The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Measurements >> Phase Noise and Jitter Measurements >> Eldo RF Phase Noise DSB/SSB definition confusion
https://designers-guide.org/forum/YaBB.pl?num=1133793480

Message started by neoflash on Dec 5^th, 2005, 6:38am

Title: Eldo RF Phase Noise DSB/SSB definition confusion
Post by neoflash on Dec 5^th, 2005, 6:38am

Hi, Folks:

I am using eldo_rf to simulate the phase noise of ring oscillator.

It will generate report as sphi and sphi_ssb. According to definition, single side band noise should be 3db greater than double side band noise.

However, the results are reversed.
db(Sphi_dsb) = db(Sphi_ssb)+3dB;

I believe Mentor have a bug, any one know more information on this bug?

thanks

Title: Re: Eldo RF Phase Noise DSB/SSB definition confusi
Post by vborich on Dec 5^th, 2005, 12:45pm

There is widespread confusion regarding this topic.

Technically, phase noise is short for Spectral density of phase noise. You'll commonly see this denoted as
S_phi(f) in technical literature. On the other hand, designers usually refer to the noise-to-carrier ratio Lambda(f) simply as "phase noise", even though this quantity measures the total noise power at offset f and therefore includes the contribution, however small, of amplitude noise as well.

In the small-angle approximation, and assuming negligible amplitude noise, S_phi(f) is twice the noise-to-carrier ratio at small offsets or, equivalently, 3 dB HIGHER.

Because the noise power in the noise-to-carrier ratio definition is measured in a single sideband, and because S_phi(f) approximatelly equals twice the single sideband noise-to-carrier ratio, and because the noise-to-carrier ratio is usually referred to as "phase noise", S_phi(f) has come to be known as "double-sideband phase noise".

So Eldo-RF is indeed correct.

Your simulator may in fact compute just one quantity and simply scale the other by 3 dB. Or it could take a more rigorous approach and compute both of them directly. It's easy to verify: Sweep your analysis over a broad range until you hit the noise floor, and then plot them on the same graph. You'll see ~3dB difference near carrier and then the plots will slowly diverge as you approach the knee.

Vuk