AWR2243BOOST: Outputs for a AWR2243BOOST + DCA1000EVM for real time evaluation

Hi,

For real time evaluation, we recommend you use one of our programmable sensor platforms such as AWR1843BOOST board using the mmWave SDK demo.

DCA1000 is a capture card that allows you to capture raw data from a sensor board such as AWR2243BOOST.

The captured raw ADC data can be post processed using matlab included in the mmwave studio setup.

It is not possible to stream the data in real time. However you should be able to capture long durations of raw data that can be used for algorithm development.

Thank you

Cesar

Hi Cesar,

Thanks for the reply. I am planing to use the "AWR2243BOOST" this is not a programmable board for the AWR224 ?

Sorry i do not understand the sentence  "DCA1000 is a capture card that allows you to capture raw data from a sensor" , this means that with the "AWR2243BOOST i am not able to get the raw data? Moreover,  from the web page (https://www.ti.com/tool/DCA1000EVM) it says "Real-time data-capture adapter".

My questions were more in the direction of "what do you mean by "raw data" ? can i have access to information such as distance, position or velocity?

The only way to post-processes the data is using matlab? what should i need to buy/do in order to be able to posr-process the data using python and or c++?

Many thanks in advance,

Victor

Hi,

Can you please provide more information on your project?

The raw data is the ADC data. This data must be processed in order to compute range, velocity, angle...

You can develop code to process this data in any high level language such as python or C++.

The matlab postprocessing capabilities are provided with mmWave studio for ease of use.

DCA1000 enables to capture raw ADC data real time. However it does not allow streaming this data real time. It only allows capturing the data to a file on the PC.

thank you
Cesar

Hi Cesar,

Cesar said:

Can you please provide more information on your project?

We are mainly interested on including a Radar on a drone to detect the distance to non-desired objects on the flight path of the drone. The project is in the early stages, so we are just trying to understand which kind of products from texas instrument can be useful for our purpose and until which level we should go. In other words, should we stay on the AWR2243? or do we need to go for the board AWR2243BOOST with the DCA1000 ?

An important point to understand is the difference between the different elements such as the AWR2243 or the AWR1843. It is clear that the bandwidth and the sampling rate is different. But for example  I see that the AWR1843 has some processing elements such as "radar accelerator" while the AWR2243  has not these elements. This means that we need additional processing component for the AWR2243?

Cesar said:

You can develop code to process this data in any high level language such as python or C++.

The matlab postprocessing capabilities are provided with mmWave studio for ease of use.

What is exactly the level of "matlab postprocessing capabilities" ? Do they provide the level of data such as range/velocity/angle? Is the source code of matlab available? 

Cesar said:

DCA1000 enables to capture raw ADC data real time. However it does not allow streaming this data real time. It only allows capturing the data to a file on the PC

Thanks for this information. Do I need then some external processor (from texas) to be able to stream the data in real time?

Many thanks for your help,

Regards,

Victor

Hi,

I think it is very important for you to understand that AWR2243 is a sensor only device. It does not have any processing. With AWR2243 the raw ADC data is sent through CSI2 to a host processor that should perform the processing.

This is the reason why I don't recommend using the AWR2243 for this application.

I recommend using a sensor that has integrated processing such as AWR1843.

Before we discuss further the raw data capture it is important for you to understand the differences between AWR2243 and AWR1843

thank you
Cesar

Hi Cesar,

Thanks for the clarification. Now i understand that the AWR1843 has some processing capabilities which as you said i guess this makes it better for our application. However, has the AWR1843 the possibility to join some of them in a cascade to improve angular resolution as in the case of AWR2243? I see in the web page that for AWR2243 is written "Cascade capable" but not for the AWR1843.

Following then the pervious questions...which type of output data will provide the AWR1843 respect to the AWR2243 ? Which kind of processing is done in the  AWR1843? 

What is exactly the level of "matlab postprocessing capabilities" ? Do they provide the level of data such as range/velocity/angle? Is the source code of matlab available? 

Cesar said:

DCA1000 enables to capture raw ADC data real time. However it does not allow streaming this data real time. It only allows capturing the data to a file on the PC.

Regarding the real time Do I need then some external processor (from texas) to be able to stream the data in real time?

Many thanks,

Regards,

Victor

Hi,

You are correct, the AWR1843 can't be used in a cascade system. If you are interested in a cascaded system, you would probably need to work with a third party to develop your solution. Cascaded systems are highly complex.

Regarding AWR1843, we provide a mmWave SDK that includes a great demo that runs on one of our Evaluation Boards.

Please download the SDK and review the documentation

https://www.ti.com/tool/MMWAVE-SDK

After downloading the SDK you can review the documentation

mmWave SDK Module Documentation

The demo provides as output the (x,y,z,velocity) for each detected point. This output is streamed in real time through the UART to a host PC.

Thank you

Cesar

Hi Cesar,

Thanks for your reply and the link. I´ve downloaded all the documentation.

Please correct me if i am wrong, then the AWR1843 is able to give directly (x,y,z,velocity) for each detected point as well as the range profile?, while the  AWR2243 only provides the raw data from the ADC?. Is this correct?

Is there some documentation to be able to process the raw data to get (at least) range profiles ?

Many thanks,

Victor