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.

TXS0108E: SD card interface issue with TXS0108EPWR level translator

Part Number: TXS0108E
Other Parts Discussed in Thread: TXS0206A

Hi,

We are facing sd card interface issue with TXS0108EPWR level translator, Issue is unable to initialize the SD card from FPGA. Then I probed the SD card interface signals and found that the response for from SD card is wrong and then there is no further transactions on the interface.Finally identified the issue at TXS0108EPWR level translator as some glitches at 3.3V level (SD card side), but interface connectivity is staight forward as below.

captured waveforms of level translator (both 3.3V and 1.8V level signals); Channel2 - 1.8V and Channel1- 3.3V level signals

It seems glitch comes after sd_cmd command so sdcard controller (FPGA) consider as response and no transaction if CRC not matched

We are not expecting glitch at one side of level translator, please suggest us to slove this issue.

Regards

Shiva

  • Hi Shiva,

    The TXS has internal One-Shots that get triggered to help have a faster transition rate. If it is falsely triggered, what you are seeing may happen. False triggering can occur if the input is too noisy, output is not properly matched or long traces are being used. Can you provide me with a schematic? Are there any external pullups on the lines? Can you provide scope shots of the input transitioning?

    Thanks!

    -Karan

  • Hi Karan,

    Even PCB trace length also very short as ~20.19mm (795.17941mils), pls refer below snap shot for details.

    Where, U12 - TXS0108E (leel translator) and CON!) - SD card connector.

    Didnt observe the much noise on signals, please refere below wveforms for both 1.8V and 3.3V leel signals.

    Regards,

    Shiva

  • Hi Karan,

    Even PCB trace length also very short as ~20.19mm (795.17941mils), pls refer below snap shot for details.

    Where, U12 - TXS0108E (leel translator) and CON!) - SD card connector.

    Didnt observe the much noise on signals, please refere below wveforms for both 1.8V and 3.3V leel signals.

  • Hi Siva,

    I am not able to see your snap shot. Can you please try to upload it again.

    What do you mean be glitch? Can you show me a scope shot of the output of the translator?

    Thanks!

    -Karan

  • Hi Karan,

    PFA waveforms for both side of level translator.(1.8V and 3.3V)

    glitch/unknown transition on SD-CMD lane

    schemtics:

    Issue: Observed 20pSec glitch/unknown transition immediately after the command from FPGA on SD_CMD pin (at 3.3V level), this glitch transfferd to another side of level translator (1.8V) as siganl level change, which goes to FPGA as sd card response and stopped transitions due to CRC fail.

    We didnt observe any noise on signals and no pullups on board as well.

    Thanks and Regards,

    Siva Kumar. 

  • Hi Karan,

    PFA waveforms for both side of level translator.(1.8V and 3.3V)

    glitch/unknown transition on SD-CMD lane

    schemtics:

    Issue: Observed 20pSec glitch/unknown transition immediately after the command from FPGA on SD_CMD pin (at 3.3V level), this glitch transfferd to another side of level translator (1.8V) as siganl level change, which goes to FPGA as sd card response and stopped transitions due to CRC fail.

    We didnt observe any noise on signals and no pullups on board as well.

    Thanks and Regards,

    Siva Kumar. 

  • Hi Siva,

    It looks like your 1.8V signal goes high, causing the 3.3V signal to go high but then the 1.8V signal goes low, causing the 3.3V signal to go low, creating that glitch. Can you try removing the input series resistor to see what happens? I am not sure which pulses on the 1.8V side are expected and which aren't?

    Thanks!

    -Karan

  • Hi Karan,

    I have removed series resistor and tried, but no luck. Still am seeing this issue. Is there any suggestions to debug further.

    Regards,

    Siva Kumar.

  • Hi Siva,

    Could there be any bus contention happenings? Is there anything else connected to the output 3.3V side? The outputs of the TXS are very weak which will cause it to not be able to over drive any other signals.

    Thanks!

    -Karan

  • Hi Karan,

    SD card signals are directly connected to SD card connector from level translator (3.3V side). Pls check layout and scheamtics below.

    schematics:

    Layout:

    Nothing is connected to level translaotr 3.3V side except SD card connector (refer above scheamtics and layout). so no possibility to occur bus contention at 3.3V level side.

    Appreciate your suggestions to solve this issue.

    Regards,
    Siva Kumar.

  • Hi,

    There is some external noise causing the output of the TXS to switch low. Is there a possibility there is a signal coming in from the SD card side that might have over driven the output? Can you swap out the TXS part? Can you also test this same condition on another board?

    Thanks!

    -Karan

  • Hi Karan,

    Yes, i tried on another two baords and observed same behaviour.

    As i mentioned earlier, nothing is connected at 3.3V side of level translator except SD card connector. So how it can be psosible?

    Could you please suggestions to get rid of this issue.

    Regards,
    Siva Kumar.

  • Hi Siva,

    what is the operating frequency here? TXS0108E is a generic level translator which are not built for SD card application specifically. What are the Vcca/Vccb status during this? 

    TXS0206A supports SD card interface with up to 60Mhz frequency support on the clock pins. could you consider this? 

  • Hi ShreyasRao,

    SD card initialization sequence operation frequency is ~225KHz on CMD pin.

    Voltages are 1.8V@Vcca and 3.3V@Vccb.

    Is there any solution or workaround for TXS0108E level translator issue to check the SD card functionality on existing board.

     

    Regards,
    Siva Kumar.

  • Hi Siva,

    This is relatively low frequency (225Khz) and should be supported by the TXS0108e device given the operating Vcca / Vccb voltage levels of 1.8 & 3.3V. 

    Since the 3.3V signal has a glitch, could you place a series resistor on the 3.3V CMD line to see if it can be suppressed out?