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.

TFP410-EP: TFP410-EP design review

Part Number: TFP410-EP
Other Parts Discussed in Thread: TFP410

Hi TI pals

We finished the SMT build 2 days ago, and we have 24 bits DRGB input to TFP410.

How do we confirm that the data has transformed to HDMI without I2C control directly?

The design review had finished before,  we measured the HDMI output, there's no output signal 

  • Ho,
    Have you measured power? How much current?
    Are you using I2C, or ISEL=0? (is correct interface mode set from your source?)
    Is RST asserted after power up?

    Regards,
    Wade
  • Compal TFP410 design.pdfHi Wade

    The power is 3.3V & RST is high

    I'm using non-I2C mode(ISEL=0)

    Please review again as attached.(The '@' means non-pop component)

    Is HDCP needed? The signal is inter-routing, it's not connect to other device.

    The next transformer IC (from ADI) is non-HDCP

  • Ho,
    It looks like DSEL is left floating. I don't think this should cause major issue since it is don't care in the mode you are using, but it should be biased to valid level. If it is oscillating, it can cause extra current and internal noise.

    Otherwise, everything seems okay with schematic.
    Have you verified good clock and data (IDCK+,DE, HSYNC, VSYNC, DATA[0:23] to device? They are capacitively loaded to gnd. Are signals good? (depends on strength and other loading of module driving)
    Have you probed HDMI data lines?
    HDCP is not required.

    Regards,
    Wade
  • Hi Wade
    Is DSEL should be pull down as valid level voltage as you mean?
    I'm sure that the DRGB signal is OK. And the HDMI data is without any change.
    I will probe it again later
  • DSEL can be pulled low or high. It is a don't care in your configuration.
    It just should not be left floating.
    Regards,
    Wade
  • HI Wade

    If RTFADJ is not 510 ohm. Is the output all disappear?

  •   Hi Wade

    Actually, the receiver is a HDMI switch HD3SS215IRTQR. And it's for 1 input,  2 output.

    If C-port can't add dc-couple caps, can I put on both A-port & B-port to cover it?

    Pleas see the attachment.

    Target:

    1. J6 CPU 24 bits DRGB without audio signal to TFP410 transferring  to HDMI.

    2. HDMI to HD3SS215IRTQR

    3. It's all inter-tranfer, there's no connectors

    TFP410 with Switch.pdf

  • Ho,
    I have not been able to locate information on behavior of device if the RTFADJ is not set appropriately. It would be good idea to set to 510 to verify behavior.

    I do not understand your comment regarding DC coupling caps? Do you mean without AC coupling?
    I am not familiar with this device, so I do not know if the ac caps are required.
    Do you have access to the HDMI signals on the TPF410 side of port C? Ie, is the TFP410 driving the HDMI signals?

    Regards,
    Wade
  • Hi Wade

         Sorry, AC coupling is correct. Maybe we should ignore the device first.

    I replied to the wrong(myself) window above. Please see the last I post , 

    the waveform on the TFP410 pins. Because I use the non-I2C control mode, how to

    confirm that TFP410 is working on DE-mode or HV-mode.

    I will ask a new ques. for the switch

  • I believe you are indicating that the HDMI probed signals are on the TFP410 pins, and that the TFP410 has no output.
    I would recommend correcting the RTFADJ and testing to confirm that it is not impacting the outputs.
    Also, how many boards are you validating on this first build? It may be good idea to verify a couple in case of possible damage to components during assembly.

    I am going to be out on vacation tomorrow and monday, so there may be a delay on any additional followup.
    Regards,
    Wade
  • Hi Wade
    We built over 100 boards, and I have tried to verify 5 pcs, it's the same as others.
  • Hi Wade

    Have you come back? Please analyze it as soon as possible

    Review the sch & confirm that it's OK for our purpose. The HDMI output

    still without any voltage changing

  • Ho,
    I am back. However, I do not see what the issue is that is causing system to not work.
    Have you set RFADJ to 510ohm?
    Have you tied DSEL high or low? (I don't think this is related though)
    Have you verified incoming data is valid as well as clock? Verified setup and hold times are met?

