• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 21 subscribers
  • Views 1975 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.

TXS02612: abnormal CMD signal

Bruce li
Genius 5580 points
Part Number: TXS02612

Dear TI Experts,

My customer are debugging TXS02612 for accessing TF card, and find abnormal 2 level voltage level on CMD line, seems one is 1.8V and the other is 3.3V. This issue existed in all TXS02612 boards at all time

They suspect this abnormal CMD signal has caused small portion of their product cann't boot from TF card.They tested all other 5 DATA/CLK line of TXS02612, and the signal qulity are all clear.

From the datasheet, we can see the CMD can work in Open-drain and Push-pull mode, customer have tried to add Pull-up/down resistor, yet the phenomenon is consistent...

Below is the waveform and schematics about this case, could you kindly help to review and give some advice? Thanks a lot.

  • 0
    Prodigy 110 points
    Hi Bruce,
    My guess based on the limited information you've provided is that the CLK line has >70pF of load capacitance.

    When this is the case, the one-shot in the TXS translator doesn't have enough drive strength to get the output above the turn-off threshold of the channel's nFET, so when the one-shot turns off (after just a few nanoseconds), the output will match the input voltage (~1.8V). It looks like longer pulses are able to reach 3.3V, but it takes a while to charge the parasitice capacitance on the load (note the slope from the ~1.8V level up to the 3.3V level).

    When the holidays are over, I'm sure one of TI's experts will comment for you.

    You might want to add some details about the load, and get a screeshot that shows both the input and the output of a channel (just the output doesn't really tell anyone anything useful....)
  • 0
    TI__Guru 62170 points
    Hi Bruce Li,

    The TXS has internal pullup resistors within them to help in translation. The pullups are mentioned in the datasheet page 4.
    It is either 40k or 4k depending on the signal going high to low or low to high respectively.

    In the schematic shown, I see there are external 10k pullups on all the CMD and data lines. I would advise to remove them from the system.
    What is the frequency of operation? According to the datasheet specs, the device is rated for up to 60Mhz (under rated loading conditions).
    There are TVS diodes D6-D12 connected on each line. This will add to the line capacitance impacting the switching.
  • 0 in reply to ShreyasRao
    TI__Guru 62170 points
    I would also consider removing the series 22ohm resistors to see if it can improve the switching response. The series resistors in combination with the TVS diodes input capacitance can result in a low pass filter to slow down the output response.
    As Rigor suggested, please send any other details, if possible.
  • 0 in reply to ShreyasRao
    Genius 5580 points
    Dear Rogor and Shreyas,
    Happy new year! Thanks a lot for your support on this case. Below are more infromation and customer's further test results:
    1) input signal is from processor -- ZYNQ7000(xilinx)'s GPIO directly
    2)The communication frequency of TF card are 25M and 50M, similiar result under different data rate
    3)Customer tried to remove TVS, remove 10K pull-up resistors (also tried to change to1K), remove 22ohm resistors. This will make faster rise time , and some TF card can be booted again, yet they can still see the 2 level voltage level on the CMD line, simaliar signal as before.

    Above all, customer want to know if there is any potential risk about the CMD line abnormal signal?
    Also, they only found this case on CMD line, other 5 DATA/CLK line are fine. They want to know,
    are there any differences between CMD line and other DATA line? I'm appreciated if you can provice further more detailed structure and spec description about CMD line inside TXS02612. Thanks a lot.
  • 0 in reply to Bruce li
    TI__Guru 62170 points
    Hi Bruce,

    "Above all, customer want to know if there is any potential risk about the CMD line abnormal signal?"
    Since there are cards which cannot be booted and this is attributed to the CMD line, this will be considered as risk.

    As expected, the rise time of the CMD lines has become faster, but since the different voltage levels or glitches exist, i suspect the driver device is sending them and not due to the TXS device itself.
    Please send both the input CMD signal and the output CMD signal on the same scopeshot, if possible.
    There are no difference on the CMD and data ports internal architecture as you can see on the datasheet page 4.