Login/Register
Hint: separate multiple terms with commas
Date
E.g., 12/19/2018
Date
E.g., 12/19/2018
Feedback
Global
Global
China (简体中文)
Japan (日本語)
Korea (한국어)
Taiwan (繁體中文)
TI Training >
Applications & designs
Products
Tools & software
Live training
Toggle navigation
Hint: separate multiple terms with commas
Date
E.g., 12/19/2018
Date
E.g., 12/19/2018
Global
China (简体中文)
Japan (日本語)
Korea (한국어)
Taiwan (繁體中文)
Email
2.4 DCA1000 Training Video
In this video, we will be covering the basics of raw data capture using TI's mmWave sensors, the DCA1000 capture card, and mmWave Studio. We will be covering the requirements, the hardware and driver setup, setting the static IP address, capturing the radar data, and additional information such as the data flow and data file format. For the hardware requirements, we're going to need a TI mmWave EVM and a 5 volt, 2.5 amp power supply with a micro USB cable. Additionally, we will need the DCA1000 EVM, a 5 volt, 2 amp power supply, micro USB cable, RJ45 ethernet cable, and a 60 pin Samtec cable. For the software, we will need mmWave Studio, Matlab Runtime Engine version 8..5.1, and Code Composer Studio version 7.1 or higher or the XDS Emulation Software Package. Now that we have covered the hardware and software requirements, we can talk about how to connect the mmWave EVM and the DCA1000 to your PC. We want to connect the mmWave EVM and the DCA1000 as shown in the image. First, attach the spacers and L brackets. Next, attach the 60 pin Samtec cable to J3 of the DCA1000 and J1 of the mmWave EVM. The connectivity to the PC is established by two micro USB cables and an ethernet cable. Connect one of the micro USB cables to J8 of the mmWave EVM and your PC and the other to the J1, the radar FTDI port, of the DCA1000 in your PC. Connect an ethernet cable to the ethernet port on J6 of the DCA1000 and the PC. To power on the devices, connect the 5 volt, 2.5 amp power supplies to both the DCA1000 and mmWave EVM and a power source. Ensure that switch three on the DCA1000 is set to DC jack 5 volt in. Make sure that the SOP jumpers on the mmWave EVM are set so that SOP0 and SOP1 are closed and SOP2 is open. With the devices powered on and connected to a PC, we can now verify that the COM ports are properly detected. Navigate to the Device Manager on your PC and check to see that you see the COM port numbers for the AR DevPack and XDS110, as shown in the image. Your COM port numbers may differ from the numbers shown in the image. In some cases, the FTDI drivers need to be installed manually. If the COM ports for the AR DevPack do not appear under the ports, COM, and LPT ports, and you have a yellow Warning icon, right click on AR DevPack port and select the Update Driver Software option. Choose the Browse My Computer For Driver Software option and enable the Include Subfolders tick box. Select the FTDI folder in your mmWave Studio Install folder for the FTDI drivers. You will need to install the drivers for each port individually. In some cases, you might need to install them twice for each port. When the drivers have been successfully installed, they will appear in the ports, COM, and LPT tab of Device Manager as shown in the image. With the DCA1000, the data is transferred over ethernet as UDP packets. The FPGA on the DCA1000 has been programmed with the destination address of 1921683330. For this reason, we must configure the target PC to a static IP address that matches the expected destination address of the FPGA in order to receive data from the EVM. I will now walk through how to setup the static IP address. Open your Network and Sharing Center and click on the Change Adapter Settings option. Open the Local Area Connection, click on Internet Protocol version 4 TCP IPv4, and select Properties. Select the Use the Following IP Address option and change the IP address and subnet mask to the values shown on the screen. Once done, click OK and close the Local Area Connection Properties window. With the static IP address set and the devices successfully connected to the PC, we can now begin to capture some data. Open mmWave Studio in the mmWave Studio Install folder under mmWave Studio, Runtime, mmWave Studio EXC. Before moving any further, we should check that we have the mmWave EVM set to the correct SOP mode. We want to set the EVM for SOP2, development mode. In order to do this, make sure that you have the SOP jumpers connected on SOP0 and SOP1, and removed from SOP2. Once mmWave Studios open, please verify that the FTDI connectivity status is connected. If it does not say connected, verify that you have properly installed the FTDI drivers. We will now begin by connecting RS232. First, select the correct COM port under RS232 options. In my