Characterization Procedures

Characterization of the RF Source

These parameters must be characterized:

-         Power stability and band power

-         Frequency stability and frequency offset

-         Grid to camera angle alignment

-         Beamwidth (azimuth cut at elevation=0°)

To characterize the RF source, follow the following steps:

1.   Mount the RF source into the 3d printed purple mount (itself mounted on top of a rotation actuator), located on top of the pedestal, at the right side of the anechoic chamber.

2.  Install the frequency extender on the robot and connect it to the Keysight MXA 9020b (use the LO output of the frequency extender and the Ext. Mixer input in the MXA).

3.   Move the extender so that the waveguide is 70 cm away from the feed horn of the frequency extender (see M.Sc. Thesis Felipe Carrero for more information). With the help of the robot, move the actuators until you find the maximum power output (check rotations, both of the source and of the extender, and check vertical alignment of the extender with the RF source).

4.  Configure the frequency extender source to External Mixer, edit the tables of harmonics and set the first box to 24, enable Signal ID to remove harmonics generated by the frequency extender. Optional: use the Recall button and then the "Correction" option to add the frequency extender and cable losses to the values shown by the extender.

5.  Connect the RF source to the Astronomia network via Ethernet cable. The Pi is configured by default to the IP 192.168.0.150. Login via ssh -XY 192.168.0.150 and navigate to 150GHzsource/adf5355software/

6.  To characterize power stability and band power you must run RFsource_characterize_measbandpower_series.py from the RPI:

a.  You can edit this code to see the readme and change the output file name. A number of samples to take at each given frequency will be requested on start.

b.  This file will take the requested number of samples consecutively at each a given frequency before moving on to the next one, for each frequency in the frequency list.

7.   To characterize frequency stability you must run RFsource_characterize_measfreqstability.py from the RPI:

a.  You can edit this code to see the readme and change the output file name. A number of samples to take at each given frequency will be requested on start.

b.  This file will take the requested number of samples consecutively at each a given frequency before moving on to the next one, for each frequency in the frequency list.

8.   To measure the grid alignment you must place a vertical screen parallel to the wire grid and turn on the lasers (switch at the back). Take pictures of the diffraction patterns that appear on the screen. Make sure the camera is focused the distance at which the diffraction patterns appear on the screen. These pictures can later be analyzed with proper software (a MATLAB version has been implemented).

9.  To measure the azimuth pattern you need to run RFsource_characterize_measureAz.py from the Beaglebone board that control the robot in the anechoic chamber. Instructions to log into the board are pasted in a wall close to the chamber. You’ll find the script in the /mmwave-lab/Scripts/ folder of the Beaglebone. By default, this will move the actuator between -30 and 30° with 0.2° resolution and log an average of the measured power in the MXA to a text file. Confirm that the actuators move correctly prior to using this script: log into the beaglebone (run "robot" from the NUC computer in the lab) and then run "robot" again to enter the motion control program of the robot.

Check these results against prior characterization. In general, the power output of the RF source at 150 GHz should be of the order of -68 dBm at 70 cm without any corrections applied (mixer loss, free space path loss, extender horn gain) or ~ -57 dBm with mixer loss correction applied in the MXA (with the Recall function). The average power stability across the 130 to 160 GHz band should be of the order of 0.1 dB. The frequency stability across the band should be of about 1 ppm and the average frequency offset should be of around 5-10 kHz. The azimuth cut of the beam should be consistent with a simulation of an open-ended waveguide at 150 GHz, with a beamwidth of ~47° (-3 dB point).

For more information, check the thesis of Felipe Carrero or ask him directly.