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.

TLK2711-SP: TLK2711A GTX_CLK floating

Part Number: TLK2711-SP

Hi

I am using the TLK2711A in my previous design.

Problem is that now I was told that there is a chance that the clock source of the TLK (towards the GTX_CLK pin) can be powered off while the TLK is still under power.

At this point this mean the GTX_CLK will be left floating !

I saw in the datasheet there is no internal pull up/ down over this pin.

Is this a problem?

Luckily, I made a preparation (which I didn't populate so far) for a parallel (R in serial with C)  termination at the GTX_CLK pin.

For a GTX clock of 100Mhz , can you please suggest which values I can use in the two (R+C) free pads I currently have on board in order to both terminate the clock properly and as well  solving the GTX floating issue?

Thanks a lot

Amnon

  • Amnon,

    I am not sure what you have available. Does one of the C or R have a pad to ground?
    If possible, connect large resistor to use a pulldown on the signal to ground.
    The size might need to be influenced by your clock driver. However, when the clock driver is high impedance, it should not much current to pull it to ground.
    A 33kohm (or even larger) resistor to ground should keep input biased low when undriven and have minimal impact on signal integrity of clock signal.
    To be certain, you could simulate the signal integrity of the clock with IBIS to insure that clock is not being impacted.

    If this answers your question, please click "Verify it as the answer"
    Regards,
    Wade
  • Hello Wade
    Thanks for the very elaborated answer which resolved my issue to the full
    but I wanted to ask again to be sure, this is in did a real issue (even a problem) to leave the GTX pin floating?
    And yes, the R+C is connecting GTX to GND so I understand one solution is to put a 36.5Kohm (this I have already in the BOM) resistor over the R pads and a jumper (0 ohm) over the C pads?
    If so, I will do hyperlynx to verify the clock shape is untacked
    Many Thanks
    Amnon
  • Amnon,
    It is not a good practice to leave an input floating. It can cause inrush current on the input transistors if the signal floats into the indeterminate region. This can lead to reliability problems. It can also cause internal oscillations.
    Reference this logic app note for some additional insight:
    www.ti.com/.../SCBA004

    Regards,
    Wade