AWR2243: mmwave cascade radar evm

Hi Cherry,

From your question, it seems that the customer is trying to calibrate the cascade device. Have they followed the user guide? Are they facing any issue here?

Is this board calibrated for the radar parameter 64chirp only? -> No. This is a sample lua script that can generate data which can be used for calibration on a particular EVM. Using the data collected, a new calibration matrix is generated which will be used for post processing there on.

I am unable to understand your question regarding the impact of changing the number of chirps in a capture. Could you please explain it in terms of input, expected output and obtained output?

Regards,

Kaushik

Hi Kaushik,

Thanks for your support.

Customer uses 2243 board TXBF mode to capture data, radar sets parameters to chirp number 128 before acquisition, and the official sample code cascade_TxBF_signalProcessing.m file is used to process the data after successful acquisition. In this file for radar data, there is %load MIMO calbibration params calibrationFilePath = '.\PostProc\calibrateResults_high.mat';

The processing results are imaged as follows:

Again in TXBF mode, when the customer changes the chirp number to 64 to capture radar data, the official code cascade_TxBF_signalProcessing.m is still used, and the radar processing is the same as in reality. It works properly as shown in the following figure: 

So the customer suspects that the calibration file calibrateResults_high.mat is an issue. When radar captures data in TXBF mode, the parameter 128chirp will be abnormal, and the parameter 64chirp will handle it properly. (All the above radar data was acquired with different chirp parameters, all other parameters are the same) 

Thanks and regards,

Cherry

Hi Cherry,

Has the customer generated a new calibration file following the provided guidelines? If the customer suspects that it is due to the calibration file, you can regenerate a new one and try it with that.

Alternatively, you can check the same by disabling the calibration step and checking the output. This way we can debug along the right lines.

Regards,

Kaushik

Hi Kaushik,

Kaushik Gowda said:

Has the customer generated a new calibration file following the provided guidelines?

They just followed our guidelines for the calibration. 

Customers have used the official calibration files that come with the software, and the resulting calibration files that they have calibrated themselves, and both have had problems.

And they look up and find that the number of chirp multiplied by the radar angle range is greater than a specific value, such as 2048, the problem occurs. Is this likely the cause of the above issue? 

Thanks and regards,

Cherry

Hi Cherry,

Just to be clear on the chirp no. wrt device configuration

As per last image: total no. of chirp in a frame is calculated = (end_chirp_tx - start_chirp_tx) * nChirp_loops.

Now with this request customer to correlate their 128 chirp no. with which they are facing issue.

And on the other hand, I would suggest customer to try processing captured data with their own Matlab script to eliminate dependency which may have some known/unknown issue/limitation.

Regards,

Jitendra