IWR6843ISK: z-coordinates for detected points are always 0.0

Hey Cagri,

Looking at your channelCfg line in your config file, it seems like you are not utilizing all of your tx antennas. If you put 7 in txChannelEn field in the channelCfg line (and add another chirpCfg line respectively), does that enable z value data to come off the sensor?

Regards,
Luke

Hi Luke,

I see what you are saying. I think there was a mismatch between what I was seeing on the mmWave visualizer configuration screen and the *cfg file that was being actively used.

I have attached the corrections you have suggested to the new *.cfg. If this looks good to you, I will go ahead and test it.

Thanks for your swift response to this.

Regards,
Cagri

3TX4RX_active_config.cfg

Hi Luke,

I used the *.cfg file with 3 TX and 4 RX antennas active and captured the attached data on mmWave visualizer.

Even though I get the data displayed as 3D scatter plot as expected, The z-component of all the detected points in the cloud are always zero (as I indicated with the mouse pointer in the attached screenshot)

I am not sure if there is another configuration parameter I need to configure differently, Please advise.

Thank you for your help.

Regards...

Hello Cagri,

I am going to see if I can replicate your findings using your cfg file.

Regards,
Luke

Thanks Luke. That will be very helpful.

One more thing (which I assume you have noticed as well), the color coding I have used in terms of elevation can be clearly seen in the points the 3D scatter plot (above). However, all the points are mapped to z = 0 coordinate for some reason. I have decoded the TLVs for a sample capture and confirmed that the z coordinates are indeed mapped to 0.0. So the elevation information (indicated by changing colors) seems to be lost somewhere for some reason.

Regards...

Hi Luke,

I also tried one more thing a few minutes ago, which I think will help you shed light onto what the issue is.

I remember a while back I did manage to capture non-zero z-axis data using mmWave Visualizer, which used a narrower bandwidth (i.e., 700 MHz) centered at 62 GHz. I am attaching the *.cfg file I used for that old data capture.

The corresponding 3D scatter plot for one sample capture is include below. As you can see the color coded elevation information for each captured point is accompanied by non-zero z coordinates as one would expect.

Therefore, it seems the z coordinates in the case of full bandwidth point captures seem to be the culprit. This is just a guess based on what the data is showing and there may be another root cause for this of course. I wanted to share this with the hope that it will help with your investigation.

Regards...

RadarConfig_700MHz_BW_centered@62GHz.cfg

Okay, thank you for providing more context, I will use that config as well. 

I am very curious to see what the cause is here. I will be getting back to you within a few hours. 

Thanks Cagri,
Luke 

Hi Luke,

Thanks for looking into this. That's great news.

It sounds like using the Industrial visualizer moving forward will make sense.

Will I need to change any firmware on my antenna board to be able to use that or is the industrial visualizer fully backward compatible with version 3.5+ SDKs?

Regards...

Hi Cagri, 

I would recommend upgrading to 3.6 if possible. This is an LTS SDK version so you would not have to worry about changing firmware again. However, I believe it should work with 3.5, though I am not certain. 

Regards,
Luke

Sounds good. I will try 3.5 first and upgrade to 3.6 if necessary. I will be in touch in case of any issues.

Thanks for the guidance Luke.

Regards,,,