Hello Team,
I would like to ask TI experts about the preliminary test results (maximum detectable range) and related cfg settings.
The summary of test results and questions are as below:
* Device & condition
- Board : AWRL6432Boost - Visualizer ; V5.4 - Flashed image : MPD demo prebuilt image (SDK 5.4.0.1)
* Cfg : 3 kinds
(Common) We had to use 1Tx - 3Rx antenna configuration & SIMO.
- channelCfg 7 1 0 (use only Tx 1)
- chirpComnCfg 8 0 0 256 0 28 0 (use SIMO)
- frameCfg 64 0 2300 1 250 0 (Normal mode frame)
(Split) We adjusted max range by simply lowering chirp slope.
- chirpTimingCfg 6 63 0 75 59 (estimated max range of 10m)
- chirpTimingCfg 6 63 0 50 59 (15m)
- chirpTimingCfg 6 63 0 36 59 (20m)
(Cf) Other items are almost identical to those of mmwave demo cfg (MotionDetect.cfg in SDK 5.4.0.1).
* Test method & brief results :
- A person walks up to 20m back and forth in front of the radar.
- The TLV data was recorded by using visualizer (V5.4), then we parsed range, and extracted the maximum value.
- The max range was characterized as 7.5 ~ 8m for 3 cfg cases (without any tendency).
* Questions :
(1) Short max range
- Even though the estimated max range was 10, 15, and 20m, the real range was only 7.5 ~ 8m.
Is this because of using only one Tx antenna and SIMO instead of MIMO (BPM)? Or was there any lack of cfg settings above?
(2) Max range is not proportional to the estimated for the 3 cfg settings (No tendency)
- According to tuning guidelines in the toolbox (V2.0), increasing of # ADC samples and ramping time is also described for longer range.
Do we need to adjust them other than the chirp slope?
(3) Others
- Actually, we are searching the application of tracking human AND vehicles using the radar.
We found the 'Exterior_Intrusion_Monitoring' application and related cfg in the toolbox (V2.0), but this seems to be optimized for tracking walking pedestrian (or possible intruders).
Is there any efficient cfg settings for tracking nearby cars as well as pedestrians?
