This application note is meant to act as a guide to testing Radar sensors built using TI’s Radar chips. This document is expected to act as a high-level guide to help setup testing of radar sensor units during production. It captures generic requirements of radar testing; the actual test set up and software required to run the tests varies depending on the actual application of the radar. The user is expected to design the test software and determine the appropriate limits for any tests based on the application of the radar.
All trademarks are the property of their respective owners.
Described below are the basic production checks done on a radar module at high-level:
Sensor performance validation can cover the entire range of sensor hardware and software subsystems. This document is not intended to serve as a reference for hardware bring up or validation. Instead, this section describes a high-level overview of the production line testing done on fully validated and characterized hardware to screen for issues on a production ready sensor. These issues generally arise from handling defects, environmental factors, assembly defects, and manufacturing defects.
After the assembly, the sensor board is powered up. The boot-up current drawn from the supply is measured and made sure that it is within the limits. The test program is executed which takes the sensor board through the sequence of tests described in the following sections.
The Radar Test GUI is a program that controls the Radar chip and processes the data that the Sensor captures. It is a custom program developed by the user suiting their end application and test hardware interface available.
Described below are key factory calibrations that help improve the performance of the sensor by compensating for imperfections introduced by the manufacturing process.
A transmitter antenna designed to generate peak power at bore-sight may show the peak power tilted away from bore-sight in practice, thus the antenna beam pattern can exhibit a tilt (skew). The magnitude of tilt depends on the antenna and feed design. The tilt could change monotonically with the RF frequency, and the tilt could also change with variations in the PCB. For example, the pattern of glass fabric that is inherent in some of the PCB dielectric materials can interplay with the antenna transmission line geometry and can cause alteration in the radiation pattern, including beam tilt. This is especially true with low-cost PCB materials. Beam tilt is also more pronounced for a high-gain antenna.
The frequency-dependent tilt may be systematic and quantifiable, but the tilt due to PCB artifact may be somewhat random. Beam tilt on the final sensor can include effects of radome and casing. This beam tilt may be a parameter used as a pass/fail criterion or can be corrected in the mechanical mounting of the Radar sensor if required such that the beam tilt of the mounted sensor is effectively zero.
The beam tilt can happen for both transmit and receive antennas. It is typically measured for the required combination (or every combination) of Tx and Rx antennas by using the loopback measurement setup shown in Figure 2-2 explained in the next section.
Procedure: The beam tilt can be measured either in azimuth or in elevation. In this example, assume that the beam tilt is in elevation for the sensor that is being tested. The procedure involves stepping through the elevation angle and measuring the signal strength at each step.
The target (corner reflector) is kept at a distance so that it generates an IF of around 1 MHz. The chirp parameters are chosen appropriately. A narrow-band RF sweep is chosen that sweeps in the center of the band. Narrow band is chosen to minimize the frequency-dependent beam tilt.
Next, the elevation angle is swept. In this example, we use a sweep of 40 degrees. Using a turn-table, position the sensor at one end of the elevation angle sweep at +20 degrees. Use one of the transmitters for generating the chirp. ADC output from all receiver chains are simultaneously captured and processed to measure the signal strength. The signal amplitude from the receivers are averaged. This procedure is repeated by stepping through the elevation angle by some incremental value, such as 0.5 degrees, until the end of the elevation sweep at -20 degrees. The angle corresponding to the maximum received signal strength is the beam tilt for that transmitter.
The above procedure is repeated for each of the transmitters. The sequence may be re-arranged to optimize the procedure. For example, Tx 1-3 measurements can be done for each elevation position before changing the elevation position. This reduces the mismatch in measured beam-tilts between Tx. Sufficient averaging is done to improve the accuracy.