cal unit port assigned to a
given test port must be the same in all assignments.
Again, this minimizes the number of port reconnec-
tions between the calibration stages.
Same as for One Path Two Port
calibrations, using the test port
with the lowest number as "node
port".
(Every other calibrated port could
serve as "node port")
Among the minimal ones, the "star-shaped" solutions with fixed (but arbitrary) "node
port" are also optimal w.r.t. the (average) length of the "Chain of Throughs" and hence
Calibration
