IWR6843ISK-ODS: Area scanner demo has limited range

Hi Tom,

Any reason you are not using gui visualizer from 4.2.0 industrial tool box.

Thanks

Yogesh

Hi Yogesh,

I need a 3D point cloud GUI with ideally ~10m of range, but at least more than the 3m I'm getting from 3.6.2 Area Scanner. The reason I'm not using 4.2.0 is because I haven't found any GUI in 4.2.0 that gives a 3D point cloud.

If you know of an 3D point cloud GUI in 4.2.0 I'm interested to hear about it, otherwise I need to figure out how to extend the range on the 3.6.2 GUI.

Thanks,

Tom

Hi Tom,

The gui visualizer at C:\ti\mmwave_industrial_toolbox_4_2_0\labs\area_scanner\68xx_area_scanner\gui is 3-D visualizer. Let me know if you are having issue running the same.

Thanks

Yogesh

Hello Tom,

You can try adjusting the CFAR threshold in your configuration file. If you are using the default cfg provided in the ODS out of box lab the default threshold is set at 15dB. You can try lowering it until you see the detection points you expect. Note that if there is a lot of clutter in your scene the other objects may be stronger than a person standing still - making detection difficult. 

You can try the new area scanner lab as suggested by Yoges in the 4.2 toolbox which does background subtraction to make it easier to detect addition of new static objects to the scene.

Amanda

Hello Amanda and Yogesh,

I tried both of your solutions, and neither achieved what I need.

I couldn't get the Area Scanner demo from 4.2.0 to run. I tried two different chirp configs and both gave an error before the visualizer GUI was launched. The error with the chirp config in the out of box demo, \mmwave_industrial_toolbox_4_2_0__win\mmwave_industrial_toolbox_4_2_0\labs\out_of_box_demo\68xx_ods_point_cloud_demo\chirp_configs, is given in this log:

COM4 opened.
COM3 opened.
Opening configuration file
Parsing configuration file...
COM4 opened.
COM3 opened.
Sending cfg file to device...
sensorStop
flushCfg
Done
dfeDataOutputMode 1
Done
channelCfg 15 7 0
Done
adcCfg 2 1
Done
adcbufCfg -1 0 1 1 1
Done
lowPower 0 0
Done
profileCfg 0 60 7 3 24 0 0 156 1 256 12500 0 0 30
Done
chirpCfg 0 0 0 0 0 0 0 1
Done
chirpCfg 1 1 0 0 0 0 0 2
Done
chirpCfg 2 2 0 0 0 0 0 4
Done
frameCfg 0 2 32 0 100 1 0
Done
guiMonitor -1 1 1 1 0 0 1
Done
cfarCfg -1 0 2 8 4 3 0 15.0 0
Done
cfarCfg -1 1 0 4 2 3 1 15.0 0
Done
multiObjBeamForming -1 1 0.5
Done
calibDcRangeSig -1 0 -5 8 256
Done
clutterRemoval -1 0
Done
compRangeBiasAndRxChanPhase 0.0 -1 0 1 0 1 0 -1 0 -1 0 1 0 1 0 -1 0 -1 0 1 0 1 0 -1 0
Done
measureRangeBiasAndRxChanPhase 0 1. 0.2
Done
aoaFovCfg -1 -90 90 -90 90
Done
cfarFovCfg -1 0 0.25 15
Done
cfarFovCfg -1 1 -13.39 13.39
Done
extendedMaxVelocity -1 0
Done
CQRxSatMonitor 0 3 4 63 0
Done
CQSigImgMonitor 0 127 4
Done
analogMonitor 0 0
Done
lvdsStreamCfg -1 0 0 0
Done
sensorStart
Index exceeds the number of array elements (2944).

Error in getTLV (line 62)

Error in parseBytes_AS (line 102)

Error in area_scanner_visualizer (line 258)

MATLAB:badsubscript

The other chirp config I tried was the default area scanner config. It gave an error that it doesn't recognize the cli bpm command.

With the 3.6.2 Visualizer I tried modifying the CFAR function in the config. I tried 5 and 10 dB. At these values it detected me at slightly farther distances, with more points detected, but once I was past 5m neither of them saw me. I also tried 0dB but with that value the demo froze. Even with holding a metal sheet in front of myself I couldn't get past 5m. Do you think that the range is limited by the SNR? The radar is mounted at 1.35m above a concrete floor in a large room with no clutter nearby. Have you ever gotten this radar to detect 3D points beyond 5m?

I would appreciate if you can advise me on either getting the 3.6.2 or the 4.2.0 demo working to detect me at a farther distance. Thanks for your help,

Tom

Tom - the binaries, cfg files, and GUIs are not necessarily interchangeable. You should be using the ones that correlate with each other.

Is your sensor mounted overhead? It so this is likely why you are not seeing further distances.

No, my sensor is mounted perpendicular to the ground facing forward.

I tried using the 4.2.0 GUI with the 4.2.0 cfg file. But I got the error that bpm was not recognized by the cli.

To summarize so far, I haven't been able to run the 4.2.0 area scanner GUI with any chirp config since it gives me an error. I have been able to run the 3.6.2 area scanner GUI with the 4.2.0 out of box chirp config, but my range is too short even after changing the CFAR threshold.

Tom - please ensure you are flashing the 4.2.0 area scanner lab binary on an ES2.0 ODS board and using the default config with the GUI provided as outlined in the quickstart guide for the lab. The binary/cfg/GUI combination has been verified to work on an ES2.0 ODS board - it will not work on ES1.0

You should also double check that whatever config you are using does not have static clutter removal enabled since this will prevent you from getting detections from still objects.

Hi Amanda,

You were right about the 4.2.0 Area Scanner: I was using the wrong binary. But after working extensively with this setup and reading documentation, I still can't get static detection past 3.5m. My requirement is 5m. What do you think is the maximum possible range? The only other example I have of this demo is the video in the user's guide, where it's not clear to me if it statically detects beyond 3.5m.

My setup is:

- 4.2.0 Area Scanner GUI

-4.2.0 Area Scanner ODS prebuilt binary

-4.2.0 Area Scanner ODS chirp config

- MMWAVEICBOOST

- IWR6843ISK-ODS.

My demo settings are:

I tried changing several parameters in the CLI commands in the config file.

Like you suggested, I first tried changing the SNR threshold in the cfarCfg command. According to the SDK documentation, the second parameter in cfarCfg indicates if it is for the range or the doppler direction. I changed the threshold down to 5dB for the range direction (procDirection=0), and this did not work. The static detection still couldn't detect anything past 3.5m, however the dynamic points became a lot more common. (I also tried lowering the threshold for the Doppler direction, procDirection=1, but that didn't help either)

Another parameter I tried changing was in clutter removal. I changed the second parameter from 1 to 0, to turn off clutter removal, and this didn't change anything either.

Next I tried playing with the parameters in staticDetectionCfg. I couldn't find documentation for this command so I had to guess based on the comments in the function MmwDemo_staticDetectionConfig in the file mmw_cli.c, which I assume is the parser for this cli. I tried raising and lowering the local peak threshold, heatmap different threshold, and significant threshold. Regardless of what I did, the static detection range did not pass 3.5m.

I attached a picture of my setup, in case you can identify anything that would be affecting my range.

Thank you for your help,

Tom