• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 98 subscribers
  • Views 574 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.

AM3352: SPI I/O termination

Nomo
Guru 16800 points

Part Number: AM3352

Hello,

We have some questions for the structure of the IO termination on AM3352.
We are focusing on A17(SPI0_SCLK), B17(SPI0_D0), B16(SPI0_D1) and A16(SPI0_CS0) on AM3352BZCZ.

1.Could you share the figure showing the structure of the IO termination of A17(SPI0_SCLK), B17(SPI0_D0), B16(SPI0_D1) and A16(SPI0_CS0) on AM3352BZCZ?

2.In case the above terminations are set to SPI and in case of SPI read, could you tell me the point where the clock signal of SPI is looped back?
  Also, is this point effected by the reflection from the output termination?

The followings are background.

- AM3352 sends a read command to the SPI Flash memory which is connected to A17, B17, B16 and A16.
- Sometimes, the read data is different from the expected data.
- The SPI clock frequency is 48MHz, and the mode is in burst read.

- We've already checked the waveform and the waveform seems to be good (We've used the feature in an oscilloscope)
- However, the data in AM3352 is not the expected data. (For example, the expected data is 0x2E, but the data in AM3352 is 0x5E)
- The frequency of this phenomenon is only 1 time per about 3MB.
- This phenomenon appears only in the case of that the dumping resistance is 0 ohm which is marked by a red rectangle in the following figure.
- However, in case the resistance changes to 22 ohms, this phenomenon seems to be not happened.

We want to know your view for this result.

Best Regards,
Nomo

  • 0
    TI__Guru**** 393215 points
    Hi,
    You query has been assigned to a TI engineer. Please note that response may be delayed due to Christmas and New Year holidays.
  • 0
    TI__Guru*** 125430 points

    Nomo said:
    2.In case the above terminations are set to SPI and in case of SPI read, could you tell me the point where the clock signal of SPI is looped back?
      Also, is this point effected by the reflection from the output termination?

    The loopback happens near the pad, i.e. it is part of the padconf control (rxactive bit).  Yes, the loopback signal will be impacted by the series resistor value.  That series resistor is located close to the AM335x in order to provide the best looking waveform at the memory.  However, if you probe close to the AM335x you will likely see a "knee" in the waveform, possibly even a non-monotonicity.  Increasing that series resistor will generally impact the location of that knee.  By moving the knee away from the transition threshold that eliminates this type of issue.

    Nomo said:
    - However, in case the resistance changes to 22 ohms, this phenomenon seems to be not happened.

    The schematic checklist recommends a 33 ohm resistor on this signal.  A 22 ohm resistor is also reasonable.  I'm not surprised you are having issues with 0 ohms.

  • 0 in reply to Brad Griffis
    Guru 16800 points

    Hello Brad-san,

    Thank you for your reply.
    Your explanation is very clear and useful.

    Best Regards,
    Nomo