Post Go back to editing

Issue with ADRV9025 Build on ZCU102 Evaluation Board - Not getting Constellion

Category: Software
Product Number: ADRV9026
Software Version: Petalinux Version 2022.2

Hi ,

We have built the ADRV9025 project by running the following project in Vivado version 2022.2 on the ZCU102 Evaluation board:

 https://github.com/analogdevicesinc/hdl/tree/main/projects/adrv9026 

We also took the Petalinux build in the 2022.2 Vivado version. While we are able to get the spectrum, we are not getting the constellation.

Details:

  • Bandwidth: 100 MHz
  • Sample Rate: Changed from 245.76 MHz to 122.88 MHz

Additionally, we are not able to find the file needed for the 100 MHz test purpose.

Could you please assist us in resolving these issues?

Thread Notes

  • While we are able to get the spectrum, we are not getting the constellation.

    Is it TDD or FDD?

    Additionally, we are not able to find the file needed for the 100 MHz test purpose.

    How are you able to get the spectrum if you don't have the waveform file? We can not provide Test waveform files due to MATLAB licensing restrictions. However we have some files available here. LTE waveform files 

    Most probably the JESD data mapping may be incorrect while you are receiving on the deframer side. Please check this part.

    Is your JESD link is up and deframer status is as expected?

  • Hi,

    This is Peter. I am also a part of this team. Thank you for replying quick. Here is our response and observations:

    The process that we followed is as follows:  
    1. An 'NR1.1' file was generated using a VXT, with details as shown :
    FDD, BW 40MHz, Subcarrier Spacing : 30kHz

    2. This file was saved in the .mat format using IIO Oscilloscope, with 1 SF samples i.e., 122880 samples.

    3. This was then pushed to the dac buffer o/p, and uploaded to the tx side of FPGA

    4. The output was then plotted. The BW spectrum for 40MHz was observed, however the constellation seemed to degrade at times, it was inconsistent. This was observed in the National Instrument RF analyzer.  

    PFA the video of the constellation observed.

    However, this was not seen with the R&S analyzer. Also matlab plotting of this file was also not giving any constellation.

    5. Parallelly we also tried to generate a matlab file for the same,
     i) complex integer IQ format,
     ii) signed integer IQ format,

    However we were not able to get to any constellation with this.

    Also to answer your previous queries : "Is your JESD link is up and deframer status is as expected?"

    The answer to both is YES. PFA the screenshot

  • Looking at the video you shared, it seems to me that the LO Leakage is so high. Have you enabled LOL Init cal and tracking cal? Do you have Tx connected back to ORX for the external LOL tracking cal to run?

    Enable this and check.