Designer’s Guide Community

BackerShu

Community Member

Offline

Posts: 64

What does SSB exactly mean in pnoise simulation?
Nov 18^th, 2011, 12:10am

Hi, guys

I have some uncertain about the Noise Type option in pnoise simulation in Spectre RF.

Specifically, when Noise Type of "sources" is chose, spectre RF will do SSB noise analysis as stated in the Tool menus. My question is what does this SSB mean?

1) conventional SSB (I make this term up): the negative side is folded back. So the result should be the same with DSB.
2) half DSB (also made up by me): the negative side is ommited. So the result should be 3dB better than DSB.

Does it means any one of above, or the real meaning is something else?

Thanks,

Back to top

Thanks and Regards!
BackerShu

https://www.linkedin.com/pub/shu-guanghua/26/b72/711

IP Logged

RFICDUDE

Community Fellow

Offline

Posts: 323

Re: What does SSB exactly mean in pnoise simulation?
Reply #1 - Nov 18^th, 2011, 3:53am

The definition of SSB and DSB NF always causes some confusion.

The SSB and DSB refers to the conversion of the "source" noise that is used to compare to the total noise at the output of the mixer. The NF is the total output noise power compared to the converted reference source noise power.

For DSB NF, the converted reference source noise power from both sidebands is included in the the NF calculation [flo+frf and flo-frf], so the reference noise power that the total noise is compared to is higher because of the sum of the two reference sidebands.

For SSB NF, only the converted source noise power from one sideband is used in the NF calculation. There is still thermal noise from the source at the opposite sideband (if there is no RF filter at the source); therefore, this sideband noise is considered as another SNR degradation component at the mixer output. The total mixer output noise is the same for both DSB and SSB NF measurements, but for SSB the reference noise power is typically 3dB lower due to only considering the reference noise from one sideband therefore the SSB NF is typically about 3dB worse than DSB NF.

Hope this helps.

Back to top

IP Logged

BackerShu

Community Member

Offline

Posts: 64

Re: What does SSB exactly mean in pnoise simulation?
Reply #2 - Nov 18^th, 2011, 12:11pm

Thanks RFICDUDE. This helps a lot, I have some other question to make sure I am fully understand you.

In mixer case, as you stated, the SSB only consider half of the source noise, so the SSB NF is 3dB worse than DSB NF.

Actually, I was doing a VCO simulation, and I want to see the phase noise from pnoise simulation. Since actually there is no "source" in an VCO, can I still understand the result under the same way with that in mixer, meaning the real phase noise would be 3dB worse ? If so, what the "source" mean in a VCO?
If not so, what the real phase noise would be?

Back to top

Thanks and Regards!
BackerShu

https://www.linkedin.com/pub/shu-guanghua/26/b72/711

IP Logged

YCY

Community Member

Offline

Posts: 32

Re: What does SSB exactly mean in pnoise simulation?
Reply #3 - Nov 19^th, 2011, 4:04pm

Hi BackerShu

As far as I know, when you choose "sources" mode in the pnoise analysis
for VCO simulation, you are calculating SSB noise power to the carrier power ratio.
I am not sure what you mean by "real phase noise".
If you mean the phase noise measured by the instrument such as PSA or SSA,
that should be SSB phase noise, whose definition will be the same as that you get from the simulation.

BTW, if you use "modulated" or "jitter" mode to analyze VCO's FM phase noise,
you will get the DSB phase noise, which will be 3 dB higher than SSB phase noise.

Hope this is helpful to you.

Back to top

IP Logged

BackerShu

Community Member

Offline

Posts: 64

Re: What does SSB exactly mean in pnoise simulation?
Reply #4 - Nov 19^th, 2011, 4:28pm

Thanks, YCY!

Some more confusions.

By real phase noise, I mean the measurement results. So to your understand, this should be the same with the phase noise simulation, if I chose SSB option.

Here comes the parts I am confusing.
Confusion 1: What should be the physical meaning of phase noise? is it SSB phase noise or DSB phase noise?

Confusion 2: As stated by RFDUDE, in the mixer simulation, the SSB option would cause 3dB difference with measurement result. So, it comes that in VCO simulation, the result get when SSB option is set is equal to the measurement result, while it's not true with Mixer. Is this really the case in Cadence? I have doubt about this.

