The analyzer provides a calibration wizard for each calibration type. The guided calibration consists of the following steps:
Select Physical Port Connectors and calibration kits at all calibrated ports.
Compile Calibrations: Select a calibration type and the physical ports to be calibrated (skipped for predefined calibrations).
Measure Standards: Acquire measurement data for all standards required for the selected calibration type.
Calculate the system error correction data (error terms) from the measurement data of the standards and apply the result to the active channel.
A successful calibration will supersede the previous calibration, discarding all previous system error correction data. To keep older correction data you can transfer them into a Cal Pool using the Calibration Manager.
The system error correction data determined in a calibration procedure are stored on the analyzer. You can read these correction data using the remote control command [SENSe<Ch>:]CORRection:CDATa. You can also replace the correction data of the analyzer by your own correction data sets.
Performing a manual calibration
Performing an automatic calibration
The first dialog of the calibration wizard displays a table to select the connectors and calibration kits for all calibrated physical ports.
The check boxes above the table allow you to select equal connector types and genders at all ports. For some multi-port calibration types, the port connector types must be equal, e.g. because they require a Through standard with known characteristics.
The table contains the following rows:
Physical Port Number #
The ports (and therefore the number of table rows) are determined by the active calibration type selected in the Start Calsubmenu. For a user-defined calibration (Start Cal – Other...) all measurement ports of the analyzer are displayed regardless of their port configuration.
To perform a system error correction, the physical port must be available as a source and receive port. This condition is met if it is selected as a measured port in the Balanced Ports and Port Groupsdialog (Meas is on).
Connector
provides a drop-down list to select the connector type An (f) behind the connector type denotes female connectors, an (m) denotes male connectors. Symmetric (sexless) connectors (e.g. PC7) are not labeled. User-defined connectors can be added or removed in the Available Connector Types dialog, which is opened from the Channel –Calibration –Calibration Kits dialog. at the port and its gender. If Same Connector at All Ports is active, the connector types at all ports (but not their gender) are always adjusted to the current selection. If Same Gender at All Ports is active, the genders at all ports (but not the connector type) are always adjusted to the current selection.
Ref. Imp. shows the reference impedance for the selected connector.
Calibration Kits
provides a drop-down list to select a calibration kit. The list contains all calibration kits available for the selected connector type. The assignment of a calibration kit to a connector type must be the same for all physical ports: If a calibration kit is changed, the analyzer automatically assigns the new kit to all ports with the same connector type.
Completness: The selected calibration kit must contain all standards needed for the active calibration type. If it doesn't the analyzer displays an error message.
Waveguide cutoff:If a user-defined waveguide connector is assigned to one of the calibrated ports, then the start frequency of the active channel must be above the waveguide cutoff. Otherwise the analyzer displays an error message.
Import Kit opens the Import Calibration Kit dialog to load and (if desired) activate a cal kit file. Next > opens the second dialog of the wizard to continue the calibration procedure:
In a user-defined (Other...) calibration, the Compile Calibrations dialog is called up.
In a predefined calibration, Compile Calibrations is skipped and the Measure Standards dialog is called up.
Detector selects the detector settings for the S-parameters which the analyzer uses during the calibration. The average (AVG) detector eliminates noise contributions which are superimposed on the measured signal; it is needed for noise figure measurements using option R&S ZVAB-K30. The Normal detector is recommended for all other applications.
If the calibrated channel has already been assigned to a cal group, the correction data overwrites the cal group data, so the new calibration will affect all channels assigned to the cal group. The network analyzer generates a notice message "New calibration will overwrite cal pool!" when opening the Select Physical Port Connectors dialog.
Checks on switching to the next dialog
When the Next> buttonis pressed the analyzer checks the calibration kits and the matching of the calibration standards and possibly displays a notice box (confirm with OK). This happens:
If one of the calibration kits is described by ideal kit parameters or typical values.
If different connector types are defined at the ports but the selected calibration type requires uniform connectors.
If a predefined 7-term calibration method TXX (TOM, TRL, TNA, TRM) is selected but one of the calibration standards is mismatched. A standard is assumed to be mismatched if it is described by S-parameter data (*.snp file) with a reflection factor > –30 dB.
Remote control:
[SENSe<Ch>:]CORRection:COLLect:CONNection<port_no>:PORTs [SENSe<Ch>:]CORRection:COLLect:CONNection<port_no>:GENDers [SENSe<Ch>:]CORRection:COLLect:CONNection<port_no> [SENSe<Ch>:]CORRection:COLLect:SCONnection<port_no> [SENSe<Ch>:]CORRection:CKIT:<conn_type>:SELect "<Ckit_Name>" [SENSe<Ch>:]CORRection:COLLect:DETector
The second dialog of the calibration wizard appears for user-defined (Other...) calibrations only. It is used to compile a list of calibration types and physical ports in order to perform all calibrations in a single measurement sequence. A measurement of a particular standard that is required for different calibration types can be reused.
