• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 82 subscribers
  • Views 1288 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.

CCS/TMS320F28377D: Does the SDFM Module pin need an additional pull-up resistor?

Hui Chen
Intellectual 850 points
Part Number: TMS320F28377D


Tool/software: Code Composer Studio

Hi,everybody:

    When we read RAM with 28377 of the EMF2, the SDFM Comparator will cause a false alarm. We can't observe the noise signal with oscilloscope. When we read datasheet chapter 13.2"The internal pullups can be configured in the GPyPUD register."We found inspiration and Configure the internal pull-up resistors.SDFM work OK! No fault.!

    So I look forward to further understanding of the meaning of this sentence about The internal pullups.Do I need to configure an external pull-up resistor?If  need to configure an external pull-up resistor, what is the value of the resistance?

thank you

regards 

  • 0
    Intellectual 850 points

    For more details, please refer to
    https://e2e.ti.com/support/microcontrollers/c2000/f/171/t/620744

  • 0 in reply to Hui Chen
    TI__Guru 62975 points
    Tina,

    The internal pull can provide a small load on the trace to help dissipate the energy introduced by cross-talk between channels. Are you enabling the pull-ups on the SDFM pins only? Enabling input qualification on the SDFM pin may help too.

    It is hard to say whether you should use external resistors or not. It really depends on the magnitude of the cross-talk. The best approach would be to identify the aggressor signal and improve the layout to eliminate the noise.

    -Tommy
  • 0 in reply to tlee
    Intellectual 850 points

    Tommy

    Thank you very much for your reply.I think your suggestion is very correct. It's great that we expect this noise to be measured.Can the crosstalk noise be observed by an oscilloscope?How do you observe this noise by operating an oscilloscope?In fact, only two modules are disturbed in the 8 SDFM modules. We can't see the noise on the disturbed signal line by oscilloscope or find it different from other normal modules.Could you give me a little more help?

    -Tina

  • 0 in reply to Hui Chen
    TI__Guru 62975 points
    Tina,

    The first step would be to create a simplified test case that can reliably produce SDFM faults when using EMIF. Failures induced by EMIF writes are preferable because they will give you the most control over subsequent experiments.

    With the simplified test, it is then possible to change the PINMUX options of EMIF pins (from EMIF to GPIO input with pull-up enabled) one at a time to see which EMIF signals are causing the failures.

    Errors caused by EMIF reads will require hardware modification for isolation. The signals can either be disconnected at the memory pin or termination can be added to the trace as close to the memory pin as possible (either a capacitor to ground or a series resistor).

    -Tommy
  • 0 in reply to Hui Chen
    TI__Guru 55700 points

    Tina,

    Check whether your system has any source of switching noise (like say switching FET switches). If so, try probing both SDCLK / SDDATA pins along with a signal which accurately shows the switching of FETs etc. You can trigger on these switching noise signals and capcture SDCLK / DATA signals.

    What value of external pull required really depends upon your noise level. General rule of thumb is lower the pull resister value high the noise immunity as it provides a stronger pull up.

    Regards,

    Manoj

  • 0 in reply to Manoj Santha Mohan
    Intellectual 850 points
    Manoj，I have another question.The SDFM module signal is COMS electronic level, and has strong noice immunity. We can't observe the noice signal by oscilloscope, so the noice signal is mV . So why can't DSP correctly recognize the logic of the signal?
  • 0 in reply to Hui Chen
    TI__Guru 55700 points
    Tina,

    In the earlier post, it was said that SDFM caused false trips when EMIF is being used. When inquired, it was said that SD-Cx / SD-Dx traces cross EMIF signals. That is when I suggested that noise might be coupling and might be causing these false trips. I'm not privy to your application / system board design. So, It would be difficult for me to suggest you how to capture noise signal if any.

    Regards,
    Manoj