• State TI Thinks Resolved
  • Locked Locked
  • Replies 11 replies
  • Answers 4 answers
  • Subscribers 102 subscribers
  • Views 3474 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.

AM5718: HDMI interface to TPD12S016

Rogerio Almeida
Mastermind 26315 points

Part Number: AM5718
Other Parts Discussed in Thread: TPD12S016, TCA9800

Customer is designing AM5718 and  looking at the AM5718 IDK but is currently bumping into some issues.

 

Looking at the HDMI portion of the schematic, it seems there’s a logic-levels mismatch between the AM5718 and TPD12S016 – please find below more details:

 

-       TPD12S016

o    Max Low Level Input Voltage (VIL) on SCL_A, SDA_A (Pag. 6 of Datasheet):

§  VIL = 0.082 x VCCA = 0.148V when VCCA = 1.8V

§  VIL = 0.082 x VCCA = 0.271V when VCCA = 3.3V

 

-       AM5718

o    Max Output Low Level Threshold open-drain at 3mA sink current (VOL3) (Pag. 168 of Datasheet)

§  VOL3 = 0.36V in standard mode at 1.8V

§  VOL3 = 0.4V in standard model at 3.3V

 

Same story holds true for other signals like CEC_A, CT_HPD and LS_OE on the TPS12S016. Please let me know your comments,.

  • 0
    TI__Guru**** 393215 points
    Hi Rogerio,

    The key point here is that AM571x maximum output voltages are specified at certain currents (3mA in above case). I am certain that the TPD12S016 sinks no more than a few microamps.
  • 0 in reply to Biser Gatchev-XID
    TI__Mastermind 26315 points

    Biser,

    thanks a lot. in case field/customer does not add new comments, we can close this post in a couple of days.

  • 0 in reply to Biser Gatchev-XID
    TI__Intellectual 2635 points

    Hi Biser,

    thanks for the reply - we already noticed a different characterization of AM5718 output levels, in particular for GPIOs:

     

    - at +1.8V and 2mA , Vol is 0.45V

    - at +3.3V and 100uA, Vol is 0.2V

     

    I agree that the TPD12S016 input pins (e.g.: CT_HPD and LS_OE) sink no more than a few uA; but regarding I2C and CEC_A signal, 10k pull-up resistors are present inside the TPD12S016 on the supply rail (+1.8V in our case).

    Considering that at +1.8V on these pins the Vil for TPD12S016 is about 0.15V the current sink is about 165uA.

     

    Starting from the available data for AM5718:

     

    - Vol = 0.36V for an I2C pin at +1.8V and 3mA current sink

    - Vol = 0.2V for a GPIO pin at +3.3V and 100uA current sink

     

    I'm wondering whether at +1.8V and 165uA current sink on an AM5718 pin it is possible to guarantee the Vil = 0.15V for TPD12S016.

     

    Do you think you could get further documentation on the AM5718 output voltage levels as a function of current sink and power supply?

    Please let me know

    Thanks

  • 0 in reply to Fabio Bucci
    TI__Guru**** 393215 points
    I'm sorry, we don't have additional documentation on the AM5718 output voltage levels as a function of current sink and power supply. This design has been used in several EVM boards without any issues.
  • 0 in reply to Biser Gatchev-XID
    TI__Intellectual 2635 points
    Hi Biser,

    Now, I understand we don’t have additional documentation on the AM5718 output voltage levels as a function of current sink and power supply but need anyway to provide them with solid rationales behind the AM5718 and TPD12S016 connection which is the same as our EVM.

    I cannot justify saying this design has been used in several EVM boards w/o any issues so just trust it.

    Please let me know

    Thanks
  • 0 in reply to Fabio Bucci
    TI__Guru**** 393215 points
    I have asked the factory team to comment further. They will respond here.
  • 0 in reply to Fabio Bucci
    TI__Mastermind 42545 points
    Hi,
    We are reviewing the specs for the AM5718 and the TPD12S016 to try and resolve the discrepancy. We will pull in the design team and try and get you an answer as soon as possible.
    Regards, Bill
  • 0 in reply to Bill Taboada
    TI__Intellectual 2635 points
    Hi Bill,

    it's been a while but I haven't got any feedback yet - customer is upset and is about to replace the IC.

    Do you have any update?

    Thanks
  • 0 in reply to Fabio Bucci
    TI__Mastermind 42545 points

    Hi,

    We have checked and found that the information in the datasheet is not complete enough to resolve the question. It is clear that heavier loads can cause the voltage level to violate the specs of the other part. Resolving this issue will require some additional simulation using the IBIS models and that will take some time. I have requested that we start this process and will update the post periodically on our progress but I don't have an answer for you today. This post is not resolved, I hit the button by accident. 

    Regards, Bill

  • 0 in reply to Bill Taboada
    TI__Mastermind 46460 points

    The TPD12S016 does not meet the I2C standard VIL(max) levels when using 1.8V signaling.  It does have proper input sampling levels when using 3.3V signaling.  Its input thresholds are supply referenced so that when the supply voltage is reduced to 1.8V, the sampling levels are too low.  A level translator such as the TCA9800 can be used to obtain the proper 3.3V signaling levels for I2C.

    Tom

  • 0 in reply to Tom Johnson 16214
    TI__Expert 3075 points

    The following is placed here for archival purposes.

     

    • The AM5178 VOUT levels  are likely spec’d at 3 mA to indicate they can actually sink 3 mA without harm. There are few logical interfaces these days that draw anywhere near that. This is an assumption on my part since I am not familiar with the AM5178 but we have the same kind of specs for fault indicators on some of our power switches.

     

    • The AM5178 Max Pulldown impedance is, indirectly, spec’d. as follows;
      • VDDS = 3.3 V  -->   RPULLDOWN = 0.4/.003 = 133 Ohms
      • VDDS = 1.8 V --> RPULLDOWN = 0.36/.003 = 120 Ohms
      • AM5178 is connected to the I2C “A” ports, which have a 10 k pullup (nom.) to VCCA, per the EC Table. ( see below )
      • With a bit of voltage dividing we can determine the logical voltages at the input to the TPD12s016….. which are well below the threshold.

     

    I understand the customer desire tounderstand the specs pyramid and avoid any negative margins and/or misread signals in their design. Hopefully this can help assure them.

     

    VCCA

    AM5178

    RPULLDOWN

    RPULLDOWN

    (RPULLDOWN +10k)

    VIN_LO

    (mV)

    1.8

    120

    .012

    21

    3.3

    133

    .013

    43