I am a relatively new user. The manual adds nothing for me, no examples, no dicussion.
I want to look at the 10kHz out phase noise for a 10Mhz signal. I had thought that the proper way to do this is to use a Delta Pair Marker, with one marker on the peak of the 10Mhz signal, and the other marker on the noise baseline 10Khz away (which is how screen shots of the Rigol DSA815 show it).
But on the Siglent, if I first select a delta pair marker, with one 10Khz away, and then select (under the Marker Function key) the Noise Marker soft-button, the delta pair goes away.
This side-by-side screenshot between Rigol and Siglent only confused me more;
https://www.eevblog.com/forum/testgear/comparison-rigol-dsa815-tg-with-siglent-ssa3021x-spectrum-analyzer/msg965789/#msg965789
Who can teach me the proper steps on the Siglent, for showing phase noise at 10Khz out from the main signal? Which sequence of keys
creates the proper screen display and number?
And finally, when using that Noise Marker soft-button, the screen display is approx -103.37 dBm/Hz, and not the expected dBc/Hz.
Again, I am fairly new at this, with open ears.
Sincere Thanks for any mentoring.
Neal
I am far away from my SSA so I can not make examples with UI.
But I can see it need bit clarify some thing.
When you look noise floor near or far from signal you are testing it is not phase noise what you see. You may get some imagine abot phase noise if take care many things and have experience and knowledge enough so that you can make wise guess main thing you see is phase noise.
What means dBc. What means example -10dBc and what mean example with delta cursors that other marker2 is -10dB related to marker1. If marker 1 is carrier (if we name this signal as "carrier") top and marker2 is sideband we can say marker2 level is -10dBc. But how SA as it is machine can know marker1 is positioned to "carrier" so how it can namethis result as dBc. Semantics yes. Human who use SA is only one who can tell it IS dBc. dB is dB if it is dBm or dB.
And then. What "mystery" is example noise level is x dBc/Hz.
If you have two markers. One is carrier level. Say example its level is -10dBm then you have other marker far away from carrier and we assume that there is random noise. Your RBW is 100Hz. You see marker show this noise level is say example -90dBm. But, because your window what measure this level (resolution band width) is 100Hz you need subtract 20dB then you can say level is -110dBm/Hz (not excatly, there is reasons -later..)
Now your carrier was -10dBm and your noise looking marker is jumping over noise averagely level -90dBm then do bandwidth correction -20dB and we can say it is -110dBm/Hz and now if we name this other marker signal as carrier we can say other marker is -100dBc/Hz related to this other marker (but this is not right value for Noise because our RBW filter is gaussian type). But how SA can name it dBc. SA do not know, only operator know it is The Carrier. So why Rigol name it dBc. Who have told to SA that this other marker is just over carrier signal. How about if both markers are positioned to different levels of noise. Do it then happy told that there is xdBc/Hz. Oh well it have just fooled you like Mr T.
When we say dBc/Hz (and even if we measure using 1Hz RBW) it is not same. Why. You measure using somehow gaussian type filter what IS suitable for sinusoidal like signals but This Noise is NOT. dB/Hz is defined using rectangle filter shape. How to do, there is not real rectangle filter for use. When SA "sweep" it use this same filter over whole span. Through RBW filter is measured sinusoidal type "carrier" and then also some distribution of random like noise where we see just noise. Correction factor is unknown, it need deeply know RBW filter shape and some other things in system so that it can define.
Thumb rule is, it need around 2.5dB correction. SA need do this but only for Noise level.
If we tell to SA that some marker we are using for measure Noise level, it need do this correction what take care about difference between gaussian type RBW filter and noise level definition so that level is what we see if we look it through pure rectangle window filter and also take care about this kind of noise after detector etc Rayleigh like distribution.
And just this is roughly bit over 2.5dB so it need do this correction between these cursors when we tell that we measure levels between sinusoidal base signal and noise. If it do not This correction then it works wrong. If it tell carrier distance to noise, like dB/Hz without c it is right because user can add this c because user of course know what he is doing but SA do not know anything. It can name it dBc in case that user tell to SA that there is sinusoidal type carrier what he is comparing to random noise but only in this case.
This I told now because time ago IRL some people was wondering why one SA measure wrong, even if he take very long average so that noise was nearly like pure line if he measure levels it display wrong numbers. No, it was not wrong. It was time to learn.
SA what he was using did noise correction to other marker but not at all to other marker and not experienced user was totally fooled just due to noise correction and all known RBW filter width correction. Ten times more wide RBW and 10dB.
So, user need know what he is doing and measuring so that wagons do not start pull horse.
As told many times. Keysight renamed HP Application note AN-150 is more than important to read. All these are well explained there and after basics many things also what handle digital IF etc SA things is very compatible with Siglent SSA.
https://www.ab4oj.com/test/docs/5952-0292.pdfIt is really suitable to read. No need understand and read all. It can use many times, first some things and then more and more deep things specially if SA do or display something what is difficult to understand. What are detectors, what is what and how these are best to select depending what are doing.