• State Resolved
  • Locked Locked
  • Replies 10 replies
  • Subscribers 34 subscribers
  • Views 687 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.

Original question:

SN65DSI86-Q1: SN65DSI86-Q1 HPD pin voltage

SN65DSI86-Q1: HPD recommended 51k value doesn't meet the minimum 2.2V input spec.

WST
Expert 1015 points
Part Number: SN65DSI86-Q1
Other Parts Discussed in Thread: SN65DSI86

The DisplayPort HPD from the data sheet is:

On my current design I have a 51.1k resistor tied to the HPD pin. The other side of the resistor is tied to pin18 of a DPort connector and is sitting at 3.26V. I measure 1.68V at the HPD pin of the IC.
This is consistent with a value of about 50k for the internal resistor. This seems to be working with my system, but it doesn't pass the threshold requirement of 2.2V for Vhpd_plug.

Using the suggested 51k nternal resistor and external 3.3V with a minimum value of 51k internal you would have only 1.65V at this input. The Vhpd_plug says 2.2V is the threshold.
Going the other way, with 69k internal you would still only have 1.9V at the input.

If the 3.3V was increased to 3.6V, the value with internal 51k would be 1.8V and with 69k would be 2.07V.

I must be missing something here. I'm assuming the maximum for this input is 3.6V, so it seems like a smaller series R should be used to meet the 2.2V spec.

By my calculations, with a 3.3V HPD and lets say 3% low 3.2V at the HPD pin (18) of the connector, then to get a guaranteed 2.2V you would have:
with 51k internal then Rext = 23.2k (maximum)
with 69k internal then Rext = 49k (maximum).

I'm thinking about lowering the R value to 21.5k to guarantee the 2.2V level as desired.
With my measured 3.26V that would give me 3.26*(51/(21.5+51)) = 2.3V and with 69k internal I would have 2.49V. These are both over the 2.2V and well below a 3.6V maximum.

If I do the calculations assuming a 5V HPD signal at pin 18, then this works out better, 51k internal has2.5V at the IC and 69k internal would have 2.875V.
Is this why 51.1k is suggested?

The monitor I currently have attached only puts out 3.3V and I think the spec is a minimum of 2.4V at pin 18.

Is there a value I could use for Rext that would cover 3.3V HPD levels at pin18 and still protect with 5V levels?
If I use 21.5k, then with 69k internal I would be at 4.2V at the pin and I think that is too high.

Maybe there is another way to handle this, and maybe 1.68V is enough.

  • 0
    Expert 1015 points

    Hmmmm.... in looking at this more, there is a note for the pin that says 'Measured at 51k series resistor'. Maybe this is the same as the voltage at pin 18, the input side of the resistor. [I wish the spec sheet was clearer].  If that is the case, then I am fine with a 3.3V HPD at pin 18, and this matches more closely the 2.0V min detection spec I saw for this pin (see below).

    This would be then @2.2V with 51 internal and 51 external 0.9V at the SN65DSI86 pin, and with 69 internal and 51 external 1.26V at the SN65DSI86 pin. 
    If the spec is really referring to the other side (the connector side) of the 51k external resistor, then 0.9V must be enough at the input pin to say HPD is high.
    Connected to a 5.3V pin with the 69k internal resistor the voltage at the SN65's pin would be 3.0V, so still below a 3.6V limit.

    I'm guessing I was reading the spec sheet wrong and if I connect to a 3.3V HPD level at the pin I should be fine.

    Thanks for any help!

    Something I found for this pin's levels:

    Table 4-39 Required Detection Levels for Hot Plug Detect Signal

    Item                                                     Value

    High voltage level (Source)                 Minimum 2.0 Volts, Maximum 5.3 Volts
    Low voltage level (Source)                 Minimum 0 Volts, Maximum 0.8 Volts

  • 0 in reply to WST
    TI__Genius 13310 points

    Hello, WST,

    The HPD input requires a 51-kΩ 1% series resistor.

    Thanks,

    Zach

  • 0 in reply to Zach Dryer
    Expert 1015 points

    I have that, 51.1k, but I was just confused about the specification. Is the 2.2V spec'ed at the 'input' side of the resistor (the one connected to the display port interface) or the SN65DSI86 pin? I'm assuming because of previous discussions that it is at the connector side.

  • 0 in reply to WST
    TI__Guru**** 178126 points

    WST

    The VHPD_PLUG spec is at the DSI86 pin.

