AWRL1432BOOST: ADC HIL

Hi,

We will get back to you as soon as possible. By Friday the latest.

thank you

Cesar

Hi Liam, 

If I gathered correctly, you captured raw ADC data using mmwavestudio+DCA1000, and then you tried to feed that data back to the demo using CCS and tried to step through the code? Is that correct?

Additionally, what demo are you trying to feed it to?

If you are running your demo in CCS to load the ADC data please be sure to do the following before you attempt to load ADC data:

1. flash C:\ti\MMWAVE_L_SDK_05_03_02_00\tools\Ate_Calibration\xwrL14xx\calibration_example.release.appimage onto your device and hard reset your device by reconnecting it  

2. flash C:\ti\MMWAVE_L_SDK_05_03_02_00\examples\empty\xwrL14xx-evm\m4fss0-0_freertos\ti-arm-clang\empty.debug.appimage and hard reset your device by reconnecting it. Be sure to change the pin to functional mode before reconnecting it. 

3. Continue to run what you were running previously in CCS

Hi,

Thank you for your reply. I am still working on it.

1. I am using mmwave_demo with 1432 board. After I flashing Ate calibration and empty images, it didn't occur an error. But I still didn't see the result, like cfar result, point cloud.

2. Besides, I also want to save ADCBuffer in CCS memory. Here I found adc data format is real sampling data in uint32_t type with 12 valid bits. 

If now I have 256 ADC samples, 2 Tx, 3Rx, 64 doppler bins, how should I choose the size when saving memory. Should I choose a TI style, uint32_t with 256*2*3*64 size? And then read only 12bits in matlab.

3. And whether I can use the adc data parsing scripts in  C:\ti\mmwave_studio_04_01_00_06\mmWaveStudio\Scripts\MatlabExamples\matlabParsingScript\dataPostProc_lowPowerDevice.m to parse the memory ADCBufData I save with ccs without any changes in reshaping raw data?

4. A comment in above script saying the bits are Swizzling Mode. I didn't find document describing the Swizzling Mode. Could you provide me with a path?

Regards

Liam

Hi Liam, 

Glad to know that the factory calibration error is sorted. Now we can move on to the next ones

1. Are you loading the config after running the demo? Did you disable low power cfg in your profile to 0? If not the device will enter sleep mode and sever connection with CCS. Did you also enable adcDataSource 0 adc_data_name.bin to 1 in the cfg along with the correct file name and path?  What visualizer are you using? Did you check that the visualizer has the correct COM port set up?

When the visualizer is getting processed adc data from the adc bin file vs from the radar in real life the point cloud can be seen but very slowly. Make sure your grid is large enough to see the points at the respective ranges. 

2. If you look at the image you shared the data format for ADCBufData is DPIF_DATAFORMAT_REAL16  where #define DPIF_DATAFORMAT_REAL16 (uint32_t), so choose that data type when saving and loading data. If you want to see the size of ADCBufData, you can load in the expressions window to see its size.

For 3 and 4 I will have to get back to you by the end of the week. Thankyou!

Please also note that in order to see the heatmap you need to change in needed bit in guiMonitor cfg in the profile 

Hi,

Thank you for all detailed suggestions. I think question 1 has been solved. 

Regards

Liam

3. I believe you should be able to use the script to parse raw data from CCS. The script is intended to be used with the package provided in mmwave studio 4 1 00 06, any modifications or reuse is not supported by us. 

Thank you for your reply.

Regards

Liam