The selected calibrations and ports appear in the List of Calibrations. A (N) behind the ports number denotes a node port (see Add Calibration dialog). The list of Standards to be Measured is automatically generated according to the List of Calibrations.
The dialog provides three buttons to extend or modify the List of Calibrations:
Add Calibration... opens the Add Calibration dialog to add a new calibration to the list.
Modify Calibration... opens the Add Calibration dialog to modify the selected calibration.
Delete Calibration removes the selected calibration from the list.
Two buttons at the bottom of the dialog move back and forward in the calibration wizard:
< Back opens the Select Physical Port Connectors dialog.
> Next opens the Measure Standards dialog.
If one of the calibration kits is described by ideal kit parameters or typical values, the analyzer displays a notice box.
[SENSe<Ch>:]CORRection:COLLect:METHod:DEFine "<cal_name>", REFLshort | FOPort | FRTRans | OPTPort | TOSM | TOM | TRM | TRL | TNA | UOSM, <port_no>[,<port_no>][,<port_no>][,<port_no>]
The Add Calibration dialog adds a new calibration to the list in the Compile Calibrations dialog or modifies an existing calibration.
Type contains a drop-down list to select a calibration type.
Calibrate Ports provides a list of all test ports of the analyzer (no external device ports). Ports with unsuitable port configuration (see Select Connector – Physical Port Number #) are grayed.
Node Port is available for One Path Two Port calibrations where it selects the fully calibrated port (i.e. the source port). All ports selected in the Calibrate Ports table can be used as node ports.
Checks on closing the dialog
When the OK button is pressed the analyzer checks the calibration kits and the selected calibration (s) and possibly displays a notice box (confirm with OK) or a status message. This happens:
If one of the selected calibration types is redundant. A calibration is redundant if another selected calibration provides more complete correction data. Redundant calibrations are deleted from the list of calibrations.
The last dialog of the calibration wizard is used to perform the necessary measurements of standards and to calculate the correction data.
Measured Standards displays the list of measured standards compiled in the previous dialogs. With the exception of the Isolation (optional) measurement, all standard measurements are required perform the selected calibration.
Structure of the Measured Standards list
The list of measured standards has a tree structure.
The first level contains all physical ports where one-port (reflection) measurements are required and all physical port combinations where two-port (transmission) measurements must be performed.
The second level contains check boxes for the standards to be measured at each port or port combination.
For a sliding match, a third level contains check boxes for the different positions of the sliding element.
Double-click a physical port symbol to expand or collapse the list.
Calibrations using a match or sliding match
If the calibration kit contains a sliding match standard, the Sliding Match appears in the Measured Standards list whenever the selected calibration type requires a Match. A click on the node expands the check boxes for the different positions of the load element. The number of different positions is defined in the User Interface tab of the System Configurationdialog.
The sliding match is a one-port standard consisting of an air line with a movable, low-reflection load element (sliding load). This standard is used because a no perfect match is available over a wide frequency range. However, a series of measurements at a given frequency with equal mismatch and varying phase yields reflection factors that are located on a circle in the Smith chart. The center of this circle corresponds to perfect match. The network analyzer determines and further corrects this match point following I. Kása's circle-fitting algorithm.
To obtain the reflection coefficient for a perfectly matched calibration standard, the sliding load must be measured at least at 3 positions which should be unequally spaced to avoid overlapping data points. Increasing the number of positions to 4 – 6 can improve the accuracy. Moreover it is recommended to use the predefined load positions of the standard.
The calibration is valid (Apply is available) if either the match or three positions of the sliding match have been measured. However, it is often desirable to acquire calibration data from both standards. The analyzer combines the data in an appropriate manner:
The match results are used up to the lower edge of the specified frequency range of the sliding match (Min Freq).
The sliding match results are used for frequencies above the Min Freq. In general, the sliding match will provide better results than the match within its specified frequency range.
Isolation (optional) measurement
For two or multiport normalizations, the Measured Standards list contains an Isolation (optional) standard in addition to each through standard. Measurement of the isolation is optional; the normalization correction is calculated as:
Correction value = (Transmission coefficient DUT – Isolation) / (Transmission coefficient Through – Isolation).
The isolation term accounts for a possible crosstalk between a pair of analyzer ports. The isolation standard is not a physical standard; it is recommended to terminate the analyzer ports with 50 Ω to measure the isolation.
Waveguide calibration
When calibrating waveguide ports, ensure that the calibration standards are connected correctly. The electric fields at each of the waveguide transitions must be parallel. For example, if you use a frequency converter and a waveguide calibration kit, the rectangular waveguide cross sections of the frequency converter flanges and of the waveguide calibration standards must have the same orientation.
