AWR1843BOOST: Issue in radar configuration to capture raw data using DCA1000

u4564981

Part Number: AWR1843BOOST
Other Parts Discussed in Thread: DCA1000EVM,

Hi,

I'm using 1843 with DCA1000 to capture raw data via LVDS using the mmw demo project. I've followed the constraints of radar configuration (bandwidth < 4 GHz, slope < 100, num_loops < 256, duty cycle < 50%, etc.) for the radar configuration; however, when I try to program the configuration, the message received at serial after sensorStart command is Debug: Init Calibration Status = 0xffe, and I receive no raw data from the DCA. I'm using all 4 rx channels and the configuration snippet is:

channelCfg 15 1 0
profileCfg 0 77 47.0 3 50.4 0 0 67.5 1 256 6000 0 0 30
frameCfg 0 0 239 0 50 1 0
lvdsStreamCfg -1 1 1 1

However, the everything works fine for a test low-throughput specification. In this case, the serial output after sensorStart command comes out as Done. The modified specification is:

profileCfg 0 77 76 6 90 0 0 20 1 256 3200 0 0 30
frameCfg 0 0 64 0 50 1 0

Am I missing some constraint which is leading to data pipeline failing. I found in some earlier posts that 0xffe status is fine; but with serial output as Debug state, I'm not getting any output whatsoever. Also, I'm curious what causes Debug output to be raised if there's no error.

Thanks

over 5 years ago

0 Kyle Cousino over 5 years ago

TI__Mastermind 24955 points

For the current release of the mmWave SDK, version 3.3, please refer to the User Guide found in the "docs" folder.

Take a look at section 3.3.2 "mmWave demo with LVDS-based instrumentation".

This section will provide instructions on how to configure the DCA1000EVM with the Out-of-Box Demo as well as how the output data is structured.

There are additional steps that are needed just beyond changing the configuration (.cfg) file.

Regards,

Kyle

0 u4564981 over 5 years ago in reply to Kyle Cousino

Intellectual 445 points

Hi Kyle,

Sorry if it was not clear in my post - I have data-capture setup working. If I used the second configuration, then everything works and fine and I can get data out. However, when I try to step up the game by using first configuration in the post, the configuration programming shows Debug: Init Calibration Status = 0xffe and I don't get any data out.

For the first configuration, I've adhered to all configuration specifications/constraints I'm aware of, but still the config programming is not successful. So my questions are:

Is there any constraint getting violated with the following profile and frame configuration?
1. 1. profileCfg 0 77 47.0 3 50.4 0 0 67.5 1 256 6000 0 0 30
    frameCfg 0 0 239 0 50 1 0
what causes Debug output (Debug: Init Calibration Status = 0xffe) to be raised in radar configuration?

Thanks.

0 Kyle Cousino over 5 years ago in reply to u4564981

TI__Mastermind 24955 points

What are your chirpCfg CLI commands that you are sending to your AWR1843BOOST EVM?

The chirpCfg is a mandatory CLI command.

The process for constructing a radar frame is as follows: profile --> chirp --> frame

Were a profile is used to construct N number of chirps and then M number of chirp loops are used to construct a radar frame.

Please provide your chirp configurations that you are using or make sure to include them in your .cfg file.

Regards,

Kyle

0 u4564981 over 5 years ago in reply to Kyle Cousino

Intellectual 445 points

Sure, please find the full configuration. I modify only profileCfg and frameCfg from the working configuration profile.

sensorStop
flushCfg
dfeDataOutputMode 1
channelCfg 15 1 0
adcCfg 2 1
adcbufCfg -1 0 1 1 1
profileCfg 0 77 47.0 3 50.4 0 0 67.5 1 256 6000 0 0 30
chirpCfg 0 0 0 0 0 0 0 1
chirpCfg 1 1 0 0 0 0 0 0
frameCfg 0 0 239 0 50 1 0
lowPower 0 0
guiMonitor -1 1 1 0 0 0 1
cfarCfg -1 0 2 8 4 3 0 15 1
cfarCfg -1 1 0 4 2 3 1 15 1
multiObjBeamForming -1 1 0.5
clutterRemoval -1 0
calibDcRangeSig -1 0 -5 8 256
extendedMaxVelocity -1 0
lvdsStreamCfg -1 1 1 1
compRangeBiasAndRxChanPhase 0.0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
measureRangeBiasAndRxChanPhase 0 1.5 0.2
CQRxSatMonitor 0 3 5 121 0
CQSigImgMonitor 0 127 4
analogMonitor 0 0
aoaFovCfg -1 -90 90 -90 90
cfarFovCfg -1 0 0 8.92
cfarFovCfg -1 1 -1 1.00
sensorStart

Thanks.