case, this corresponds to COM port 8, which is the application user UR COM port. Once selected, click on Set, followed by Connect. If the connection is successful, you should see the RS232 connectivity status changed to connected. With the UR connected, we can now begin loading the firmware. First, select the appropriate BSS and MSS firmware options. You can do this by going to your mmWave Studio Install folder, selecting the RF Eval firmware folder, and selecting Radar SS for the BSS option, and then choose based on your device. In my case, I'm using the 1642 ES1, so I'll select that firmware. Do the same thing for the MSS firmware, but under the Master SS folder. Next, load the BSS firmware first, followed by the MSS firmware. If successful, you should see the BSS firmware version and MSS firmware version appear. We can now connect Spy and power up the RF subsystem. First, click on Spy Connect. If the connection is successful, you should see the Spy connectivity status change to connected. If Spy fails to connect, there is a high probability that you have selected the wrong firmware. Next, click on RF Power Up. We can now move to the Static Configuration tab. In the Static Configuration tab, we can configure the channels, the ADC, and the internal RFLDO. First, begin by selecting the desired number of Tx Rx channels and the desired ADC configuration. Once done, click on Set. If you are supplying the RF power rails of the mmWave device with 1 volt, then you should select the RFLDO Bypass option. In cases where you are supplying 1.3 volts, do not select the RFLDO Bypass option, as this can damage the device. In this case, we will not be selecting the RFLDO Bypass option. Next, configure your LPADC mode and click Set. Once done, click on RF in it. We can now move on to the DataConfig tab. In the DataConfig tab, we can configure the data path, the clock, and the LVDS lanes. With the DCA1000, data will always be captured over four lanes with the 12 and 1443 devices and two lanes with the 1642 devices. For now, we will leave these as default values and click Set for data path configuration, Set for clock configuration, and Set for LVDS lane configuration. We can now move on to the Sensor Configuration tab. In the Sensor Configuration tab, we can define the chirp and frame parameters. We will begin with the Profile section. In this section, we can define the RF parameters of the chirp. The device is capable of supporting multiple chirp profiles, with a maximum of four profiles supported. This can be accomplished by changing the profile ID. In this case, we will use a single chirp profile with an ID of zero, and click on Set. The Chirp section allows us to add additional variations to the chirp, as well as specify the sequence of the chirps. For example, if we wanted to have a sequence of eight chirps with the first four chirps corresponding to profile ID zero and the next four chirps corresponding the profile ID one, we would begin by setting the first four chirps by specifying the start chirp and the end chirp. In this case, zero to three will correspond to the first four chirps, and click on Set. We could then change the profile ID to one and specify the start chirp as four and the end chirp as seven and click on Set. This would create the sequence of eight chirps with four chirps for profile zero and four chirps for profile one. In this case, we will leave the default values corresponding to a single chirp profile with profile ID zero. Click on Set to set your chirp config. Once you have set your profile and your chirps, you can go back and view your previously configured profiles and chirps by selecting Manage Profile and Manage Chirps. We can now move onto the Frame section. In the Frame section, we can specify which of the chirps will be used in a frame, as well as the total number of frames, the number of chirp loops per frame, and the frame periodicity. The frame periodicity will also be used to determine the duty cycle. This can vary based on how much time you need in between your frames to do any processing of data. Additionally, we can add a trigger delay and select the trigger for software trigger or hardware trigger. With a hardware trigger, the frame will be triggered by an external source on the digital sync end pin, which is typically used in Cascade applications. In our case, we will leave the default values and click Set when you are done configuring your frame. Be sure that you go in the order of profile, followed by chirp, followed by frame. Once you've finished configuring your chirps and frames, you can save your configuration by selecting the SaveConfig button. You can load previously saved configs by selecting the Load Configuration button. We will now move on