• State TI Thinks Resolved
  • Locked Locked
  • Replies 13 replies
  • Answers 7 answers
  • Subscribers 33 subscribers
  • Views 3614 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

IWR1642BOOST: Antenna Pattern is not centralized in elevation

Kang Liu
Prodigy 90 points
Part Number: IWR1642BOOST
Other Parts Discussed in Thread: IWR1642, , AWR1642BOOST, AWR1642, IWR1443

Hallo, 

During our Radar development, we have measured your Product of IWR1642 we bought from TI online Shop .

We find that the antenna performs totally different and very worse as described in the data sheet.

elevation pattern is not at 0° centralized, but with bias of 17°

 

Is this design from TI correct?

 

Kang 

 

IWR1642_Antennapattern.pdf

  • 0
    TI__Guru*** 143834 points
    Hi,

    The information provided in the datasheet is based on testing.

    How do you perform your testing?

    Thank you
    Cesar
  • 0 in reply to Cesar
    Prodigy 90 points

    Cesar,

     

    The test process is described as follows:

    We use the mmWDemo program as firmware. 

    The sensor is mounted on the table and it could be turned by the motor, so that the sensor scans the corner reflector in front of 1.5m. 

    The data we recorded is the target Intensity shown in the demo figure range profile. 

    After that we have a plot, the x axis is the motor angle, the y axis is the intensity of this corner reflector. 

    We repeat this process for azimuth and elevation.

     

    thanks,

    Kang

  • 0 in reply to Kang Liu
    TI__Mastermind 37500 points
    Hello Kang,
    We have sent your PDF to the antenna development engineer, we will send a response after he reviews this. Please inspect the mmwave sensor to antenna, to make sure there is no solder mask, or material on top of the RF traces. Please also send us the board serial number.

    Regards,
    Joe Quintal
  • 0 in reply to Radio-Joe
    TI__Mastermind 37500 points
    Hello Kang,
    If you could send a close up picture of the U2 mmwave sensor to the RF antenna traces, that may help us. There is a picture like this in the EVM User guide, www.ti.com/.../swru521a.pdf, Figure 9.

    The RF traces, have a silver immersion finish (it can tarnish). There should be no delaminated RF trace sections, or deformed patches. There should be no scratches through the traces. It should have no soldermask covering the RF traces.

    To try to isolate the antenna patterns, you can select 1Tx, using the .cfg files, you can manually change from Tx1, or Tx2.
    channelCfg - this should have a Tx mask of 3 (both Tx1, Tx2)
    Rx mask of 15 (all 4 Rx1, Rx2, Rx3, Rx4)
    chirpCfg - here you can select Tx1, or Tx2 only.
    These assume only one chirp type is done per frame. TDM MIMO alternates chirps, to effectively have additonal receivers.

    with 1 Tx, and 4 Rx the software can resolve the specific GUI monitor. (bottom left of Visualizer)

    If you are using the mmWave SDK data capture demo, could you please restrict the RF Slope, so we have several narrowband sweeps?
    76-76.567Ghz (needs to be one sweep) FreqStart 76
    DFE Out rate 5e6
    256 adc samples per chirp
    complex 1x
    rampend time = adcstart(4.5us) + 256adcsamples/5e6 + excessramptime (1us) = 56.7u
    idle time 130 us - this allows time to send data, its is larger than min value
    Synthesizer Slope is 9.994Mhz/us
    RF bandwidth is 566.66Mhz (Synthesizer Slope * rampend time)

    if these are repeated at 1Ghz start frequencies, 76,77,78,79,80

    This can be done with the mmwave Data capture demo, or with the dfp Radar Studio (with additional hardware, Devpack, TSW1400, and Radar Studio software)


    Regards,
    Joe Quintal
  • 0 in reply to Radio-Joe
    Prodigy 90 points

    Hi Joe,

    Thanks for your reply.

    the antenna pattern we have measured is done by configuring 1Tx and 1Rx and only with 1GHz bandwidth.

    for this example, we use the configuration

    channelCfg 1 1 0
    adcCfg 2 1
    adcbufCfg -1 0 0 1 0
    profileCfg 0 77 7 16 110 0 0 20 1 256 3000 0 0 30
    chirpCfg 0 0 0 0 0 0 0 1
    frameCfg 0 0 128 0 100 1 0

    Another topic is the antenna trace status, I have a close up picture here

    this is the antenna from one of our IWR1642boost boards

    this is the second antenna of our IWR1642 boards

    this is from our AWR1642boost board

    All of IWR1642boost boards have problem with elevation bias, one is 17 degree, the other is 24 degree.

    The AWR1642 boost works well.

    you can the board serial numbers in the above pictures.

    could you figure out whether the RF traces have problem?

    Thanks,

    Kang

  • 0 in reply to Kang Liu
    TI__Mastermind 37500 points

    2526.mmWave_hw_design_guide_rev_9.pdfHello Kang,

    I have attached a document that has the antenna dimensions.   The visual inspection should have no cuts or scratches in the RF FMCW, antenna stripline extensions, or antenna patch.   The rectangular patches, and patch to patch antenna stripline should match the dimensions within 1mil.  