YCY wrote on Nov 19^th, 2011, 4:04pm:

Hi BackerShu

As far as I know, when you choose "sources" mode in the pnoise analysis
for VCO simulation, you are calculating SSB noise power to the carrier power ratio.
I am not sure what you mean by "real phase noise".
If you mean the phase noise measured by the instrument such as PSA or SSA,
that should be SSB phase noise, whose definition will be the same as that you get from the simulation.

BTW, if you use "modulated" or "jitter" mode to analyze VCO's FM phase noise,
you will get the DSB phase noise, which will be 3 dB higher than SSB phase noise.

Hope this is helpful to you.

Back to top

Thanks and Regards!
BackerShu

https://www.linkedin.com/pub/shu-guanghua/26/b72/711

IP Logged

BackerShu Community Member Offline Posts: 64	Re: What does SSB exactly mean in pnoise simulation? Reply #5 - Nov 19^th, 2011, 5:25pm Are there some materials covering the explanation for SSB option that I need? I tried to look up this in Spectre RF manual but I can not find it. I'll be appreciated if someone can point them to me.
Back to top	Thanks and Regards! BackerShu https://www.linkedin.com/pub/shu-guanghua/26/b72/711 IP Logged

YCY

Community Member

Offline

Posts: 32

Re: What does SSB exactly mean in pnoise simulation?
Reply #6 - Nov 19^th, 2011, 8:04pm

BackerShu wrote on Nov 19^th, 2011, 4:28pm:

Thanks, YCY!

Some more confusions.

By real phase noise, I mean the measurement results. So to your understand, this should be the same with the phase noise simulation, if I chose SSB option.

Here comes the parts I am confusing.
Confusion 1: What should be the physical meaning of phase noise? is it SSB phase noise or DSB phase noise?

Confusion 2: As stated by RFDUDE, in the mixer simulation, the SSB option would cause 3dB difference with measurement result. So, it comes that in VCO simulation, the result get when SSB option is set is equal to the measurement result, while it's not true with Mixer. Is this really the case in Cadence? I have doubt about this.

Physical meaning of phase noise is the PSD of phase error,
which will be the same for both upper and lower sideband.
SSB phase noise indicates only one sideband is considered,
while DSB phase noise includes both upper and lower sideband,
thus it will be 3dB higher than SSB phase noise.

So, for what you have asked in the beginning of this thread,
SSB phase noise means half DSB phase noise.

When we measure or calculate phase noise of a VCO, SSB phase noise is always used.

Back to top

IP Logged

BackerShu

Community Member

Offline

Posts: 64

Re: What does SSB exactly mean in pnoise simulation?
Reply #7 - Nov 19^th, 2011, 9:55pm

Thanks YCY!

As you state, if the SSB phase noise in Sprectre RF only calculates the upper sideband noise. I agree it should be 3dB lower than DSB phase noise.
But I still don't convince that the measurement phase noise would be the same as that in SSB phase noise we get in Spectre RF. I think the measurement results should be DSB phase noise, which means that when we simulate the phase noise in Spectre RF, we should add 3dB to the result we get. In this way, we can predict the measurement results at best.

Could you please provide some results of the comparison between the simulated SSB phase noise and the measurement phase noise to attest your statement?

Thanks again!

Back to top

Thanks and Regards!
BackerShu

https://www.linkedin.com/pub/shu-guanghua/26/b72/711

IP Logged

YCY

Community Member

Offline

Posts: 32

Re: What does SSB exactly mean in pnoise simulation?
Reply #8 - Nov 20^th, 2011, 12:23am

BackerShu wrote on Nov 19^th, 2011, 9:55pm:

Thanks YCY!

As you state, if the SSB phase noise in Sprectre RF only calculates the upper sideband noise. I agree it should be 3dB lower than DSB phase noise.
But I still don't convince that the measurement phase noise would be the same as that in SSB phase noise we get in Spectre RF. I think the measurement results should be DSB phase noise, which means that when we simulate the phase noise in Spectre RF, we should add 3dB to the result we get. In this way, we can predict the measurement results at best.

Could you please provide some results of the comparison between the simulated SSB phase noise and the measurement phase noise to attest your statement?

Thanks again!

Hi BackerShu

You may take a look at the data sheet of instruments such as Agilent E5052A:
http://www.livingston-products.com/products/pdf/131270_1_en.pdf
SSB phase noise is always measured but not DSB phase noise.

The following thread may also be helpful to you
http://www.designers-guide.org/Forum/YaBB.pl?num=1244213153

Back to top

IP Logged