Because of the holidays, TI E2E™ design support forum responses will be delayed from Dec. 25 through Jan. 2. Thank you for your patience.

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.

TLK100: Low temperature startup issue

Part Number: TLK100

Hi, I'm experiencing issues with this component during startup in cold conditions (-23 degrees Celsius).

We have used several hundreds of these in our pcb's for the past 5 years, but during a startup in cold environment we are now experiencing issues. 

It seems that the oscillator starts up, but the controller itself does not - and it does not provide a 25MHz reference out either (to provide a clock reference for the next PHY circuit)

After a reboot, everything seems to work as it should and it takes a couple of attempts to replicate this issue, and somethimes it seems to help to allow the PCB to cool down before a new issue arises. Of the 20 boards in test, only one or two seems to have this issue....

 

Do anyone have any errata information on this IC regarding temperature?

/Best Regards
Eirik Larsen

  • Hi Eirik,

    It seems that the oscillator starts up, but the controller itself does not - and it does not provide a 25MHz reference out either (to provide a clock reference for the next PHY circuit)

    Can you please explain the application setup? If the oscillator is not providing 25MHz reference clock to the TLK100 PHY, it will not function. We are not known of any errata on this issue and never come across this one.

    Also, please note that this part is NRND and there will be limited support.

    Regards,
    Rahul

  • Hi Rahul, 


    To clarify the setup: Two TLK100's are used. #1 are clocked by a crystal oscillator at 25MHz. On a working PCB, CLK25OUT from #1 provides clock reference for #2. 

    However, during these cold start conditions, sometimes - but not always - and some of the PCBs have this issue but not all:

    The oscillator at #1 starts up and oscillates at 25 MHz. The #1 TLK100 itself does not seem to work at all (No LED Activity). On the Clock output (CLK25OUT of #1) there is no clock signal. The #2 TLK100s LED starts, but probably due to the lack of a clock reference it does not start either - LED is constant on. (Only LED_LINK is used)

    BR

    Eirik

  • Hi Eirik,

    From my understanding you are determining if there is clock or no clock with the LED_LINK ? or are you measuring the clock signal?

    Measuring the output clock from the crystal will also determine if the clock is being supplied to the PHY #1.

    Also, when you say that the LED is constant on for PHY #2, when the LED_LINK is configured, it should be ON for Good Link and OFF for No LINK.

    Regards,
    Rahul

  • We are measuring the clock signal with an oscilloscope, but the LED_LINK was only an observation that did not make sense. Here is a sketch with some comments for the given situation. Note that even though the LED_LINK on PHY2 lights up there is not possible to get a connection, but that makes sense since it does not have any input frequency on XIN...

    BR 

    Eirik

  • Eirik,

    Do you have a schematic that you can share? We are a little confused between : oscillator vs crystal? Are you using crystal? Because oscillator only gets connected to XI and not XO?

    Also from the diagram shared above, can you share the waveform of XIN during the problem case? We can check if it meets the requirements.

    Also as part is NRND, do you have plans to move to in-production parts : like pin compatible DP83822?

    --

    Regards,

    Vikram