TRL extensions: calibration with two lines, combination with TRM
Checking one of the boxes in the list causes the analyzer to stop the measurement in all channels except the active one and measure the standard according to the active channel settings. The progress of the calibration sweep and the result can be monitored in the diagram. In case of an error (e.g. if the measurement result shows that the calibration standard was not connected properly), Abort Sweep immediately terminates the sweep.
After completing the sweep the analyzer generates a short sound and a green checkmark appears in the checkbox. Measurements can be repeated as often as desired. Newer results overwrite older measurement data.
Most channel settings including the trigger settings remain valid for calibration sweeps and power calibration sweeps. To start the calibration sweeps without delay select Free Run trigger mode.
Checks for the calibration sweep
If the generator frequency range of the active ports exceeds the validity range of the standard model (defined by Min. Freq. and Max. Freq. in the Add/Modify Standarddialog) the analyzer displays a notice box (confirm with OK).
The Measure Standards dialog provides further controls:
Show Measurement Diagram displays or hides the diagram to the right of the list of Measured Standards. Hiding the diagram leaves more space for displaying the characteristics of the measured standards.
Show Phase displays or hides the phase trace (in addition to the dB Mag trace) in the measurement diagram.
Keep Measurement Data for >Repeat Previous Cal< causes the raw measurement data of the standards to be stored after the calibration is completed. This enables the Repeat Previous Cal... command, which can be used to optimize a previous calibration without repeating the measurement of all standards. If Keep Measurement Data... is not active, then the raw measurement data of the standards is deleted and the analyzer only stores the system error correction data. Deleting the raw data saves disk space.
Apply is enabled as soon as data has been acquired for all standards. The button starts the calculation of the system error correction data and closes the calibration wizard. The calibration is applied to the active channel. In the Port Configuration dialog, only the calibrated ports are enabled for the measurement. The current instrument settings are stored with the correction data. To avoid incompatibilities, older system error correction data is deleted unless it has been transferred into a Cal Pool using the Calibration Manager.If an unknown through standard is measured, Apply opens an additional dialog to check the calculated length offset parameters; see Unknown Through Standard.
The Keep Measurement Data for >Repeat Previous Cal< setting is valid for the current calibration only. To activate this function in general, use the parameter in the User Interface tab of the Sytem Config. dialog (menu System – System Config).
Checks during the calculation of correction data
Incompatibilities between the selected calibration type, the standards and the channel settings may cause the calibration to be inaccurate. The analyzer auto-detects potential sources of errors and displays appropriate, self-explanatory notice boxes.
[SENSe<Ch>:]CORRection:COLLect[:ACQuire]:RSAVe [SENSe<Ch>:]CORRection:COLLect[:ACQUire]:SELected [SENSe<Ch>:]CORRection:COLLect:SAVE [SENSe<Ch>:]CORRection:COLLect:DELete ["<cal_name>"] [SENSe<Ch>:]CORRection:STIMulus?
The frequency range for a TRL calibration is restricted due to singularities in the system of equations to be solved. Singularities occur whenever the length difference DL between the through and the line is an integer multiple of half of the wave length:
As a rule, singularities are avoided with sufficient accuracy if the phase shift resulting from the (electric) length difference between the through and the line standard is between 20° and 160°. This corresponds to a ratio of 1:8 for the start and stop frequency of the calibrated sweep range.
To shift the calibrated sweep range to smaller or larger frequencies, you can use a longer or shorter line. To extend the calibrated range, use one of the following methods:
Perform TRL calibration with two or three different line standards. With an appropriate length of the lines, the ratio for the start and stop frequency of the calibrated sweep range can increase to approx. 1:64 (for 2 lines) or 1:512 (for 3 lines).
In the low-frequency domain where TRL becomes inaccurate, replace TRL by TRM calibration.
The methods can be combined or used separately. The Measured Standards list for TRL calibration is extended if the calibration kit in use contains the necessary standards:
A 2-line (3-line) calibration requires two (three) different lines of matching gender. The lines must be measured between any combination of two ports.
A TRM extension at low frequencies requires either a match or a sliding match standard. The standard must be measured at each port.
The complete Measured Standards list for a two-port calibration is shown below. The calibration is valid as soon as the standards for TRL with 1 line have been measured. The extensions are applied automatically if the necessary standards have been measured.
Calibration with two or three lines
If several lines with different lengths are measured, the analyzer automatically divides the calibrated range into segments; see below. The calibration data of the longest line is applied to the lowest segment, the calibration data of the shortest line to the highest segment.
Calculation of segments
The calibration sweep segments for two lines with electric lengths llong and lshort (llong > lshort) are obtained as follows (the through standard is assumed to be of length lthr):
The longer line can be used up to a frequency flong where its transmission phase is equal to 160 deg. This frequency is equal to flong = 4*c0/[9*(llong–lthr)].
