R&S FSPN26 is used to evaluate the phase noise of LMX2572 evaluation board at 3.139GHz.
It is found that changing the power configuration word will change the phase noise. From pwr = 10 to pwr = 63, the maximum phase noise difference is 7dB.
Whether it is VCO output or divided by 4 output, divided by 8 output, its phase noise will not change with power varying, only VCO being divided by 2 will have this phenomenon.
What I wonder is whether this phenomenon is normal?
Here are the TICS pro setting and measured phase noise with different power settings.