• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 23 subscribers
  • Views 337 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.

LMX2594: Large spurs at odd offsets

Josh Merti
Prodigy 190 points
Part Number: LMX2594
Other Parts Discussed in Thread: USB2ANY

I am seeing some large (~ -60 dBc) at strange offsets when using the LMX2594.  The image below is a spur with an offset of 39.281 MHz and is 63 dBc.  This measurement was taken on the LMX2594 eval board with a 250 MHz clock and a Fpd of 125 MHz.  The TICS Pro settings follow the measurement.  Do you have any ideas what this spur is from or how to reduce it?

  • 0
    TI__Mastermind 33660 points

    Josh,

    This is a huge spur, and I'm not sure what's causing it. Nothing about your configuration suggests it should be there.

    Usually when we see very large out-of-place spurs like this, we suspect a few possibilities:

    • The power supply could have some free-running microcontroller inside adding a spur at a very high offset which no one noticed before - try switching the power supply
    • The USB2ANY can sometimes carry noise over from the host PC - try the measurement with the USB2ANY disconnected, after loading register settings
    • Try setting the OUTB_MUX to VCO as well - this shouldn't have a large effect but it's something to try
    • Confirm that the spur is still present even when the carrier is removed - might be an artifact on the spectrum analyzer

    Test these and let me know what you see.

    Regards,

    Derek Payne

  • 0 in reply to Derek Payne
    Prodigy 190 points

    It appears to be the spectrum analyzer.  I tried using a different company's spec an and the spur went away.  I still see the smaller spurs (~ 70 dBc) but I think those are expected.  I read some forum entries that suggested using the MASH_SEED field to lower those spurs.  I had some questions about that.

    1. Is there any way other than trial and error to determine the best setting for the MASH_SEED field to reduce spurs?

    2. Is the MASH_SEED setting consistent for the same fractional value regardless of N (so for example N of 45 with F of 23/25 would have the same setting as N of 46 with F of 23/25)?

    3.  Does MASH_SEED affect the ability to use the SYNC feature?

  • 0 in reply to Josh Merti
    TI__Mastermind 33660 points

    Josh,

    1. PLLatinum Sim can estimate the spurs based on a given MASH_SEED value. Set feature level to advanced and check the spurs tab, there should be an entry for MASH seeds which includes the ability to optimize for MASH seed spurs.
    2. Yes, the MASH spurs are tied to the fractional portion of the N-divider, and the integer divide is essentially independent and does not affect the MASH spurs. That said, there's also minimum N-divide values permitted for a given MASH order, and if the MASH order changes then the spurs will change. You would also have to reset the MASH after changing frequencies to ensure that the appropriate MASH_SEED value is loaded after an N-divider change.
    3. Enabling MASH_SEED requires VCO_PHASE_SYNC, which puts a portion of the channel divider in the included divide and potentially changes the SYNC category. That said, if you are using the fractional divider with non-zero numerator, you're already at best dealing with category 3 SYNC, so in practice this isn't a new limitation caused by MASH_SEED.

      Part of the benefit of using MASH_SEED in the first place is that you can combine it with the SYNC feature to achieve a specific phase offset between input and output, or between two devices with similar configuration but different MASH_SEED values. So it should work with the SYNC feature.

    Regards,

    Derek Payne