0 Kyle Cousino over 5 years ago in reply to u4564981

TI__Mastermind 24955 points

It looks like you might need to make a slight modification to your frameCfg command.

In the Doxygen for the 1D FFT, it states that:

"Due to the Ping/Pong scheme implemented in rangeProcHWA DPU, numChirpsPerFrame/numTxAntennas should be an even number."

Based on your previous post, your number of chirps per frame is 239 and your number of TX antennas is 1. This would result in an odd number. Please change to ensure this is an even number.

Additionally, when I run your configuration from your post above, I receive the following debug messages in Code Composer Studio:

-----------------------------------------------------------------------------------------------------------------

[Cortex_R4_0] **********************************************
Debug: Launching the MMW Demo on MSS
**********************************************
Debug: Launched the Initialization Task
Debug: mmWave Control Initialization was successful
Debug: mmWave Control Synchronization was successful
[C674X_0] Debug: DPM Module Sync is done
[Cortex_R4_0] Debug: CLI is operational
Debug: Sending rlRfSetLdoBypassConfig with 3 1 0
Error: DPM Report 4 received with error:-40112 arg0:0x64 arg1:0x8003e88
{module#9}: "./mss/mss_main.c", line 2046: error {id:0x10000, args:[0x1652c, 0x1652c]}
xdc.runtime.Error.raise: terminating execution
[C674X_0] Error: DPM Report 4 received with error:-40112 arg0:0x64 arg1:0x8148dc
xdc.runtime.Main: "./dss/dss_main.c", line 256: assertion failure
xdc.runtime.Error.raise: terminating execution

-----------------------------------------------------------------------------------------------------------------

If I search for error code #40112, this code evaluates to the following:

"Error Code: Out of L3 RAM during detection matrix allocation."

This means that after the device has allocated enough space for the radar cube matrix, when it goes to allocate additional memory for the detection matrix, there is not enough space for this detection matrix.

This can be resolved in one of two ways:

Decrease the number of chirp loops
Decrease the number of ADC samples per chirp

Also as a comment on the .cfg file you shared in your previous post, line 9 appears to be unnecessary as no TX antennas are enabled for that chirp.

Regards,
Kyle

0 u4564981 over 5 years ago in reply to Kyle Cousino

Intellectual 445 points

Thank you for looking into the problem. I thought total number of chirps = number of loops + 1. Because some example implementations show number of chirps = 256, and the sdk in frameCfg mentions max number of loops as 255. Should this number be 256 then?

The insufficient size of detection matrix is plausible explanation. However, the raw data capture from LVDS does not need in-built data processing pipeline, so I could probably allocate more memory to detection matrix for the sake of overcoming this problem. I had tried working with the mmw demo code in CCS earlier. To incorporate the mmw demo code in CCS, I attempted to modify the available projectspec file of sdk 3.1 as the projectspec file for sdk 3.3 wasn't available. However, I ran into issues there as well as some symbols remained resolved (my post: https://e2e.ti.com/support/sensors/f/1023/t/890004). Could you please share your projectspec file for sdk 3.3 mmw demo 1843 code or let me know the issues in my projectspec file in that post?

Then, I could check if increasing detMatrix memory allocation fixes the problem. As the data in ADC buffer is sent for each chirp in DCA based LVDS data capture, I believe I could achieve 256 chirps per frame, with each chirp of 512 complex samples for my application.

Thanks

0 Kyle Cousino over 5 years ago in reply to u4564981

TI__Mastermind 24955 points

You can download a .projectspec file for the AWR1843BOOST Out-of-Box Demo from the Industrial Toolbox on TI's Resource Explorer.

The link for the Industrial Toolbox is found here: http://dev.ti.com/tirex/explore/node?node=AJoMGA2ID9pCPWEKPi16wg__VLyFKFf__LATEST

Download this to your PC and extract to the C:\ti folder

You will find the .projectspec files in the file directory listed below.

C:\ti\mmwave_industrial_toolbox_4_2_1\labs\out_of_box_demo\18xx_mmwave_sdk\src

Regards,

Kyle

Sensors

Sensors forum

AWR1843BOOST: Issue in radar configuration to capture raw data using DCA1000