The shorter line can be used from a frequency fshort where its transmission phase is equal to 20 deg. This frequency is equal to fshort = c0/[18*(lshort– lthr)].
The border between the two frequency segments fdiv is calculated as the geometric mean of flong and fshort, i.e. fdiv = sqrt(flong * fshort).
The formulas are also applied if flong < fshort.
For a TRL calibration using three lines with different length, the allowed frequency ranges are calculated in an analogous manner in order to obtain three (ideally overlapping) frequency ranges. The borders between two adjacent frequency ranges are calculated as the geometric mean of the frequency limits flong and fshort of the two ranges.
A second or third line in the list does not mean that you have to measure two or three line standards. If the calibrated frequency range is small enough, the calibration is valid as soon as the analyzer has acquired correction data for a single line line standard. The match and sliding match standards are not necessary for TRL calibration, however, they must be measured if TRL is combined with TRM calibration.
TRL calibration with two lines
Defining cal kits with several line standards
Predefined calibration kits usually do not provide several line standards with different lengths. To generate a calibration kit appropriate for TRL calibration with two or three lines, proceed as described in Creating a user-defined calibration kit.
Low-frequency extension with TRM
TRL calibration becomes inaccurate if the electrical length difference between line and through standard corresponds to a phase shift below 20°. In practice this means that TRL is only practicable above a threshold frequency fTRM which depends on the lengths of the longest line and through standards. The threshold frequency is given by:
fTRM = c0/[18*(llong– lthr)]
where llong denotes the electrical length of the longest of the used line standards, lthr the length of the through. The analyzer assumes lthr << llong and calculates fTRM = c0/(18*llong). At frequencies below fTRM, TRL calibration is automatically replaced by TRM, provided that the necessary calibration data has been acquired. For a line with llong = 16.666 cm, the threshold frequency is fTRM = 100 MHz.
TRL calibration with low-frequency TRM extension
Accuracy conditions for the line(s)
The length error of the line, converted into a transmission phase error, must be below the minimum difference to the singularity points 0 deg or 180 deg multiplied by two. Suppose that an approximately known line standard causes a transmission phase 30 deg at the start frequency and of 160 deg at the stop frequency of the sweep. Its length error must cause a phase difference below (180 deg – 160 deg)*2 = 40 deg.
If an unknown through standard (or an unknown reciprocal mixer; see Mixer Vector Cal) is measured, the Apply button in the Measure Standards dialog opens an additional dialog.
The Unknown Through Characteristics dialog shows the delay time or transmission phase that the analyzer determined during the calibration sweep.
For a non-dispersive standard (Dispersive check box cleared), the Delay Time can be determined unambiguously, provided that the transmission phase difference between two consecutive sweep points is below 90 deg (make sure that the spacing between the sweep points is small enough). In this case it is sufficient to press Apply in order to calculate the system error correction data and close the calibration wizard.
For a dispersive standard (Dispersive check box selected), the delay time is frequency-dependent, however, the analyzer can determine two solutions for the transmission phase at the start frequency of the calibration sweep. The two solutions differ by 180 deg; the right solution must be selected from a drop-down list.
If you are not sure which one of the two phases for a dispersive standard is correct, select one, press Apply and check the measured result. If the transmission phase looks incorrect, use Channel – Calibration – Repeat Previous Cal... to select the alternative solution.
The delay time or phase determined by the analyzer can be corrected manually e.g. by entering a value derived from the mechanical length of the standard. Doubtful delay times or phases are displayed with a question mark.
[SENSe<Ch>:]CORRection:COLLect[:ACQuire] [SENSe<Ch>:]CORRection:COLLect[:ACQUire]:SELected
The following labels in the trace list inform you about the status or type of the current calibration.
Label
Meaning
Cal
The system error correction is applied without interpolation. This means that a set of measured correction data is available at each sweep point.
Cal int
The system error correction is applied, however, the correction data for at least one sweep point is interpolated from the measured values. This means that the channel settings have been changed so that a current sweep point is different from the calibrated sweep points. It is not possible to disable interpolation.
Cav
The system error correction uses variable calibration methods to calculate a measurement parameter. This happens for example, if a Z-parameter is calculated from S-parameters which are partly factory calibrated and partly normalized. This can also happen, if an S-parameter is measured with different port impedances, or if the analyzer uses a one port calibration at port 1 and a normalization at port 2.
Ca?
The system error correction is applied, however, the accuracy is questionable because of of the following applies:
The attenuator settings during the calibration differ from the attenuator settings during the measurement.
The bandwidth settings during the calibration differ from the bandwidth settings during the measurement.
The point delay settings during the calibration differ from the point delay settings during the measurement.
Cal Off !
The system error correction is no longer applied (e.g. turned off by the user). See also Calibration Overview.