HI, Nathan:

From the Doxygen documentation (file:///C:/ti/mmwave_sdk_03_06_02_00-LTS/packages/ti/control/mmwavelink/docs/doxygen/html/structrl_profile_cfg__t.html), we should be able to program 1024 samples. 

Another concern is the radarCube memory limitation, you can copy/paste your configuration to TI sensing estimator, and click on AUTO_POPULATE button. 

For some reason, I can not see your attachment.

Best,

Zigang

Hey Zigang,

Thanks for your reply.

They are working to better understand the performance limitations of the devices processing abilities. Specifically, looking for clarity on the following:

1. Maximum data processing rates – they want to understand the upper bounds of operation and how they align with the devices processing capabilities.

2. Support for long chirp operation – they plan to work with chirps consisting of 1024 samples each, within relatively long frames (on the order of tens of milliseconds). In the first phase, they are only interested in range FFT processing. 

they find it quite difficult to determine from the current SDK/documentation what is feasible and what is not, especially in terms of chirp/frame duration and range FFT throughput.

We’d greatly appreciate any guidance or insights you can provide on these topics.

BR,

Nathan K

Hi, Nathan:

I was able to access their configuration file now.  Here is what sensing estimator tells  us.  The ramp time and ADC start time is all too short.  

They needs to play with it, and send a valid chirp configuration first.  The longer ramp time seems okay. 

channelCfg 1 1 0
adcCfg 2 1
adcbufCfg -1 0 1 1 1
profileCfg 0 60 3 3 90 0 0 10 1 1024 12500 0 0 36
chirpCfg 0 0 0 0 0 0 0 1
frameCfg 0 0 16 0 500 1 0 

By the way, they can read the app notes below to understand more about chirp design.

https://www.ti.com/lit/an/swra553a/swra553a.pdf?ts=1753107116967 

Best,
Zigang

Hey Zigang,

The app note refers to the following - MMWAVE-DFP Firmware | TI.com which doesn't support the AWR6843 device, what do you suggest?

BR,

Nathan K

Hey Nathan,

Zigang will be out for the next week, so some other engineers will help for the time being. Regarding the app note, it is applicable for AWR6843 as well.

Regards,

Kristien

Hey Kristien,

I was referring to the MMWAVE-DFP (MMWAVE-DFP Firmware | TI.com), do we have something similar for the AWR6843?

BR,

Nathan K

Hey Nathan,

The DFP is contained within the MMWAVE-SDK itself under <MMWAVE_INSTALL_DIR>/firmware. I will note this DFP package is sparse in terms of information, but the mmWave Radar Interface Control document linked under the MMWAVE-DFP TI.com page contains in-depth information on the firmware which is applicable for most Gen 1 and 2 devices, including the AWR6843. You can also directly download it through the link below:

https://dr-download.ti.com/software-development/firmware/MD-XRrFQlHOcy/06.04.01.00/mmWave-Radar-Interface-Control.pdf

Regards,

Kristien

Hey Kristien,

All the customer wants currently, is to know if they can configure a long chirp with 1024 or so samples and for the first stage, only range FFT -

Chirp -> FFT-> chirp -> FFT and so on. They would like to know the capabilities of the device and its limitations. They are, and me as well, are getting lost in all the collateral, can you please help us with this matter? 

BR,

Nathan 

Hey Nathan,

The chirp configuration provided with 1024 samples should work. As mentioned by Zigang, the ramp end time and ADC start time is invalid, so she had suggested bumping up the ramp end time to at least 90 us. This will decrease the maximum velocity, but improve the velocity resolution.

Regarding processing limitations, there should be enough memory and time to process the data via range FFT into a range profile. 

Regards,

Kristien

Hey Kriestein,

Thank you. Would you be able to share a configuration that demonstrates that?

BR,

Nathan

Hey Nathan,

Zigang had shared a modification to the profileCfg command for the DEBUG.cfg you had provided. The front-end commands are given below, but the only change is the rampEndTime.

channelCfg 1 1 0
adcCfg 2 1
adcbufCfg -1 0 1 1 1
profileCfg 0 60 3 3 90 0 0 10 1 1024 12500 0 0 36
chirpCfg 0 0 0 0 0 0 0 1
frameCfg 0 0 16 0 500 1 0 

Regards,

Kristien