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.

LMX2595: LMX2595 PHASE NOISE ISSUE

Part Number: LMX2595
Other Parts Discussed in Thread: LMX2594

Hello

I have 2 queries.

1. In the evaluation board datasheet, for the loop filter assembled, the bandwidth mentioned in 285 KHz, but in simulation in PLLatinum why is bandwidth getting simulated ad 49.5Khz... I have attached the 2 images. Please clarify

The second query is same as earlier.... As per simulation file posted above the phase noise is -94 at 1K and 10K offsets... But practically I am measuring 80 at 10MHz ref.... At 100 MHz ref... the simulation and practical results are matching.... What might be the issue at lower PFD.   Please clarify. I am attaching the phase noise of the source and the measured value of LMX2595 along with TICS PRO settings...

  • Hello Narendra,
    our experts for this device are going to clarify this observation for you.
    Best regards,
    Patrick
  • Narenda,

    The EVM assumes a clean 100 MHz input reference, which is doubled to a 200 MHz phase detector frequency.

    In your case, you have only 20 MHz phase detector frequency, which completely changes the loop dynamics; typically for a case like this, you would re-design the loop filter. Note that if you go to PLLatinum Sim and type in 200 MHz for the phase detector frequency without re-designing the filter, the loop bandwidth will increase.

    Also, for the 10 MHz reference, PLLatinum Sim is assuming that it is noiseless, which I doubt is the case. In this case, you can go to the PLL section and enter the phase noise profile for this.

    Regards,
    Dean
  • Hi Dean,

    I changed the sine 10MHz Input to CMOS... Now the simulation and practical results match (practical  degrades by 2 dB).. I have attached the phase noise of the CMOS  source.. Please suggest the best loop filter configuration and values. Where I can get the best phase noise and least spurs at 13.3 GHz with this 10MHz CMOS Input.

  • Hi Narendra,

    I see Dean is supporting you in another post, let's closed this thread.
  • Hi Noel,

    One additional query,

    When using a CMOS source of 10MHz, the phase noise improves but there are additional spurs appearing. I have attached the plots of the spurs at 3 spots 13.05 GHz , 13.3 GHz and 13.55 GHz. Please suggest what can be done to remove these spurs

  • Narenda,

    These spurs look like 10 MHz multiples of your input frequency and the fact that they are impacted by the input suggests they are the Fosc spurs. These are impacted by the input. What the LMX2594 ideally wants to see is low amplitude, differential, high slew rate signal, like LVDS.

    The CMOS has a faster slew rate and this improves phase noise, but I suspect the higher amplitude is causing the spurs. You can do a quick diagnostic of reducing the charge pump current and seeing if the spurs go down. If they do, then you might be able to reduce these with another pole in your loop filter. If they do not (which is what I suspect), this suggests the spurs are going around the loop filter and the biggest thing you can do to impact these spurs is change input format/level/frequency.

    Regards,
    Dean
  • Hello Dean,

    Thanks for the suggestion will try it out and get back.