    HPD has a minimum of 60-kΩ ±15% internal pulldown resistor, so you can use a series resistor to meet the VHDP_PLUG requirement. The 51kOhm is only an example that actually has been implemented and worked, however, you can use another resistor value with the voltage divider network.

    Thanks

    David 

  • 0 in reply to David (ASIC) Liu
    Expert 1015 points

    Can you read through me original question. Your answer doesn't make sense. The VHDP_PLUG value is NOT met with a 51k resistor if the 2.2V needs to be at the pin and the HPD pin on the connector is at 3.3V. I suspect the VHPD_PLUG spec is at the other side of the 51k resistor; that is the only way it works.

  • 0 in reply to WST
    TI__Guru**** 178126 points

    WST

    My point is that you don't have to use 51k. You can use another resistor value to meet the 2.2V requirement. 

    So 3.3V x 60 /(R + 60) = 2.2, with R then being 30k. 

    Thanks

    David

  • 0 in reply to David (ASIC) Liu
    Expert 1015 points

    And I was saying above, that if I did that, with say a 21.5k so that I could meet the spec with a 3.3V HPD at the connector, then if there was 5V at the connector from an old monitor (I think 5V is allowed), then the voltage at the IC pin would be worst case (with the 69k +15% value) of 4.2V, too high for the IC. There is no value of resistor between 21.5k and 51k that works for both 3.3V HPD at the connector and 5V HPD at the connector. 51k doesn't even work with 3.3V at the connector, you only get 1.65V at the pin.


    Also, I found this earlier comment by another TI support person that disagrees with your VHDP_PLUG location:
    (1) SN65DSI86: the HPD pin voltage confuse - Interface forum - Interface - TI E2E support forums
    I'm thinking they are correct because of the "Measured at 51k series resistor" in the datasheet.

    I'm just wondering who is correct. If need be, I'll have to add a buffer from the connector and then divide the output, or just let the maximum be 3.3V at the pin [Is that OK? - the datasheet doesn't say]. Again, the 2.2V makes more sense to be related to the connector, since this matches the HPD voltage spec of HDMI/DPort of 2.2-5.3V.
    If that is incorrect and the maximum voltage on the HPD pin can only be 3.6V or less, then the 21.5k would be ok. But I still don't see how a 51k works or why the datasheet would suggest that.

  • +1 in reply to WST
    TI__Guru**** 178126 points

    WST

    I double checked on the HPD implementation and you are correct. the 2.2V spec'ed in the DSI86 datasheet is on the connector side, not at the DSI86 pin side.

    Thanks

    David

  • 0 in reply to WST
    Expert 1015 points

    OK, I ran an experiment to resolve the question.
    1) The pin description for this pin explicitly calls for a 51K resistor

    2) I lifted the DPort connector end (the end NOT connected to the IC) of the 51K resistor and connected it to a DC power supply that was initially set at 3.3V
    3) I then used the I2C bus to read register 0x5C, the HPD register of the SN65DSI86. I had a monitor plugged in, but its HPD line was disconnected.
    4) I used a DMM to measure the voltage at the IC pin and with the supply at 3.3V, the IC pin measured 1.68V and the HPD register had a value of 0x10 (detect true).
    5) I then lowered the voltage of the supply to 2.2V to simulate a monitor that had a lower HPD voltage. The voltage measured at the pin was 1.14V, but the HPD register still had a value of 0x10. NOTE: This is well below 2.2V at the IC pin.
    6) I then continued to lower the supply voltage, and when the supply was <1.55V (the IC pin was now at 0.8V), the HPD register finally went to 0x00. It seems like that is the threshold.
    7) I then powered up the board with the supply set to 2.1V (1.09V at the IC) and the register was 0x10, so there isn't a power-up issue with this low level.
    8) The earlier answer from the link above (copied below) seems to be correct, the 2.2V and 0.8V levels are before the 51.1k ohm resistor, at the DPort HPD pin 18 of the connector.

    Using the 51K resistor with a 3.3V HPD level gives 1.65V (at the lowest level of 51K internal) and that is > 1.2V actual pin threshold. It also protects the pin from high voltage if connected to a 5V HPD level (which may not exist) since the highest voltage would be around 2.9V.
    I think the datasheet should be clearer as to the Vhpd_plug location, say "Measured at the connector side of the 51-kOhm series resistor" or something like that.

  • 0 in reply to David (ASIC) Liu
    Expert 1015 points

    Thank you for your answer, I ran the experiment before you posted and was just about to post to confirm my thoughts. Now it makes sense.