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.

TCA9517 Vol over Spec

Other Parts Discussed in Thread: TCA9517, TCA9517A, TCA9548A

The TCA9517 datasheet mention the MAX VOL of SDAB is 0.6V. From the picture, we measure the LOW voltage is about 0.76V at the beginning. What reason cause TCA9517 over Spec? That may cause I2C Fail. 

The design is from MCU input pull high to 1.8V to TCA9517 output pull high to 3.3V. Pull high resistor is 10k. 

Ch1: Clk  

Ch2: SDA from MCU  

Ch3: SDA output from TCA9517

  • Do you have a schematic I can view? 

    A screen shot showing just the TCA9517, pull-up resistors, and all other connections of SCL and SDA will be sufficient if you cannot share the entire schematic.

    Can you confirm the full part # of the device? Is it TCA9517DGKR?

  • Thanks for your feedback, I add the schematic of I2C part. The pull up resistors has been changed form 4.7k to 10k. The result remains the same. The full part number is TCA9517DGKR.

  • The schematic shows TCA9517ADGKR?

    Your last post says TCA9517DGKR.

    Is the part on the board the A- or non-A version?

    VOL = 0.76 V (VIL into EP1) when VCC = 3V should NOT result in an I2C failure (VIL = 0.3*VCC).

    The final terminations of SCL_EP1 and SDA_EP1 are not shown on the schematic, so I cannot be sure what else is connected or why this was identified as a failure in communication.

  • It is  non-A version TCA9517ADGKR.
    The final terminations of SCL_EP1 and SDA_EP1 is only connect to the SFP+ optical transceiver.
    We would like to know the reason cause SDA_EP1 over 0.6V(max).

  • How many boards have you built?
    How many TI TCA9517A devices have you tested?
    Is there 100% failure based on multiple boards and multiple units tested?
    If not, can you replace a failing TCA9517 device with a known-good TCA9517 device and resolve the issue?


    I am merely trying to determine if this is a Quality issue or if there is an application-specific problem that requires more debug.
  • Hi, Brian Berner,
    I am co-work with David Wang32 at this issue.

    These are 100% failure base on multilple board and multiple unit.
    Although no all the unit its peak voltage is over 760mV , but all of their Vpk are over 0.6V at beginning.
  • What is the part number of "SFP+ optical transceiver"? I need to know its VOL/IOL characteristics if there is a datasheet available.

    Is there another device in between the TCA9517 and the SFP+ Optical Transceiver? Like a TCA9548A switch?

    If I cannot see what is in between TCA9517 and SFP+ or verify that there is nothing in between them, I cannot be sure the TCA9517 is causing the problem.

    If the SFP+ Optical Transceiver is NOT present but the TCA9517 tried to communicate with it,  what is the VOL level at the B-side of the TCA9517?

    Have you tested the system with and without the SFP+ Optical Transceiver present?

    If there is a Switch or Mux (e.g. TCA9548A) in between, can you test the system without the switch by "blue-wiring" around it?

  • We had compared the waveform of with/without SFP+ optical transceiver the Vpk is same.
    And also blue-wiring to without switch TCA9545, and confirm the result is same.
  • Joe,

    Thank you for your answers, but maybe you misunderstand the reason I am asking these questions:

    I am not necessarily asking questions because I think I know what is wrong with your application.
    Rather, I am brainstorming possibilities that you might want to try.

    If you have already done the work I am asking, and you have taken scope shots, please attach more scope shots that are labeled with the appropriate schematic names and/or part pin names.

    We may see something interesting on another, well-labeled scope shot, but I guarantee I will not discover anything if you reply with "yes" or "no" answers.

    If you have tested >10 parts and you experience 100% failure, it is a good indication that some part of the application is violating the spec of an IC, not that an IC is violating its own spec.

    If you are not able to share more of your schematic and provide more scope shots, I do not have anymore suggestions at this time.

    If you are confident that the TCA9517 is violating its own spec, please contact your distributor to Return the "Failed" devices and they will be directed through our Quality channels back to Dallas.
  • Brian,

    Sorry late to reply.

    I taken the scope as attached, including with/without SFP, with/without swtich chip TCA9545 and provide the schematic of I2C bus.

    Please advise us what is the possible root case. Thanks you.