    In the radar configuration of 77-78Ghz we do not expect an elevation tilt.  The simulation does show a tilt from 80-81Ghz.

    We can arrange to have you send an EVM, we will measure it in our lab and replace it.   If you have a TI Field Application Engineer in your area, he can advise you, I will get an address for you to return the EVM for testing / replacement.

    Please note, in the measurements for the IWR1642 antenna we have Tx1,  and (IWR1443) Tx3.  The Rx dimensions are OK.   

    Related to the included, 2_28_18_1642_AntennaMeasurementCompare.xlsx, we have measured several IWR1642 boards, could you please try one of your boards with the Chirp setup included in the Excel, and send the results back.

    Regards,

    Joe QuintalAntennaMeasurementCompare.xlsx

  • 0 in reply to Radio-Joe
    Prodigy 90 points

    Hi Joe,

    Thank you very much for your data and answer.

    the elevation measurement from our side is attached. 

    The EVM is given back to TI FAE. Please evaluate and find the problem.

    Thanks,

    Kang

    elevation_iwr_awr.pdf

  • 0 in reply to Radio-Joe
    TI__Mastermind 37500 points

    1884.AntennaMeasurementCompare.xlsxHello Kang,

    We have checked several EVMs, and performed an azimuth and elevation test.   I think you are discussing with Vaibhav, and a replacement EVM is being processed for you.   The test document is attached.  

    There is a possibility that the  commanded RF value, is not the same as the radiated RF value.  If you have the equipment to test  the Tx output at a single Frequency, there is a test method in the mmwave SDK User Guide for Continuous Tx.  The suggestion is to select 78Ghz.  Since the EVM is using a crystal with 50 ppm tolerance, and has some temperature drift.  There is an expected variation in the transmit frequency.

    We are working on a test for the 40Mhz output, I will send those instructions, after I have tested them in the lab today.

    Regards,

    Joe Quintal

  • 0 in reply to Radio-Joe
    TI__Mastermind 37500 points

    Hello,

    If you use Radar Studio which can be downloaded from the Ti website, and you have the Devpack board,   You can use the sample configuration attached, and add one set of manual operations, (RegOps tab) to write a debug value, @0x4320140C, read, modify bits 16.19 with 0x7.  This will activate the 40Mhz crystal clock to the J2 connector, should be 1v amplitude, 40Mhz with 50 ppm.   This can be used to validate the 40Mhz clock.

    we have evaluated a returned EVM with the test procedure described in AntennaMeasurementCompare.xlsx, we are working on the cause of the problem.  

  • 0 in reply to Radio-Joe
    Prodigy 90 points

    Hi Joe,

    could you give me more exact position, where is the description for ?

    there is a test method in the mmwave SDK User Guide for Continuous Tx.

    Another point:

    As I understand from you explanation, the antenna pattern problem from temperature and the drift of frequency and RF is not as the command required.

    However, I saw the picture you showed, which has exact 40MHz.

    Do you have some measurement from our bad EVM board?

    Regards,

    Kang

  • 0 in reply to Kang Liu
    TI__Mastermind 37500 points
    Hello Kang,
    The Continuous Output is described in section ,wave sdk User Guide 1.1.0, on page 19, The demo also supports continuous mode. The demo directory has the capture_demo_script_lvds_cont_mode.txt file which has a list of the CLI commands used for
    configuring the system in continuous mode.

    Use the keyword "continuous" you can find the different APIs.

    The 40Mhz clock, was measured on my lab board.

    Regards,
    Joe Quintal
  • 0
    TI__Mastermind 37500 points
    Kang,
    The returned 1642 board measurements are being done in the next few days.
    Regards,
    Joe Quintal
     
  • 0 in reply to Kang Liu
    TI__Mastermind 37500 points

    Hello Kang,

    In the mmwave SDK user guide, there is the following note:

    The demo also supports continuous mode. The demo directory has
    the capture_demo_script_lvds_cont_mode.txt file which has a list of the CLI commands used for
    configuring the system in continuous mode.

    In the ti\mmwave_sdk_01_01_00_02\packages\ti\demo\xwr16xx\capture

    capture_demo_script_lvds_cont_mode.txt - has the commands for continuous mode.

    If you have a Tx source that is 1-2Mhz above the Continuous Frequency, you should be able to down convert that to the Receiver.

    You may need to set the Tx backoff to 30db, so we are reading your external horn antenna reference.

    Also without the Tx backoff, you can measure the Tx output frequency,

    (Note: removing the crystal, crystal CLKP load capacitor, and CLKM load capacitor install a 100 ohm resistor to both CLKP and CLKM

    drive an AC coupled 40Mhz +/- 50ppm 1.2v reference to CLKP for better stability.

    The clock was measured today in the lab, I will post the results tomorrow.  we have measured your board, and found the tilt, we have sent you another board.

    Regards,

    Joe Quintal