• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 21 subscribers
  • Views 447 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.

LMX2485: Drawbacks of exceeding OSC2X maximum frequency

William McAuley
Prodigy 10 points
Part Number: LMX2485


Tool/software:

Hi,

While experimenting with the LMX2485 settings to improve phase noise, I found that enabling OSC2X gave significant improvement for the same F_PD. Only problem is our input frequency is 40 MHz and the datasheet specifies 5 - 20 MHz for when OSC2X = 1. What are the expected downsides/risks of exceeding this range?

Thanks.

  • 0
    TI__Guru** 111590 points

    Hi William,

    The risk is you may not get every device's OSC2x function at 40MHz, or you may have problem at different temperature. 

    You said enabling OSC2x but then set R=2 to get the same fpd and see phase noise improvement. Sounds to me that it may be a slew rate issue. Were you using a sine wave or clipped sine wave reference clock? Can you try with a square wave clock? 

  • 0 in reply to Noel Fung
    Prodigy 10 points

    Hi Noel,

    Thanks for the reply. The input to the OSCin pin is a sine wave. Trying with a 40 MHz square wave does not improve the phase noise (although to be more exact, I'm measuring the RMS phase error not the phase noise directly), and setting OSC2X = 1 then doubling R (from 10 to 20) still gets an improvement!

    This is a capture of the 40MHz square wave. The duty cycle isn't 50/50 which I thought would cause issues when using OSC2X, but apparently it performs better?

    Is this slew rate sufficient? Do you have any other ideas for this behaviour?

    Thanks.

  • 0 in reply to William McAuley
    TI__Guru** 111590 points

    Hi William,

    Could you take a phase noise plot? That will help identify if the increase in jitter is due to in-band or far-out noise.

  • 0 in reply to Noel Fung
    Prodigy 10 points

    Hi Noel,

    I'm a bit limited in equipment, so hopefully this is good enough. Further out looks the same between cases on my equipment. Weirdly the power of the VCO output increased when using the square wave, but I wouldn't have thought changing the OSCin waveform would affect the output power?

    Sine wave, OSC2X = 0:

    Sine wave, OSC2X = 1:

    Square wave, OSC2X = 0:

    Square wave, OSC2X = 1:

  • +1 in reply to William McAuley
    TI__Guru** 111590 points

    Hi William,

    Loop bandwidth is approx. 1kHz, so the far-out phase noise is dominated by the VCO.

    The quality of the reference clock (phase noise, slew rate) will affect in-band phase noise.  Set your spectrum analyzer span smaller so as to zoom in to view the in-band details. 

    Duty cycle of your 40MHz is too bad that we should not use OSC2x. The phase noise improvement is probably due to a smaller N-divider when you enable OSC2x. N-divider contributes in-band noise, smaller N returns less noise. 

  • 0 in reply to Noel Fung
    Prodigy 10 points

    Hi Noel,

    Turns out while the 40 MHz input to the LMX2485 now had a decent slew rate, the input to the buffer driving that reference had a low slew rate. Increasing the slew rate into that buffer greatly improved the phase noise.

    Thanks for your help.