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.

TUSB211: Signal variance (jitter) in TUSB211 eye-diagram

Part Number: TUSB211
Other Parts Discussed in Thread: TUSB212

Hi TI team,

Pls. help to check above pics, it works in Level-1 mode but there are many signal variance (jitter) in the eye-diagram, (level-0 is almost same and no enough margin);

Question is what causes such signal variance (jitter)? how can we improve it?

  • Hi,

    Could you provide a simple block diagram of your test setup? What is the DC Boost selected in this test? 

  • Hi,

    We used LeCroy oscillascope WR640ZI and its Connection Kit to do this test, pls. check the attached file about the test setup.

    We use TUSB211 for USB 2.0 signal redriver, did not use any DC Boost in this test;

  • Hi,

    Have you seen this across multiple TUSB211 devices? The variance in the rising and failing edge rate can be seen slightly in the disable state eye diagram and it seems that TUSB211 is boosting this variance. Is it possible for you to measure at the type-C connector directly or use a Type-C to Type- B cable instead of the converter from Type-C to type-A. Sometimes these converters can cause significant impedance mismatch in the system. Also consider measuring on both sides of the FPC cable. Is this FPC cable fixed?
  • Hi,

    This issue have been seen with a few DUTs.

    We also tried to use a short Type_C to Type_B cable to do same test and got the same result.

    The FPC's impedance is matching as 90ohm, we also tested the eye-diagram at both FPC sides without TUSB211, no signal variance found even eye-diagram was not good, and the end side is a little poor than the start side;

  • Hi,

    Have you tried different EQ settings of TUSB211? Could you provide the eye diagrams for these settings? Could you show the DP and DM waveform as well? THere may be some common mode shift contributing to this issue. What is the pin-out for the FPC connector? I want to make sure there there is sufficient grounding between the two boards. You may also consider trying to use TUSB212. 

  • Hi Malik,

    Following pls. help to check eye-diagram for 4 EQ settings:

    And, is below waveforms of DM/DP ok for you?

    There is entire ground layer on FPC bottom side, the pin out like: GPIO Pin, DN, DP, GPIO Pin...

  • Hi,

    If possible please provide the DP and DM waveform separately. Also is there a common mode choke in the system while testing? Is there a common mode choke in between TUSB211 and the connector?
  • Hi Malik,

    Pls, help to check if attached waves are good enough for you?

    I will provide new eye-diagram without common choke next week.

  • Hi,

    Thank you for the waveform. I was looking for a some asymmetry in DP and DM waveform but could not find any. Please provide the new eye-diagrams without common mode choke when available. In the meantime I will mark this thread as TI thinks resolved. Please reply to this thread to re-open the issue and provide additional eyes. Thanks.

  • Hi Malik,

    Would you pls. help to check the eye-diagram with/o common choke? we are not sure if it is good enough or not? seems it is a little difference with previous one.

  • Hi,

    Does the test report show a failure for jitter spec and have you seen any interop issues? Would you happen to have a s-parameter model of the cable being used in the application?

  • Hi,

    The waveform of D+,D-.The C2(red) is D+ and C3(blue) is D-. Why is the waveform so uneven?

  • Hi,

    It seems there is some noise coupling onto DM/DP. I estimate that its is ~2 MHz noise which my be coming from a LDO or switching regulator in your system. This could also be caused by improper grounding of your cabling or scope probe. I would ensure that your scope probe is properly grounded and the FPC cable has appropriate ground connection. You can also measure your ground to see if the ~2 MHz signal appears, please measure in multiple test point on your board (between re-driver and host, between re-driver and connector, etc.). 

  • Hi Malik,

          Thanks for your suggest. We had found that the jitter is caused mainly by discontinuous ground. There have 2 bead between the GND of the USB connector & the GND of the board (Picture 1)

         Short the 2 bead,get picture 2

         Short the 2 bead and remove common chock,get picture 3

         Do you have any  suggestion? Thanks!

  • Hi Malik,

          Thanks for your suggest. We had found that the jitter is caused mainly by discontinuous ground. There have 2 bead between the GND of the USB connector & the GND of the board (Picture 1)

         Short the 2 bead,get picture 2

         Short the 2 bead and remove common chock,get picture 3

         Do you have any  suggestion? Thanks!

    Picture 1

    Picture 2


    Picture 3


  • Hi,

    I recommend starting with picture 3 and increasing DC Boost level, if possible, to Medium level. With Low DC Boost you could also try higher AC Boost. How are the common mode choke pads shorted together? What is the bead impedance at 100 MHz?

  • Hi Malik,

       The picture 3 is that short the 2 bead and remove common chock,at that time,short the common mode choke pads with jumper wire .

        the bead impedance at 100 MHz is 600Ω,how about  adding the bead with lower impedance?

        -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

         1. starting with picture 3 and increasing DC Boost level, if possible, to Medium level.

        -----to increase what DC Boost level? the DM/DP signal?

              How many is the Medium level?

        ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

         2.With Low DC Boost you could also try higher AC Boost.

         -----to increase what AC Boost level?  What does this do? 

        -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

        Looking forward for your reply~Thanks!

  • Hi,

    A bead with lower impedance should improve performance when the bead is added. Customer is also already using max AC boost. Can customer try using a shorter cable? Please disregard my comment on DC Boost, this only applies to TUSB212. Customer may need to move to TUSB212 for DC Boost option as their seems to be large attenuation in the lower frequency content of the signal.