• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 20 subscribers
  • Views 451 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:

LMX2592: Calibrating Issue

LMX2592: Spurious output of LMC2592 related to FREQ CAL

Charles Robidart
Prodigy 155 points
Part Number: LMX2592

Hello.  I am adding to an already existing thread.  We are experiencing what appears to be an identical problem in our application of the LMX2592.  Our spur is at 5.12kHz.  The level of the spur is very dependent upon the settings of the FREQ CAL register.  I would be willing to discuss more details of our application.  Can I ask  Mario, or the TI team to continue to resolution of this problem.  It appears to me the previous thread did NOT result in resolution.  Thanks for  your attention, Charles (Skip) Robidart.  Previous email thread below.

https://e2e.ti.com/support/clock-and-timing/f/48/p/881857/3281655?tisearch=e2e-sitesearch&keymatch=lmx2592#3281655

  • 0
    TI__Mastermind 28011 points

    Charles,

    I looked through the other thread and generated the attached TICSPro file for the LMX2592 with our settings for further debugging purposes.

    From this, I note that your fraction is 88000000/500000000, which simplifies to 22/125.  With a denominator of 125, this has no sub-fractional spurs, so expressing this fraction in larger terms probably won't help, and maybe it is what is causing the fraction.

    With a phase detector frequency of 78.125 MHz and a denominator of 125, you should get your lowest offset fractional spur at 78.125MHz/125 = 625 kHz.

    However, your actual fraction is 88000000/500000000.  Now this should give the same result, unless there is a non-zero MASH_SEED.  One way to get a non-zero MASH_SEED is to force it with the MASH_SEED bit, but there could be other ways, like toggling the modulator order.  In this case ,you can get many more spurs. So in short, consider using a smaller fraction and see if this impacts your issue.

    Activating the VCO calibration resets the MASH_SEED bit and I see some discussion that implies that when you re-calibrate, it fixes the spur.  This is why I am suspicious of this large fraction.

    Regards,

    Dean

    LMX2592 Settings.tcs

  • 0 in reply to Dean Banerjee
    Prodigy 155 points

    Dear Dean:

    Thank your for your response.  The description in the referenced thread was meant as a clue.  Our operation is not identical.  We are using the LMX2592 in an Integer N mode.  Our Ref Input is 204.8 MHz, with the R divider set to 89.  The phase detector frequency is 25.6 MHz.   All outputs we "tune" are 25.6 MHz times the N divider, and our N is always even.  E.g. we use output freq  = 5,120.00 MHz = 200 * 25.6 MHz.

    My question back to you is given we are in Integer N does your suggestion apply.  Thanks, Sklp

  • 0 in reply to Charles Robidart
    TI__Mastermind 28011 points

    Charles,

    204.8 MHz/8 = 25.6 MHz (looks like R=8, not 89)

    If you are in integer mode and MASH_ORDER=0 , then this rules out fractional spurs.

    If it's not that, then I see three possibilities:

    1.  You have another noise source on the board, like power supply, or it's coming through the XO.

    2.  There are two frequencies such that the greatest common divisor is equal to your spur frequency.

    3.  Some frequency is going through an output divider.  For instance, maybe you have 204.8 and the channel divider is on, then you can get 204.8 MHz/CHDIV.

    In general, debugging spurs is a process, not a one size fits all approach.  It's trial and error:

    1. Shut off everything off until the spur goes away. 

    2.  Disable all the channel divider segments.

    3.  Change the phase detector frequency to 204.8 MHz for diagnostic purposes

    Regards,

    Dean