    I just was re-reading your prior comments.
    You indicated:
    >The power is 3.3V & RST is high
    >
    >I'm using non-I2C mode(ISEL=0)

    However, ISEL and RST are the same pin. I am assuming that RST/ISEL is held low. Can you confirm?
    Regards,
    Wade
  • Hi Wade

        I got a document "TFP410 DEBUGGING STEPS" from Arrowasia FAE, TI's agent in Taiwan. 

    I tried the step 1 and the output appeared finally, but there is a problem, how to use HV mode?

    because we have two input source, one has DE, H/V sync, clock.

    Another has H/V sync, clock, the difference is with or without DE.

    So, it seems like that I have to let TFP410 being able to transform both DE-mode & H/V mode

    I connected another chip to transform HDMI back to DRGB, and DATA was not appear if the input

    of TFP410 without DE pin.

     

     5226.TFP410 Debugging Steps1 copy.pdf

  • Ho,
    Look at page 13 "de generator".
    I believe that you will need to enable the DE generator if not using external DE. Without I2C, this does not appear possible.
    Can you provide H/V and DE externally?

    Regards,
    Wade
  • Hi Wade

    These two input source both have H/V sync & clock.

    There's only one difference, DE, between them. As you mentioned

    above, the I2C mode become necessary, is it correct?

    How to modify the sch I sent before, and can you provide the registers

    needed to set to our SW team member directly?

  • Ho,

    I am basing this off my interpretation of the datasheet.  I would like to check with someone with more experience with the device to confirm my interpretation.   I will get back with you with more information.

    Regards,

    Wade

  • Ho,
    I believe my interpretation of the DE generation is correct. In order for device to generate DE internally, is to program registers via I2C. The functionality of the DE generation is discussed on page 16. Additionally, the registers are defined on page 18.
    There is some restrictions on capability of the DE internal generation. Please see the errata document: www.ti.com/.../sllz030

    The schematic will need to be modified to allow I2C operation.
    There are options for the design changes. Since moving to I2C mode, all configurations could be handled this way, and leave the device in I2C mode.
    Or, possibly can change in/out of I2C mode with assertion of ISEL. Then the pin strapping pins would need to work in both configurations. It would be more straight forward to handle all register settings via I2C when swapping between two configurations.

    For testing, the SCL and SDA can be wired to a I2C controller to send appropriate register settings. The ISEL would need to be enabled, but would also need to be asserted (or held )to RST state after power up before releasing. This will put device in I2C mode where the appropriate settings can be written to match the DE provided and the DE not provided configuration.

    SCL currently has a 4.7K pullup to 3.3V. SCA does not show biasing on the provided schematic, so it may also need a 4.7K pullup.

    Regards,
    Wade
  • Hi Wade

    I have reworked to I2C mode, and write a command to let PD becoming to 1(0xBD), because PD's default is 0.

    It's OK for the input with DE itself, but it's not work when I change to another input without DE and set DE generator 

    register to 1(0x40), there are some resolution parameters below DE generator, it seems like that I can't set only

    DE_GEN, the default setting of others is unable to work 

  • Ho,

    Please see the "DE Generator" section and the figure below it on the datasheet. The default settings won't work because you need to program the amount of time that DE is active, along with other parameters related to it. Page 16 of the datasheet goes over it pretty thoroughly so it should help you pick the correct settings for your application.

    Regards,
    I.K.
  • Hi I.K & Wade

         SW member is Look back to non-I2C mode, input(DE, H/V sync, clock) to TFP410. Besides I can check the input clock freq is the same 

    as HDMI output clock, how to confirm that the HDMI  data transformation is correct? Is it contains H/V sync? If another  chip receive and 

    transform back DRGB, it would be only DE & clk  or the same as original source   

  • Ho,
    Have you been able to verify the HDMI data with new setup?
    If setup is working, I would like to close this post.
    If you still have questions, please elaborate.

    Thanks,
    Wade