Calibration of Vector Network Analyzer for Measurements in Radio Frequency Propagation Channels

Making good measurements is difficult. This procedure will help develop good measurement practices. When this protocol is followed, the VNAs will debug and characterize RF devices and systems with precision.

VNAs are the most accurate way to measure an RF device or system. The theory behind this technique is well-established and following this protocol will provide consistent and repeatable results. Demonstrating the procedure will be Savio Tran, a colleague from my laboratory.

Before initiating a measurement, gather all of the components of the propagation measurement system including the cables, amplifiers, filters, devices under test, and other relevant components. Note that female standards which have an insertable port for the male port on the cable will be used in this calibration. At least a half an hour before the measurement, turn on the VNA and press the preset button.

Visually inspect all of the cables and connectors for obvious signs of wear such as nicks, dents, and imperfect connector threads and clean the connectors on all of the devices and cable ends. Before their connection, align the VNA cables with ports one and two and the device under test and tighten the connected cables by hand. If little to no resistance is felt, tighten with a 12 inch pounds torque wrench for type N connections.

To set the VNA frequency range, open the stimulus menu and select frequency. Set the start frequency to 1, 700 megahertz and the stop frequency to 1, 900 megahertz. To set the measurement type, select response, measure, and S21.

To set and adjust the port power, select stimulus, power, power, and adjust the port power to zero dbm. To set and adjust the sweep settings, select stimulus, sweep, and sweep setup, and stepped sweep. Set dwell time to zero microseconds.

To set and adjust averaging mode, select response and average and set the intermediate frequency bandwidth to one kilohertz. To set the display data format, select response, format, and log mag. Then select stimulus, sweep, number of points, and 1601 to set the number of data points in the displayed trace.

To perform a manual calibration, select response, cal, start calibration, calibration wizard, and unguided calibration. Select two port short open load thru under cal type selection and select the appropriate calibration kit so that an accurate value of the standards in the specific calibration kit are known. Then click next.

Attach an open calibration standard to the cable attached to port one and attach a short calibration standard to the cable attached to port two. To perform a measurement of the attached open, on port one, select open and type N50 female open. A trace will appear.

To perform a measurement of the attached short on port two, select short and type N50 female short. A trace will appear. After the second measurement, swap the calibration standards between the ports.

And on port one, select short and type N50 female short to measure the short on port one. A trace will appear. To measure the open on port two, on port two, select open and type N50 female open.

Next, replace the short from port one with a broadband load. And on port one, select loads and type N50 broadband load to measure the load on port one. A trace will appear.

Once the measurement has completed, remove the open from port two and the broadband load from port one and transfer the broadband load from port one to port two. Place the open from port two on port one to prevent leakage signals. And on port two, select loads and type N50 broadband load.

A trace will appear. Insert a thru calibration standard between the cables attached to ports one and two and select thru to measure the thru calibration standard. Multiple traces will appear.

Once all of the standards have a check mark above them, select next and save as user CalSet. Then enter a name for the calibration and click save. To check the calibration, connect a thru adapter without obvious signs of wear to the cables between ports one and two and select response, scale, scale, and set the per division to 0.1 decibel.

To measure the insertion loss of the thru, select the stimulus menu, trigger, and single. A single sweep will appear across the frequency range. Once the thru has been checked, select response, scale, and scale to return the per division value to 10 decibels.

To measure S11, select stimulus menu, trigger, and single, and select response, measure, and S11. To measure components and system losses, first connect the device under test to the VNA and select response, measure, and S21. To measure the device under test, select stimulus menu, trigger, and single, and select file and save data as to save the data as a displayed traces displayed format trace file.

Set the scope to displayed traces, the format to displayed format, and choose a feel type of either CSV or s2p which saves the two port S parameters. To check and analyze the test results of a bandpass filter, select marker analysis, marker search, and max to find the insertion loss of the filter. Select marker analysis, marker, marker, and set the stimulus for marker one to 1, 802 megahertz to set the center frequency of the filter being measured.

From the dropdown menu, select marker two and check the on box, moving the knob to lower the marker frequency to the point that is three decibels below the measured loss at the center frequency. Select reference and check the on box. Then select marker three, check the on box and check the delta marker box.

Then set the stimulus for marker three to the point at which the delta three is closest to zero. The filter's three decibel bandwidth will be displayed by delta three. If a manual calibration is not selected, select the electronic calibration option and attach the electronic calibration kit to the cables between ports one and two.

Select response, calibration, start calibration, calibration wizard, and use electronic calibration. Select two port electronic calibration and the appropriate module and characterization and click next. Then click measure to measure the calibration standards.

At the end of the calibration, click save as user CalSet to save the calibration data. In this representative analysis, the filter was measured after consulting the manufacturer's specifications. As observed, the insertion loss was identified as well as the three decibel points.

Here is a measurement of a damaged 10 decibel attenuator that was measured to be minus 22.70 decibels. For comparison, in this figure, the measurement of a good 10 decibel attenuator can be observed. Note that the measured value is within the specified range of minus 9.5 to minus 10.5 decibels across the entire 1, 700 to 1, 900 megahertz bandwidth.

Cable loss is another important measurement frequently performed in radiofrequency measurements. For example, in this analysis, the measured loss was minus 1.88 decibels at a frequency of 1, 750 megahertz. Be sure to always check that you have a good calibration before continuing to take measurements and to verify the measure parameters with the specification sheet after measuring a device.

Using good measurement practices with a VNA on the bench top translates into trustworthy, accurate measurements on other devices and in complex environments.