• State Resolved
  • Locked Locked
  • Replies 21 replies
  • Answers 8 answers
  • Subscribers 18 subscribers
  • Views 4349 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.

TXS0102: TXS0102 : output signal not proper

Eby Jayan
Expert 2115 points
Part Number: TXS0102
Other Parts Discussed in Thread: TXB0102, TXS0101

HI,

We are using TXS0102 to translate our MDC/MDIO lines from 2.5V to 3.3V . What we observed is that the output (3.3V) side signal  for MDC is not coming proper.

The pulse rises to 3.3V at the start and then slowly decays to 2V at the end of the pulse. The Pulse level was expected to be staying at 3.3V . instead it is ramping down from 3.3V to 2V which is even less than the input 2.5V pulse,

Please reply what will be the issue .

We tried using TXB0102 but it is not communicating 

Regards

Eby Jayan

  • 0
    TI__Guru 55990 points
    Eby ,

    The MDIO lines are open drain and hence TXB is not suitable in your case.
    Do you have pulldown resistor at the output side ? with additional pulldown , the output high voltage cannot sustain.
    if you can send out the schematics , scope shots , it would be better for review.
  • 0 in reply to ShreyasRao
    Expert 2115 points
    HI Shreyas,
    Thanks for your reply. Our MDC lines are having internal pull down at the PHY side . It is the On die pull down . To counter it what can we do ? should we add an external pull up to counter it ?
    So this internal pull down might be the reason why the Output of the TXS0102 is dropping from 3.3V ot 2V right ?

    Regards
    Eby
  • 0 in reply to Eby Jayan
    TI__Guru 55990 points
    hi Eby ,

    Yes , having pulldown or pullup resistors will cause behavioral changes on the output of TXS devices.
    I will send you a link on the app note which will be releasing very soon on the same topic. Changing from a strong pulldown to weak pulldown should solve this issue (>50kohm).
  • 0 in reply to ShreyasRao
    TI__Guru 55990 points
    The app note which was released today ! . This talks about the same problem you are describing.
    www.ti.com/.../scea054.pdf
  • 0 in reply to ShreyasRao
    Expert 2115 points

    HI,

    The pull down we are having is internal to the Chip so we cannot change the value and make it a weak pull down .

    can we add a pull up to counter this pull down ?

    Regards

    Eby

  • 0 in reply to Eby Jayan
    Expert 2115 points
    HI ,
    right now the MDC line is dropping from 3.3 V to 2.V in our board . 2V is the VIH for the PHY chip and hence it is working for now .
    Is there an Chance that this might even drop below 2.0V which will be a non recommended value for us .
    If this will remain on 2V it might be working for us .

    regards
    Eby
  • 0 in reply to ShreyasRao
    Expert 2115 points
    Thanks for the application note . Our waveform is similar to that is given the document with the pull up used in the output lines .

    Regards
    Eby
  • 0 in reply to Eby Jayan
    TI__Guru 55990 points
    Eby,

    In that case where there is internal pulldown , you can remove this chip from the design.
    As 2.5V to 3.3V can be done without the TXS device. The chip is causing further degradation to the signal pulling it to 2V due to the internal pulldown.
    having pullup will degrade the output low of the signal.
  • 0 in reply to ShreyasRao
    Expert 2115 points
    HI,
    The condition is that we cannot remove the PHY chip .Is there any other options that we can try ?

    Regards
    Eby
  • 0 in reply to Eby Jayan
    TI__Guru 55990 points
    Eby ,

    I meant the TXS chip is not required as this is making the signal worse than before since the PHY has internal pull-down resistor which cannot be removed.
  • 0 in reply to ShreyasRao
    Expert 2115 points

    HI,

    Thanks for the reply. The thing is that we have option for two channel level translator on board. If we have another design revision , it will be better for us to have a single line translator for MDIO line since it is bidirectional and connect MDC directly as it is output from the processor right ?

    In that case TXS0101 would be good for us right ?

    In our present condition , will adding any external pull up to the MDC line could do us any favor ?

    also is there any chance that the MDC voltage which is now 2V can go below that ?

    Regards

    Eby

  • 0 in reply to Eby Jayan
    Expert 2115 points
    HI,
    One more doubt , what if we short the MDC line externally from the input and out put of TXS0102 even when the TXs0102 is mounted on the board connecting to MDC and MDIO line ?
    is that acceptable from the perspective of TXS0102


    Regards
    Eby
  • 0 in reply to Eby Jayan
    TI__Guru 55990 points
    Eby ,

    I see the setup as follows currently.

    Driver (2.5V)-> TXS (level shift)-> Receiver (3.3V).

    The receiver can accept 2.5V as a logic high and hence it can work as 2.5v to 3.3V level shift by itself. The TXS is causing the signal to degrade from 2.5V to 2V level which is at the border of operation. I suggest to remove the TXS from the design and interface the MDIO signals directly into the receiver as this below :

    Driver(2.5V) - > Receiver (3.3V)

    If the receiver did not have the internal pulldown , or if it was possible to disable the pulldown , then i would suggest using the TXS. pullup and pulldown resistors are not suggested for the TXS devices .
  • 0 in reply to ShreyasRao
    Expert 2115 points

    Hi ,

    Thanks for your valuable inputs . We are having option to by pass the Buffer with a 0 Ohm resistor .

    We can do one thing , have the TXS0102 on the board and parallely load this 0 ohms resistor on the MDC lines to short the input and out put lines. While the MDIO will have the effect of TXS , the MDC will be shorted to the source. 

    is it acceptable for TXS ?

    Please see the circuit . what if we load TXS and the R621 together and leave R620 open.

    any way we require translation on MDIO line . MDC translation may be omitted as it is input to a 2.5V tolerant pin

    Regards

    Eby

  • 0 in reply to Eby Jayan
    TI__Guru 55990 points
    Eby ,

    MDIO are open drain signals, do you have the appropriate pull-up resistors in the design ?
    Why does the schematic mentions TXB ?
  • 0 in reply to ShreyasRao
    Expert 2115 points

    HI,

    The MDIO lines are having pull at the processor side and also it is internally pulled up by the PHYs

    While using the TXS the MDIO is fine but the MDC is having the issue 

    The TXB mentioned in the Schematic only . we have loaded the TXS part in the board 

    Regards

    Eby

  • 0 in reply to Eby Jayan
    TI__Guru 55990 points

    Eby,

    ok , thanks for getting back.

    Good luck and let us know if you need anything else.

  • 0 in reply to ShreyasRao
    Expert 2115 points

    Thank you for your support and valuable inputs 

  • 0 in reply to Eby Jayan
    Expert 2115 points
    HI
    One more doubt , the MDIO is latched on the Rising Edge of MDC. So will it be still fine with the MDC signal dropping . Because the Rising edge of the MDC is fine and after that only the MDC is dropping .
    So will it still work fine under this condition ?

    Regards
    Eby
  • 0 in reply to Eby Jayan
    TI__Guru 55990 points

    As long as the MDC meets the Vih and the transition rate at the receiver end to recognize as valid high, it is good.

  • 0 in reply to ShreyasRao
    Expert 2115 points

    Thank you for the valuable support