• State Resolved
  • Locked Locked
  • Replies 11 replies
  • Subscribers 35 subscribers
  • Views 1888 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.

TPD12S520RMNR DISTORTING SIGNAL

WhiteHorse
Genius 4430 points


Hi e2e,

We have a customer using the TPD12S520RMNR.  They say that when they connected their device to an HDMI source it would not work.  After they removed the TPD12S520RMNR it started working, Then they noticed that when the TVS is in the circuit it attenuates the signal level by around 70%.  Please see the attached image files with scope plots and schematic.  Can you help us determine what is causing this problem please?

Thanks for your help.

Regards,

John Wiemeyer

Schematic

1401.HDMI ESD Schematic.pdf

[

  • 0
    TI__Guru 71526 points

    Hi John,

      One thing immediately obvious to me is that pins 6, 9, 10 and 11 are all connected together to a 0.1uF cap to GND.  Only pin 11 should be connected this way, pins 6, 9 and 10 should be connected directly to GND.  If you refer to FIg 1 schematic in the datasheet, you'll see this creates back-to-back diodes on the signal pins and indeed you will see such a reduced signal.

    ~Leonard  

     

  • 0 in reply to LeonardEllis
    TI__Genius 12190 points

    Hi John,

    Leonard is correct in his assessment.

    Regards,

  • 0 in reply to Guy Yater
    Genius 4430 points

    Thanks for your help and quick response. I have let the customer know. 

    Regards,

    John Wiemeyer

  • 0 in reply to WhiteHorse
    Genius 4430 points

    Hi e2e,

    Customer came back to me and said they gave me the wrong schematic originally and provided a new one with the correct connects for pins 6, 9, 10 & 11.  It is attached.  Any other suggestions for debugging this problem?  The customer asked me to confirm that the TPD12S520RMNR does not add significant distortion as figures 9 and 10 in the data sheet indicate.  I will ask the customer if they have tried replacing the part with a new one.

    Thanks for your help.

    Regards,

    John Wiemeyer

    1680.Revised Schematic.pdf

     

  • 0 in reply to WhiteHorse
    Genius 4430 points

    The customer confirmed that they did replace the device with a new one and it didn't solve the problem.  Please give me some sugestions of things we should check.  Thanks for your help.

    Regards,

    John Wiemeyer

  • 0 in reply to WhiteHorse
    TI__Genius 12190 points

    Jim,

    Due to the signal integrity's dependency on PCB layout, can we have a screenshot or Gerber files showing the layout from the connector to the controller?

    Regards,

  • 0 in reply to Guy Yater
    Genius 4430 points

    Hi Guy,

    Sorry for hte delay.  The customer was in a hurry and could not take time to provide Gerber files or screen shot so that we could furhter debug.  He changed his design to use the TPD12S520DBTR from the TPD12S520RMNR and it is working now.  Customer says when they get time they will send Gerber data for the TPD12S520RMNR version.

    Thanks for your help.

    Regards,

    John Wiemeyer

  • 0 in reply to WhiteHorse
    TI__Genius 12190 points

    Hi John,

    Thanks for the update. The DBT package is much easier to layout due to the flow through packaging style. Perhaps that is an indicator that the issue was layout related. The silicon is the same with the RMN, just a more difficult layout due to the compactness. Its good news they went with the DBT and thanks for the update.

    Regards,

  • 0 in reply to Guy Yater
    Genius 4430 points

    Hi Guy,

    Now that the customer has their system running they need to go back to the smaller package for production.  They did a screen shot of the layout and it is attached here.  Would you take a look and let me know if you see any problems.

    Thanks for your help.

     

  • 0 in reply to WhiteHorse
    TI__Genius 12190 points

    John,

    The problems I see are the stubs to TPD12S520RMN (TVS) from the protected lines being too long. This minimizes the protection by the TVS because the current is not being forced to the i/o pin of the TVS. This will also severely degrade the signal integrity (SI) as these stubs are longer than λ/10, thereby being either capacitive, inductive, a short, an open, or some mixture of those. By my calculations, for 2.4 GHz in FR-4 λ/10 = 5.7 mm. One of the stubs is 10 mm long. These will definitely degrade the SI, as well as provide little ESD protection.

    The RMN package does require stubs. The following design screenshots, shown in the TPD12S520EVM User’s Guide, has stubs but they are only 0.14 mm. This removes any impact on SI and greatly helps to force ESD current to the TVS i/o pin.

    Top (blue) and Bottom (green) Layers (showing Top Layer on top):

    Top (blue) and Bottom (green) Layers (Showing Bottom Layer on top):

    The connector in the upper right represents the HDMI transmitter. I hope the customer can implement these changes into this design as the one you have shown has some problems.

    Regards, 

  • 0 in reply to Guy Yater
    Genius 4430 points

    Thanks for your detailed response Guy!

    Regards,

    John Wiemeyer