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.

Jitter Cleaner Question for TI Support

Other Parts Discussed in Thread: LMK04906, CDCM6208, LMK04808, CLOCKDESIGNTOOL, LMK03200, LMK04208

Hello,


We need a clock jitter cleaner (125MHz Input and Output), that also works as a simple Zero-Delay Buffer.

The phase difference between output and input needs to be constant (preferably zero) after each power-up.

Please, recommend what components can we use for this application.

If possible, it would be better if this component wouldn't require any programming after power-up.

Thank you very much,

Mircea

  • The LMK04906 could be used for as a 0-delay jitter cleaner.  However it does require programming after power up.

    73,
    Timothy

  • Hi Timothy,
    Thank you very much for the answer.
    Is there any other part that we could use as a zero delay buffer/clock conditioner, without the need to program it after power up?
    We need to run with a fixed 125MHz input and output.
    Best regards,
    Mircea
  • I'm not aware off-hand. However an MSP430 microcontroller could be used to bootstrap the LMK04906.

    See www.ti.com/.../slac541
    This is from the CDCM6208 product folder.
  • Hi Timothy,

    Can I use LMK04808 in 0-delay mode for 125MHz input and output?

    What if the jitter performance if I use it in 0-delay mode vs other modes of operation?

    Is there a reference design for LMK04808?

    Thanks,

    Mircea

  • Hello,

    The LMK04808 can be used in 0-delay mode for 125 MHz input and output.

    The jitter performance can be simulated using the ClockDesignTool or Clock Architect, while it doesn't specifically support a 0-delay configuration in simulation, this does not impact performance.

    Refer to the Tools & software tab of the LMK04808 product page for various examples in addition to the evaluation board.

    73,
    Timothy

  • Hi Timothy,

    After all this searching, we decided to go with the LMK03200 for our 125MHz, 0-Delay Clock Conditioner.

    We plan to purchase the LMK03200 Evaluation Board and test it before starting board design and production.

    I plan to use OUT_0,1,2 to deliver clean 125MHz LVDS clocks to my module, and use OUT5 for Feedback.

    Questions:

    - Is there any performance difference if I use Internal vs External feedback from OUT5?

    - Can I test External feedback on the Eval Board by removing BALUN B2, installing the appropriate terminations, and running SMA cables between OUT5 and DUT_Fin?

    - Is there any simpler configuration (device) for a 125MHz 0-Delay Clock Conditioner/Jitter Cleaner?

    - Anything else I need to know?

    Thanks,

    Mircea

  • Mircea Bogdan said:
    - Is there any performance difference if I use Internal vs External feedback from OUT5?

    I would use internal feedback because of simplicity.  The reason for external feedback is primarily of you need to route through some external device.  As for performance itself, I expect performance to be about the same, except because of additional external routing, there could be a chance of some increased crosstalk.  You could always route OUT5 for external feedback if desired if it doesn't compromise your layout and still use internal feedback.
     
    The other item about external feedback, the length of the trace now impacts the exact 0-delay relationship between input and output.

    Mircea Bogdan said:
    - Can I test External feedback on the Eval Board by removing BALUN B2, installing the appropriate terminations, and running SMA cables between OUT5 and DUT_Fin?

    Yes, although I didn't think B2 was installed by default on EVM.  Of course you could test with B2 installed if you desired, but the balun would impact timing of input to output clock 0-delay.

    Mircea Bogdan said:
    - Is there any simpler configuration (device) for a 125MHz 0-Delay Clock Conditioner/Jitter Cleaner?


    Not that I'm aware of.  Note the LMK04208 is now up on the web which 6 PLL2 outputs + 1 oscout, similar to LMK04906 but 2.9 GHz VCO.

    73,
    Timothy