The RI 7100 tester has reliable error-correction protocols built into the software to quickly and reliably correct for measurement errors in the testhead. There are applications that require the same robust error-correction for signal-separation devices outside of the normal RF testhead. This note shows a process for creating an "error adapter" array and how to use it in a testplan.

Test Configuration

For the purpose of this note, we will assume that we are going to correct for the path errors in an outboard directional coupler. The coupler is driven on its main line from port RF7 of the testhead. Its "incident" coupled port is connected to RF3, and its "reflected" port is connected to RF6 on the testhead. We will use a short, an open, and a fixed termination as our calibration standards.

Calibrating the system

The "error adapter" is a two-port S-parameter data object which holds the coefficients which are used to correct for the systematic errors in the system. Normally, these objects are held in the Test System object or the Fixture object -- depending on the path being corrected by the error adapter. When you are correcting for external fixturing, however, we recommend you place these objects in a Device Interface object.

Before we create the error adapter, we need to create a place to store it. First, create a new Device Interface object. Edit the Device Interface and "Inspect" the calibration data. On the left-hand side of the cal data inspector, click the right mouse button (RMB) and choose "add item" from the menu, as shown below:



Enter a name, such as one shown here, and you will have a place to store your error adapter. It will also be easy to retrieve when the time comes to use the error adapter to correct S-parameter measurements.

Now we are ready to create the calibration testplan. For our calibration testplan (Cal.DTP below), the Global Defaults will set the state of our test system:



Because RF3 and RF6 will be used to measure a wave parameter, each is set to "receive" mode in the defaults. RF7 is set to "src1-noise" and Source 1 is set to drive RF7. The receiver is set to S-parameter mode, and parameter b1 is selected because we will be using the "input port" button later to select the port to be measured. Nominal values are chosen for the Source 1 power, the Receive Attenuation, and IF Gain. These will, of course, be dependent on your final fixturing.

Three test sections are used to stop the tester and prompt for a calibration standard, and then measure the open, the short, and the load. The "open" measurement panel is shown below:



The critical items here are the Measure Group boxes used to bind the receiver measurement to either port RF3 or RF6. The two measured wave parameters are sent to a special calculation block which creates an S11 measurement from the wave parameters. The current frequency is joined to the measurement and it is saved in a local variable. Of course, the frequency range and number of points must be customized to match your requirements. Although this shows the "open" measurement, the "short" and "load" are identical, except for the operator prompt string and the name of the local variable used to store results.

Last, the cal standards are combined and used to calculate an "error adapter" as shown below:



In the panel above, you can see we are performing a "Reset Cal" on the cal variable we created in the Device Interface. (Note the "owner" of the Reset Cal button is set to Device Interface so that our cal variable is visible.) Also, note the "Save Cal Data" button saves the results back to the Device Interface variable (again, note the "owner" on the button has been changed). The calculation block "Open - Short - Load" gathers the measurement data for all three standards and creates an array which is used in the next step. Next the button "Cal with Calibration Kit" uses the open, short, and load standards from a selected CalKit object along with the measured standards to compute a two-port Error Adapter object. The Spline button forms a cubic spline object for the calibration so it can be used at points other than the measured frequencies.

Using the Calibration in a Testplan

Using the calibration data in a testplan is straightforward. Note the example testplan panel below:



The Global Defaults panel was exactly the same as the calibration plan. The measurement technique is the same, wave parameters are measured at ports RF3 and RF6. The ratio is computed by the special calc button "Calc S11". In this case, we use the "Cal Factor" button (with the owner set to Dut Interface) to retrieve our calibration array. It is interpolated for the frequency of interest and sent to a standard two-input calc block with the calculation "error correction" selected.