phase noise measurement results are different with the phase noise impairment setting to a VCO?

hi

I'm trying to use phase noise measurement simulink model to verify vco phase noise set by impairement, and found they are different.

which one is correct? what shall I do to make them close to each other?

thanks

jeff

Answer 1

[Aabha]

Hi @Jeff Yan,

I understand that you are observing a mismatch between the phase noise set using the ‘VCO impairment’ block and the phase noise measured using the ‘Phase Noise Measurement’ block in Simulink. Here are a few reasons due to which this might be happening:

• The ‘Phase Noise Measurement’ block measures phase noise by analysing the ‘Power Spectral Density’ (PSD) of the phase deviation from a reference carrier. Since this is a statistical quantity, short simulation times might lead to noise PSD estimates.
• The measurement resolution, which can be set by the ‘Resolution bandwidth’ parameter of the block, might be coarse/ not adjusted to match the exact noise profile. 
• The phase noise block adds impairment, but if the noise is measured after other blocks (e.g., modulators), the nonlinearities added by these blocks might distort results.

You can try modifying the parameters mentioned above to get a more accurate estimate of the noise in the system. Please refer to the following documentation links for more information about the ‘Phase Noise Measurement’ and ‘VCO’ blocks: https://www.mathworks.com/help/releases/r2025a/msblks/ref/phasenoisemeasurement.html

https://www.mathworks.com/help/releases/r2025a/msblks/ref/vco.html

I hope this helps.

[Jeff Yan]

hi @Aabha,

thank you for your help!

may i ask another question about the phase noise measurement model?

i noticed that the algorithm mentioned in the ref. doc:

The Phase Noise Measurement block uses the zero crossing points of a signal to measure the phase noise. From the zero crossing points, the phase error (φ)is extracted both at the rising and falling edge of the signal. The captured phase error is used to generate a periodic signal (sin(φ)). This periodic signal is interpolated using a fixed time step proportional to the maximum phase noise frequency offset. The power density of the modified phase noise signal, using the dsp.SpectrumEstimator System object, directly provides the phase noise profile of the signal of interest.

and i don't understand how the algorithm works here? for example, why (sin(φ)) is a periodic sequence if phase error (φ) is a noise sequence. and how does the inerpolation using 1/fmax_offset do here?

because i do find that the sim. result is related with fmax_offset, which i don't understand very well.

thanks,

jeff