Number of chirp loop per frame

Yeshaswini Palaksha

Other Parts Discussed in Thread: AWR1843, AWR1843BOOST, DCA1000EVM

Hello,

I am using serial interface to configure the mmwave RADAR AWR1843 Boost board.

In this regard, when I try to set the number of loops parameter of 'frameCfg' to greater than 64, the RADAR board never accepts.

Is there any restriction for setting the number of loops?

Thanks & Best Regards,

Yeshaswini

over 4 years ago

0 Cesar over 4 years ago

TI__Guru*** 149434 points

Hi,

The firmware supports a value from 1 to 255.

Are you using mmWave Studio to configure the sensor. There are some limitations in the mmWave Studio.

thank you

Cesar

0 Yeshaswini Palaksha over 4 years ago in reply to Cesar

Prodigy 240 points

Hello Cesar,

Nope I am not using mmwave Studio. I am using Serial interface to configure the RADRA board.

I am able to configure the board with maximum of 64 chirp loops per Frame. If I try to greater than that, configuration fails, and RADRA board never starts sending chirp.

I have made sure the active chirp time of the Frame is less than 50% of the Frame periodicity.

I am really Looking Forward to find the reason for this. If you Need any other Information About the ocnfiguration, please let me know.

Thanks & Best Reagrds,

Yeshaswini

0 Kyle Cousino over 4 years ago in reply to Yeshaswini Palaksha

TI__Mastermind 24955 points

Yeshaswini,

Can you share the configuration file (*.cfg) that you are attempting to send to your AWR1843BOOST EVM?

Regards,
Kyle

0 Yeshaswini Palaksha over 4 years ago in reply to Kyle Cousino

Prodigy 240 points

Fullscreen 6403.config.txt Download

sensorStop
flushCfg
dfeDataOutputMode 1
channelCfg 15 1 0 
adcCfg 2 1
adcbufCfg -1 0 1 1 1
lowPower 0 0
profileCfg 0 77 7.0 7.0 38.0 0 0 100.09 1 256 7200 0 0 30
chirpCfg 0 0 0 0 0 0 0  1
frameCfg 0 0 80 0 50 1 0
guiMonitor -1 0 0 0 0 0 0
cfarCfg -1 0 2 8 4 3 0 15.0 0
cfarCfg -1 1 0 4 2 3 1 15.0 0
multiObjBeamForming -1 1 0.5
calibDcRangeSig -1 0 -5 8 256
clutterRemoval -1 0
compRangeBiasAndRxChanPhase 0.0  1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0
measureRangeBiasAndRxChanPhase 0 1.0 0.2
aoaFovCfg -1 -90 90 -90 90
cfarFovCfg -1 0 0.25 8.64
cfarFovCfg -1 1 -10.59 10.59
extendedMaxVelocity -1 0
CQRxSatMonitor 0 3 11 127 0
CQSigImgMonitor 0 127 8
analogMonitor 0 0
lvdsStreamCfg -1 0 1 0
sensorStart

Hello Kyle,

Please find the attached config file.

Best Regards,

Yeshaswini

0 Yeshaswini Palaksha over 4 years ago in reply to Kyle Cousino

Prodigy 240 points

Fullscreen 1768.config.txt Download

sensorStop
flushCfg
dfeDataOutputMode 1
channelCfg 15 1 0 
adcCfg 2 1
adcbufCfg -1 0 1 1 1
lowPower 0 0
profileCfg 0 77 7.0 7.0 38.0 0 0 100.09 1 256 7200 0 0 30
chirpCfg 0 0 0 0 0 0 0  1
frameCfg 0 0 80 0 50 1 0
guiMonitor -1 0 0 0 0 0 0
cfarCfg -1 0 2 8 4 3 0 15.0 0
cfarCfg -1 1 0 4 2 3 1 15.0 0
multiObjBeamForming -1 1 0.5
calibDcRangeSig -1 0 -5 8 256
clutterRemoval -1 0
compRangeBiasAndRxChanPhase 0.0  1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 0.0
measureRangeBiasAndRxChanPhase 0 1.0 0.2
aoaFovCfg -1 -90 90 -90 90
cfarFovCfg -1 0 0.25 8.64
cfarFovCfg -1 1 -10.59 10.59
extendedMaxVelocity -1 0
CQRxSatMonitor 0 3 11 127 0
CQSigImgMonitor 0 127 8
analogMonitor 0 0
lvdsStreamCfg -1 0 1 0
sensorStart

Hello Kyle,

Please find the attached config file.

Best Regards,

Yeshaswini

0 Kyle Cousino over 4 years ago in reply to Yeshaswini Palaksha

TI__Mastermind 24955 points

Yeshaswini,

What is the SDK version that you are using to send this particular configuration file to your AWR1843BOOST?

Regards,

Kyle

0 Yeshaswini Palaksha over 4 years ago in reply to Kyle Cousino

Prodigy 240 points

Hello Kyle,

version
Platform                : xWR18xx
mmWave SDK Version      : 03.04.00.03
Device Info             : AWR18xx non-secure ES 02.00
RF F/W Version          : 02.00.00.01.17.10.05
RF F/W Patch            : 01.02.05.02.19.04.30
mmWaveLink Version      : 01.02.05.25
Lot number              : 3613400
Wafer number            : 21
Die coordinates in wafer: X = 19, Y = 5
Done

Above I have provided Version info. RF F/W Version corresponds to SDK Version I believe.

