• State TI Thinks Resolved
  • Locked Locked
  • Replies 1 reply
  • Answers 1 answer
  • Subscribers 20 subscribers
  • Views 739 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.

LMX2571: Fast changing modulated spurs

Paul Staat
Prodigy 10 points
Part Number: LMX2571

Hello,

I am using an LMX2571 on a costum board to synthesize frequencies from 600 to 900 MHz. I am facing an interesting issue with fast changing spurs on the synthesizer output which I can not explain so far. These spurs appear to be symmetrical around the carrier. They change in distance pretty fast and are getting stronger when closer to the carrier. (Up to around -50 dBc) At first, I thought it might be a problem with the loop stability but I achieve lock and decent phase noise noise performance.

Some basic info on the setup:

  • 40 MHz reference and PFD frequency
  • N = 60, num = 9586981, den = 2^24, VCO divider = 6, output frequency = 807,619 MHz
  • second order DSM
  • FSK modulation disabled
  • no dithering
  • PLL is locked

Here are my observations:

  • No improvement with different VCO dividers (not a crosstalk issue I guess)
  • Does not improve with differenct reference / PFD frequencies
  • "Smearing" of spurs in the spectrum increases with DSM order (for first order modulator, discrete spurs are not detectable anymore but close-in noise increases drastically in short bursts)
  • Very small changes in the fractional numerator do not yield improvement, changes in the order of +- 500 mitigate the spurs
  • The spurs will be reduced when decreasing the loop bandwidth (I have been reducing the charge pump current) but are still occuring.
  • Changes of PFD_DLY do not help neither

As the spurs only occur / change in dependence of the fractional numerator and the DSM, seem to be independent of the PFD frequency and are being filtered by the loop filter, I suspect the fractional engine to be the reason. Is my analysis reasonable? Can you reproduce the issue? I would like to understand the machanism causing this behaviour, so that I can predict and avoid these frequencies.

Best regards,

Paul

  • 0
    TI__Mastermind 28011 points
    Paul,

    I agree that the fractional engine is likely to be the issue. I suspect this is because you have a large irreducable fraction. When you do this, you are actually pushing the spurs to a much lower frequency. For instance, if you are looking at the E5052, sometimes it can look like some sort of movement. This fraction looks very close to a fraction of 4/7. In fact, I would try just 4/7 and see if that clears things up. Fractional spurs tend to be worse nearest integer boundaries. Now this is not an integer boundary, but they are also bad near simple fractions. For instance, fractions near 1/2 are the 2nd worst case.

    As the denominator of this device is programmable, you may get better results if you can find a value that simplifies a little more if it turns out that reducing the denominator fixes these spurs.