Because of the holidays, TI E2E™ design support forum responses will be delayed from Dec. 25 through Jan. 2. Thank you for your patience.

  • State Resolved
  • Locked Locked
  • Replies 9 replies
  • Subscribers 46 subscribers
  • Views 868 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.

ADS42B49 EVM and TSW1400

Eric Zuber
Prodigy 220 points
Other Parts Discussed in Thread: ADS42B49EVM, ADS4249EVM, ADS42B49, ADS4249

Hi - we have an ADS42B49EVM that we are using with the TSW1400 to capture data.  The A channel functions perfectly, but the data coming in the B channel looks almost like it is the wrong data format...with a sine input, the sine peaks extend toward 0, and the zero crossings switch from max negative to max positive.  The ADS eval is set for Offset Binary (and works on Channel A).  As a side note, we used this ADS EVM with a TSW1200 and I believe both channels worked properly with that system.  Would like some help with getting the B channel to work.


Thanks - Eric

  • 0
    TI__Genius 17320 points

    Hi Eric,

    This sounds strange. Will you please send some screenshots of the time domain signal captured by HSDCPro?

    Is the EVM setup for parallel programming or SPI programming?

    Regards,
    Matt Guibord

  • 0 in reply to Matt Guibord
    Prodigy 220 points

    Time capture using Channel A.

  • 0 in reply to Matt Guibord
    Prodigy 220 points

    Time capture from channel B using same input signal...241MHz input, Fs=240MHz.  EVM is set up for parallel configuration...Bin-LVDS output format selected.

  • 0 in reply to Eric Zuber
    Prodigy 220 points

    Hi - still looking for a solution to the channel B problem above.  Any suggestions?

  • 0 in reply to Eric Zuber
    TI__Genius 17320 points

    Hi Eric,

    Did you verify the EVM jumpers are set to the default positions? See table 1 in the user's guide (link below). Keep JP14 set to offset binary / LVDS.

    http://www.ti.com/lit/ug/slau477/slau477.pdf

    Is there anyway you could check the EVM again using the TSW1200 you mentioned?

    Regards,
    Matt Guibord

  • 0 in reply to Matt Guibord
    Prodigy 220 points

    Hi - Yes, we verified the EVM jumpers are properly configured to default for parallel configuration (other than using BIN-LVDS).  Also tried switching to serial config and using the ADS42B49 USB GUI.  (Attached for reference)  We also re-connected to the TSW1200 and verified that both channels are working properly...using the modified ADC TSW1200 config file as described in the TSW1200 and ADS42B49EVM post.  As a final check, we connected an ADS4249EVM which worked properly on both channels.  As an interesting side note, it seems the ADS4249 and ADS42B49 EVMs use the same FPGA configuration file as it doesn't re-download when changing the selection between the two....probably not an issue as that should be the case, but just mentioning.  Final note - both of the EVMs are Rev C's.

  • 0 in reply to Eric Zuber
    TI__Genius 17320 points

    Hi Eric,

    It sounds like the FPGA timing is marginal, which may not be unexpected due to the firmware being used for multiple boards (as you've found). I do not know how the timing constraints were handled in this firmware.

    The ADS42B49 does have a feature to delay the clock output with respect to the data. This is accessible through SPI in register 0x42. If it's timing then we should be able to meet timing by shifting the clock. Unfortunately too much shift could cause us to capture the odd and even bits on the wrong edge, so depending on which edge the timing is marginal on may determine if this feature is useful or not. Also, the controls in the GUI (v2.1) for the CLKOUT delay are geared toward the ADS4249. You can use the CLKOUT delay, you'll just need to choose the correct combination of ClkOut rise position and fall position to achieve the values specified in the ADS42B49 datasheet.

    For instance, for a delay of 350 ps, choose the following values to program the delay as 0010.

    Rise pos: LVDS +150ps
    Fall pos: default

    Let me know if this works, otherwise I may need to find a working setup here and swap boards with you.

    Regards,
    Matt Guibord

  • 0 in reply to Matt Guibord
    Prodigy 220 points

    Yes - changing timing worked.  Actually just about every different delay other than the default "No Delay" worked fine.  As a quick follow-up, the GUI version and ADS4249 affinity open a new question....the ADS42B49 links to the general ADS42xx GUI, which I was never able to get to connect to the ADS42B49 board.  The EVM users guide references the ADS42B49 specific GUI which I don't even remember how I finally located on the TI web, but it connects properly and has only the specific ADS42B49 registers.  It would be very helpful if the link on the ADS42B49 pointed to the correct GUI...this step took hours of searching and messing around before finally working.


    Thanks - Eric

  • 0 in reply to Eric Zuber
    TI__Genius 17320 points

    Hi Eric,

    Okay, great! We must not have realized that the ADS42B49 and ADS4249 had different register settings, so it was likely linked incorrectly. We'll have to get that straightened out. Thanks for the heads up.

    Regards,
    Matt Guibord