If you have any Questions, let me know.

Best Regards,

Yeshaswini

0 Yeshaswini Palaksha over 4 years ago in reply to Kyle Cousino

Prodigy 240 points

Hello Kyle,

Any idea what is the Problem?

Best Regards,

Yeshaswini

0 Kyle Cousino over 4 years ago in reply to Yeshaswini Palaksha

TI__Mastermind 24955 points

Yeshaswini,

So there a couple of issues with your configuration file.

Based on the configuration file that you uploaded, there is not sufficient time to collect the desired number of ADC samples per chirp.

In your configuration, you specify:

number of ADC samples per chirp: 256

ADC sampling rate: 7200 ksps

ADC start time: 7 us

Ramp end time: 38 us

Based on the ADC sampling rate, you would need at least 35.5 us to collect 256 samples. However, since your actual sampling time is only 31 us (ramp end time - ADC start time), the SDK demo raises an error. In this configuration, there is not sufficient time to collect enough samples in each chirp.

This can be resolved by completing one of the following:

Decrease the ADC start time
Increase the ramp end time
Increase the ADC sampling rate
Decrease the number of samples per chirp

Please keep in mind that any of these changes will have an impact on your overall chirp design.

The second issue with your configuration file is that you specify 80 chirps per frame. Since you are only using a single TX, the number of doppler chirps and number of doppler bins are the same, in this case 80. This creates an issue when it comes time to completing the Angle of Arrival (AoA) portion of the processing chain.

The AoA processing unit specifies that the size of the detection matrix must fit within the M0 and M1 memory banks. Since each of the four memory banks in the HWA is 16KB, the detection matrix must be less than 32KB.

Per the Doppler (2D FFT) processing unit, the detection matrix is defined in size as the product of the number of range bins, the number of doppler bins, and the size of uint16.

Detection_Matrix_Size = sizeof(uint16_t) * num_range_bins * num_Doppler_bins

In the case where the number of chirp loops per frame is 64, the detection matrix size is 32 KB since,

Detection_Matrix_Size = 2 * 256 * 64

Detection_Matrix_Size = 32,768 bytes

Detection_Matrix_Size = 32 KB

Once you increase the number of chirp loops per frame above 64, this will increase the number of doppler bins to the next power of 2 size, which in this case is 128. This immediately makes the detection matrix size too large and asserts the error that you are likely seeing.

Detection_Matrix_Size = 2 * 256 * 128

Detection_Matrix_Size = 65,536 bytes

Detection_Matrix_Size = 64 KB

To avoid this error, you will need to decrease the number of range bins in your profile configuration to either 128 or 64. Then you will be able to increase the number of chirps per frame above 64.

Regards,
Kyle

0 Yeshaswini Palaksha over 4 years ago in reply to Kyle Cousino

Prodigy 240 points

Hello Kyle,

Thank you for the detailed Explanation.

Regarding the sample rate and ramp end time, I had corrected. Thank you for pointing that out.

But is there any work around to avoid the Memory Limitation? I am asking this Question because I am able to configure mmwave Studio with 256 range bins per chirp and 128 chirps per Frame, and it does not complain at all.

Any comment on this?

Best Regards,

Yeshaswini

0 Kyle Cousino over 4 years ago in reply to Yeshaswini Palaksha

TI__Mastermind 24955 points

Yeshaswini,

So you can configure mmWave Studio with 256 range bins and 128 chirps per frame because you are just collecting raw data and all of the processing is done off-line. Therefore, there are no restrictions when it comes to then angle calculations since the radar is just acting as a front-end device when used in mmWave Studio.

The other option would be to use the DSP implementation of the AoA processing unit instead of HWA implementation.

You will need to modify the mmw_res.h file in the SDK Out-of-Box Demo. You will need to change the following:

original line: #include <ti/datapath/dpc/dpu/aoaproc/aoaprochwa.h>

modified line: #include <ti/datapath/dpc/dpu/aoaproc/aoaprocdsp.h>

You can try this change and see if this works for you.

Regards,
Kyle

0 Yeshaswini Palaksha over 4 years ago in reply to Kyle Cousino

Prodigy 240 points

Hello Kyle,

Thanks for this Information. Exactly, I do not want post processing to be carried out by HW unit. I owuld just like to capture the data. Infact I would like to disable entire Signal processing unit done at HW Level.

Is it possible to configure RADAR board in such a way that no post processing takes place at HW Level and it just Transfer the data through LVDS?

Best Regards,

Yeshaswini

0 Kyle Cousino over 4 years ago in reply to Yeshaswini Palaksha

TI__Mastermind 24955 points

Yeshaswini,

Please refer to the "Streaming data over LVDS" section of the XWR1843 Demo Doxygen. You can configure the radar to send out the ADC data over LVDS.

C:/ti/mmwave_sdk_03_04_00_03/packages/ti/demo/xwr18xx/mmw/docs/doxygen/html/index.html#LVDSStreamingNotes

You can leave the OOB demo as is and just enable the LVDS streaming of the ADC data. Configuring the 1843 purely as a front end device using the OOB would require significant changes to the source code.

The other option is that you use the DCA1000EVM in conjunction with mmWave Studio and capture the raw ADC data this way.

Regards,
Kyle

Sensors

Sensors forum

Number of chirp loop per frame