• State Resolved
  • Locked Locked
  • Replies 11 replies
  • Subscribers 48 subscribers
  • Views 2949 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

ADC12DJ3200EVM: Demo with KCU105 Eval kit shows poor results with noise floor @ -50dB and frequency folding for the fundamental

Manjunath Bevinamar
Prodigy 85 points
Part Number: ADC12DJ3200EVM
Other Parts Discussed in Thread: ADC12DJ3200,

Hello,

We are while setting up the demo for the TI-ADC12DJ3200EVM kit with Xilinx KCU105 board using the UHWD_2016.3 bit files we are seeing poor results.

We are seeing the very Noise floor at -50dBFs, and secondly the frequency folding is happening for the targeted frequency.

The JMODE0 is used by the ADC12XXXGUI and the settings and setup as described in the SLAU711 were correctly followed.

Setup info: ADC12DJ3200 EVM with Xilinx KCU105 EVM( Both static & DHCP Ethernet modes)

ADC sampling rate: 1600MSPS , output data rate: 3200MSPS.

ADC in JMODE0 

The noise floor remains at the same level even without the input signal connected to ADC input.

Individually the ADC and the Xilinx KCU105 Eval kit work and show proper results for the JESD204 interfaces when tested with other demos.

However when together interfaced we observe the above issue.

Appreciate for feedbacks/suggestions/advises.

Thanks and Best Regards,

Manjunath

  • 0
    TI__Mastermind 44680 points

    Hi Manjunath

    Can you provide the REV value of the ADC12DJ3200EVM. It will be in the silkscreen on the top of the board at the FMC+ connector end of the board.

    On the ADC EVM GUI, navigate to the Control tab, and click on the Enable Test Pattern Mode button. Then do another data capture.

    From High Speed Data Converter Pro, click on File>Save I32 Codes as CSV. Please save that file and attach it to your next post so I can compare with the expected data. 

    You should see a repeating 4 value sequence as follows. Adding 2048 to these values gives the equivalent offset binary value for the ADC.

    768

    512

    -769

    -513

    Doing a similar data capture and export, for the ADC in normal 12-bit data mode with no input, you should see values in the following range:

    -2

    -1

    5

    2

    -4

    -1

    3

    -2

    etc.

    There is one other thing you can try:

    • Navigate to the ADC GUI JESD204B tab.
    • Click on JESD Block Enable to disable it while making changes
    • Click on Scrambler Enable button to disable scrambling
    • Click on JESD Block Enable to re-enable the block.

    Try another capture and see if the results are better.

    Let me know how these various tests go.

    Best regards,

    Jim B

  • 0 in reply to Jim Brinkhurst1
    Prodigy 85 points

    Hi Jim,

    Thank you for the reply and info.

    Please note the revision of the ADC12DJ3200 is "Rev-A".

    Secondly we carried out the steps suggested by you and we didnt see the improvements.

    However I am attaching the captured .csv's for the normal mode, Pattern Gen selected mode and with the Scrambler-Disabled mode .

    We are still not sure what could be the root cause of this issue, and hence appreciate your help in addressing this query.

    Thanks and best Regards,

    ManjunathADC12DJ3200_GUI_PatternGenEn_Capture.zip

  • 0 in reply to Manjunath Bevinamar
    Prodigy 85 points

    Hi Jim,

    Today I were re-trying the same setup and we were able to see the results as attached.

    Appreciate if you can have a look at this as well.

    Thanks and Best Regards,

    Manjunath

    ADC12DJ32XX_GUI_JESD-Tab.zip

  • 0 in reply to Manjunath Bevinamar
    Prodigy 90 points
    Hi Manjunath,
    I have used ADC12DJ3200_EVM and KCU105 board to test the characters.The results are poor.So I expect to communicate with you about this problem.According to Jim's reply, we need to get the repeating 4 value sequence,and adding 2048 to these values gives the equivalent offset binary value for the ADC. I have got the value sequence .Do you know how to use these values next?
    Appreciate for your reply.
    Thanks and Best Regards,

    Liang
  • 0 in reply to user4632994
    Prodigy 85 points
    Hello Liang,
    Thank you for approaching me on this issue.
    It is very strange indeed to see the same issue is being observed by you as well.
    I too am able to see the pattern, but it seems the values that I could observe do not match the ones Jim was citing in earlier responses.
    I am not sure on what else could be done with the sequence other than for using it for the debugging in this case.
    May be Jim can reply us both on this to check further on mine and your captured data and how it could be used further.
    Regards,
    Manjunath
  • 0 in reply to Manjunath Bevinamar
    TI__Mastermind 44680 points

    Hi Manjunath

    We plan to re-test the ADC12DJ3200EVM with a KCU105 next week. I'll let you know as soon as I have any updates to help resolve the issue you have encountered.

    Best regards,

    Jim B

  • 0 in reply to Jim Brinkhurst1
    Prodigy 85 points
    Thank you Jim,

    We would be eagerly awaiting for your updates on this re-test.

    With Best Regards,
    Manjunath
  • 0 in reply to Manjunath Bevinamar
    TI__Mastermind 44680 points

    Hi Manjunath

    Thanks for being patient.

    After some digging we determined that the ADC12DJ3200EVM files created for the KCU105 were based on a pre-production version of the ADC EVM. Some changes to the EVM data lane polarities cause problems when using the production EVM. The 2 files contained in the following .zip file will resolve the issues you are having:

    /cfs-file/__key/communityserver-discussions-components-files/73/8750.KCU105_5F00_ADC12DJ3200EVM_5F00_Update.zip

    The .dll file should be copied into this folder location: C:\Program Files (x86)\Texas Instruments\High Speed Data Converter Pro\KCU105 Details

    The .ini file should be copied into this folder location: C:\Program Files (x86)\Texas Instruments\High Speed Data Converter Pro\KCU105 Details\ADC files

    Best regards,

    Jim B

     

  • 0 in reply to Jim Brinkhurst1
    Prodigy 85 points

    GoodResultsAfterTIFileUpdates.zipHi Jim,

    Thank you so much for taking the effort and find the root cause of this issue.

    This seems to have resolved the issues that we had so long.

    I updated the files and got the improvements in the results and attaching the csv as well as screenshots for your reference.

    I used the no input as well as the test pattern mode scenarios to test with the new files and now the noise floor is not appearing in my results.

    I hope they are as expected, and if not kindly advise.

    I would like to thank you from the core of my heart as this addresses my issues in this test setup.

    Appreciate greatly your help and support.

    Thanking you with my best regards,

    Manjunath

  • 0 in reply to user4632994
    Prodigy 85 points
    Hi Liang,
    Please see the below thresds as the issue I were observing got addressed after Jim's latest advise.
    Hope this would help you too.
    Thanks and BR,
    Manjunath
  • 0 in reply to Manjunath Bevinamar
    TI__Mastermind 44680 points
    Hi Manjunath
    The new results are correct.
    I'm glad this resolved the problem for you.
    Let us know if you need anything else as you do your evaluations of the ADC.
    Best regards,
    Jim B