to setting up the DCA1000. Make sure you have the DCA1000 tick box selected, and click Set Up DCA1000. When the window pops up, click on Connect, Reset, and Configure. You should see the FPGA version appear. With the DCA1000 connected, we want to ARM the DCA1000 and trigger the frame. Click on DCA1000 ARM and press Trigger Frame. The data will be captured in a binary file, which we will explain later in this video. To view the captured data, you can click on the Post PROC tab. This will create a new binary file that will be saved in the location specified by the dump file path. With the post processing utility, we can view the time domain plots, 1D-FFT plots, as well as many other additional plots. You can configure the plots by selecting the plot type in the dropdown menu for each of the four plots. Additionally, you can play the frame data back by clicking on the Play button. We will now talk about the data capture flow and data format. The data is captured over ethernet in the form of UDP packets. The files will be split after one gigabyte of data. So, for example, three gigabytes of data will be stored in three files with the names of adc_data_Raw_0, 1, and 2. The format of the captured file is shown and consists of a sequence number, a data length field, a byte count, and the raw mode data. When you click the Post PROC button in mmWave Studio, the ADC data will be parsed by mmWave Studio. It is important to note that only the first set of raw data will be picked. If you have multiple gigabytes of data stored, you will have to develop your own post-processing utility to process the rest of the files. Due to that ethernet protocol, the data may not be in order, or some packets may be dropped. mmWave Studio will automatically run a packet reorder and zero fill utility to correctly reorder the packets and add zero fill for any dropped packets before storing the data file. If you are noticing a large amount of dropped packets, you may want to try increasing the packet delay in the DCA1000 Connection menu. After the headers have been removed and the data is reordered, the raw ADC data will be written to the file you have specified in mmWave Studio. The data will be stored in a 2's-complement format. It will be stored in an interleaved format for the 12 and 1443 devices, and a non-interleaved format for the 1642 devices. For detailed information on how the data is organized, I recommend taking a look at the ADC raw data capture app note. In this app note, there is a lot of detail on the DCA1000 data format. For example, it defines how the data is stored in the file from the start to the end of the file for real data and complex data with the 1443, 1243, and 1642 devices. Additionally, there is a section with Matlab code provided that helps with how to interpret the data that is stored. This Matlab code is going to read the binary file and organize the data into an array based on receivers. Each receiver is going to correspond to a given row. So row one will contain all of receiver one data, row two will contain all of receiver two data, and so on. This allows for a starting point for developing your own post-processing utilities without having to worry about how to interpret the binary file. It is important to note that this code has only been designed to be used after the packet reorder and zero fill utility has been run. This concludes our presentation on data capture using the DCA1000 data capture card. For more information, please go to ti.com or ask us a question on our E2E forums.
Watch the next video in the series:
mmWave Radar Programming Model
Details
Date:
August 27, 2018
This video describes how to capture and post process raw data using a mmWave EVM with the DCA1000 Capture Card and mmWave Studio
PDFs for download
mmwave_sensor_raw_data_capture_using_dca1000_v02.pdf
Related courses and events
Practical Comparisons of DC/DC Control Modes - Solution Size Comparison
Date:
April 19, 2017
TI Precision Labs – ADCs: Introduction to Frequency Domain
Date:
April 14, 2017
Webinar - 6W-50W Class-D amplifiers
Date:
September 19, 2017
Power over Ethernet in Connected Lighting
Date:
November 8, 2018
Engineer It
Date:
March 14, 2016
arrow-top
close
delete
download
menu
search
sortingArrows
zoom-in
zoom-out
arrow-down
arrow-up
arrowCircle-left
arrowCircle-right
blockDiagram
calculator
calendar
chatBubble-double
chatBubble-person
chatBubble-single
checkmark-circle
chevron-down
chevron-left
chevron-right
chevron-up
chip
clipboard
close-circle
crossReference
dash
document-generic
document-pdfAcrobat
document-web
evaluationModule
globe
historyClock
info-circle
list
lock
mail
myTI
onlineDataSheet
person
phone
question-circle
referenceDesign
shoppingCart
star
tools
videos
warning
wiki