• State Resolved
  • Locked Locked
  • Replies 1 reply
  • Subscribers 20 subscribers
  • Views 999 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.

Fractional Clock Divider

Pete Hanish
Expert 3300 points
Other Parts Discussed in Thread: LMX2581, LMX2541, LMK04033

Does TI offer a fractional divider clocking device?

Specifically, I would like to input an ultra-clean 122.88MHz TXCO that gets divided down to a 10MHz output clock. This 10MHz output clock would serve to synchronize lab equipment to the DUT frequency domain for characterization-quality accuracy.

PLL multiplication introduces too much jitter, so the device must allow divider-only usage. No jitter cleaning is necessary, because the 122.88MHz is already super-clean. The division must be be precisely /12.288 to produce a 10MHz clock. The TXCO output is single-ended and the equipment accepts standard BNC 10MHz reference.

  • 0
    TI__Mastermind 28031 points

    Pete,

    To my knowledge we do not have exactly what you are looking for.  Here is what we do have:

    1.  Synthezier products

    The LMX2581 and LMX2541 can do the fractional division, but the VCO frequecy is too high to go down to 10 MHz.  The minimum frequency is around 30 MHz or so.

    2.   Clocking products

    If you multiply up, and divide down, this is sort of a form of fractional  PLL.  For instance,  GCD(10,122.88) = 80 kHz, but if you multiply up to 1920 MHz and then divide by 192, then you can get a phase detector frequency of 15.36 MHz, which is much better.   Something like the LMK04033 can hit this.

    3.  Standalone dividers

    If you really want the dividers, you can techincally drive one of our products like the LMX2541 in external VCO mode and use the intenal fractional divider and observe the output at the Ftest/LD pin.  You would need to look at N/2 mode to get an even duty cycle.   But there are two major issues with this:

              a.   You need to put some filtering because this will create a large amount of fractional spurs and some of them might be at fairly close offsets

              b.   The divide ratio is too small.  you want a divde of 12.288, which implies a first order modulator and very high spurs.   Also, unless you can tolerate very low duty cycle, then the Ftest/LD pin will be fast pulses unless you use the N/2 mode, but the miniimum divide for a 4/5 fractional divider as in the LMX2541 is 12.  

     

    Regards,

    Dean