IWRL6432BOOST: Cannot get Out-of-Box IWRL6432 demo working - cannot upload configuration

Hello, 

Can you let me know the path of the prebuilt binary file that you flashed?

Thanks and regards,

Josh

Hi Josh, yes: 

C:\ti\MMWAVE_L_SDK_05_04_00_01\examples\mmw_demo\motion_and_presence_detection\prebuilt_binaries\xwrL64xx\motion_and_presence_detection_demo.release.appimage

I thought it was a bit strange that Uniflash wanted a .bin file, but the .appimage seemed to upload OK

I also tried the debug version too, but same result - no output. I checked opening up the APP COM port, however I don't see any output when the board is reset, unlike what I see when running the IWR6843 people tracking demos - on that board I get a CLI through the serial port, and I was assuming the IWRL6432 would be similar. The board flashes OK, which leads me to believe the COM port is working.

Just to check my understanding, as per EVM Setup Operational Modes, there really is only one DIP switch that needs to change for flashing: that is S1-1, which needs to be OFF for flashing and ON for running. The other DIP switches can remain the same I'm assuming?

Thanks Josh - I'll check when I get back to the lab. If it is ES1.0, that's strange as I only ordered it a couple of weeks ago; but then do I understand you right - the next silicon isn't shipping on the Boost boards yet?

Thanks Josh - ordering one now!

Thanks Josh - board received, I've been able to program it (with "motion_and_presence_detection_demo.release.appimage"), and I now get a little further. I see point clouds momentarily on the Industrial Visualizer, but then it seems to hang. After a few divide-by-zero errors, I see lots of the following:

Traceback (most recent call last):
File "gui_main.py", line 2043, in updateGraph
sumOfTags = sum(self.classifierTags[trackID])
AttributeError: 'Window' object has no attribute 'classifierTags'

Then, a few:

Traceback (most recent call last):
File "gui_main.py", line 2034, in updateGraph
if(classifierOutput[trackNum][0] != 0.5 and not(self.wasTargetHuman[trackID] == 1 and abs(trackVelocity)<MIN_CLASSIFICATION_VELOCITY)):
AttributeError: 'Window' object has no attribute 'wasTargetHuman'

Then, more of the first errors, and finally:

ERROR: No data detected on COM Port, read timed out
Be sure that the device is in the proper mode, and that the cfg you are sending is valid

Sorry if I'm doing something dim here, is the Industrial Visualizer in the Radar Toolbox still the right thing to use? Perhaps I need a newer version that can handle the human classification data?

Sorry, I should add that I'm running the Industrial Visualizer from what I think is the latest Radar Toolbox (radar_toolbox_1_30_01_03)

Hello, 

You are using the correct GUI and that version of the Radar Toolbox should be compatible with the classification data so this behavior is unexpected. Please give me 24 hours to look into this and see if the issue is replicable on my end. I will get back to you tomorrow with more information. 

Can you please also let me know which configuration file you have selected when starting the demo?

Best Regards,

Josh

Thanks for the quick reply, Josh - I've been using "IWRL_Classification_MidBw", but I've also tried "IWRL_Classification_HiBw_4_ants" and "IWRL_Classification_MidBw_Sensitive" (from C:\ti\radar_toolbox_1_30_01_03\source\ti\examples\Classification\human_non-human_classification\chirp_configs)