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.

TDP142: tdp142 issue

Part Number: TDP142

Tool/software:

Hello, 
I am using TDP142 linear redriver in GPIO config mode, which means the gain is configured with resistors attached on EQ1 and EQ0.

There were a few times that I managed to get video on the display, but once I turned off and the on (reset) of the driver board, there is no more video on the display. 
That happened after let' say 20 times, where the PC just detect the monitor, but there is no resolution recognized. The strange thing is that sometimes, the driver transfers video, but it is not stable, and it is not repeatable, I mean once I turn the power OFF, and then ON, there is no longer video streaming. 


In terms of schematic, 
I am attaching here the schematic, if someone could help with reviewing the schematic would be quite beneficial. 

  • Hey Tsvetomir,

    Are you able to capture the AUX log? These inconsistent results are mot likely due to issues in link training. Additionally, try using a shorter cable to see if this resolves your issue. 

    I will have a look at the schematic and get back to you shortly. 

  • Here is the schematic review,

    for the most part everything looks ok.

    TDP142_Schematic_Review_Tsvotomir_RevA(6_25_24).pdf

  • Hello Vishesh, 
    Yesterday the board worked for 10 times, but suddenly stopped again, after resetting the power of the PCB. 

    My source is the DP output of my Nvdidia video card. I assume that it is AC coupled, or at least it should be?!

    However, the PC (windows 10 in my case), every time detects the monitor that is attached to the redriver board. Everytime it detects the new monitor, under display settings of the windows menu, but the monitor is grayed out, no resolution recognized. 

    What are your thoughts? 

    Thanks! 

  • Well, I think I found some workarounds. 

    Since I am working in GPIO mode, with strapping resistor to configure the gain of the linear redriver, I found that after any manupulation on the combination of  DPEQ1 and DPEQ0 I am able to get it work for 5 - 6 times, after power reset. 

    The settings of DPEQ1 and DPEQ0 do not matter in my case. Let's assume that they both are set to high level ( DPEQ1 = 1; DPEQ0 = 1) and every time when I power on the board, the PC only detects the monitor being powered, but it is grayed out and the resolution is not found.... 

    If I just change DPEQ1 = 0, or F (does not matter), or if I do the same to DPEQ0, and then power the board ON, then everything is working fine, just great.....

    But I realized that this is true for 5 to 6 times of resetting the power. After the 6th time, the PC only recognize that a monitor is attached, but without resolution. 

    Then, I again change one of the settings of DPEQ1 or DPEQ0, and then I again can get it work for 5 or 6 times, with power reset, and the same story repeats every time I change the settings of DPEQ1 or DPEQ0. 

    While all these tests, the SNOOPENZ pin is always pulled high, which means it is the snooping is disabled. I tried to enable it, but there is no significant improvement. 

    If I leave the board to work, without resetting the power, it is transferring video for 2 to 3 minutes, then turns off, I meant the video disappear suddenly! 

    Do you have any suggestions why, it is working every time when I change the settings of DPEQ1 or DPEQ0, but only for 5 or 6 power resets, after that I have to again change one of the settings????

    Thanks! 

  • Hey Tsvetomir,

    It looks like there is some issue with your AUX bus.

    The AUXp and AUXn lanes are used to communicate the resolution of the display as well as the speed of the trasnmission.

    Let's assume that they both are set to high level ( DPEQ1 = 1; DPEQ0 = 1) and every time when I power on the board, the PC only detects the monitor being powered, but it is grayed out and the resolution is not found....

    If the resolution is not found, then this means that the sink is not able to communicate to the source via the AUX pins. The detection of the monitor is due to the HPD of the monitor which means that the HPD is working fine. 

    Is there a possibility that repeatedly powering is disrupting the link training process. What is the timing of the power cycling? Please try to capture the AUX log this will give insight into the status  of the link and why its behaving abnormally.

    Using EQ of 1 1 is too high. Its best to tune the EQ and then leave it as is for the cable being used. This is based on the insertion loss of the cable. Additionally, do you have any control of source. Are you bale to tune the swing, pre-emphasis, and de-emphasis.

  • Hello Vishesh,

    The power cycles or plug/unplug the DP cable are let's say in 3-5 seconds intervals. I am using widows with Nvidia video card, so I do not know if I can manipulate swings, and emphasis. 

    On the other hand, having EQ 1/1 is just as an example. I tried almost every combination. Just want to mention that once I change the gain with the resistors, I am able to get video on the display. 

    If I leave it to work, without resetting the board or unplugging the cable, it will work for about 3-4 minutes, really stable and pure video 4k 60Hz is present. BUT, after ~4 minutes of working, it starts blinking and small disruptions/noises are present, then the video suddenly disappear. 

    THIS IS HAPPENING EVERY TIME, AND IT DOES NOT DEPEND ON THE EQ SETTINGS. 

    Then I have to change the EQs settings again to get it work, which makes me think of how when changing the EQs settings can make AUX work properly, but if unplugging the DP cable for ~5 times, it stop working. 

    MY SET UP IS AS FOLLOWS:

    PC(NVIDIA) --> 2-3meters DP cable --> DP sink board(TDP142) --> 30cm eDP cable --> Display.

    Do you have any ideas why this can happen?. 

    I tried to disconnect AUX from the TDP142, even to disable the SNOOP function, but there is no difference. 

    Thanks.

  • Hey Tsvetomir,

    Its odd that the video works well, and then degrades after some time. Are you able to confirm that the NVIDIA video card is able to output stable video? All the TDP142 is doing is applying a positive gain high pass filter on the data lanes to account for some of the signal degradation. If the TDP142 is configured correctly and had video output at one point in time than there is nothing within the TDP142 that will change to degrade the signal. 

    My recommendation is to verify the Nvidia video card is able to output a stable 4k 60Hz video for an extended amount of time, and then also test with another graphics card/ DP output to see if the video fails after a few minutes.

    Best,

    Vishesh Pithadiya