The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> RF Simulators >> pnoise time domain vs sources
https://designers-guide.org/forum/YaBB.pl?num=1242797715

Message started by k_p610 on May 19^th, 2009, 10:35pm

Title: pnoise time domain vs sources
Post by k_p610 on May 19^th, 2009, 10:35pm

I have following query in pnoise :

1. Lets say I am measuring the noise across a capacitor though pnoise "sources". The left side of the plate is always connected to ground. The right plate is connected to ground when clock is 0 and connected to the input of an amplifier when clock is high. As the capacitor is connected to ground when the clock is low, so I can assume that the noise at this phase of the clock is zero.Following are the results of the pnoise "sources' simulation:
1. duty cycle=0.5 total integrated noise ( 40 to 2KHz)=1.8uV
2. duty cycle=0.25 total integrated noise ( 40 to 2KHz)=0.9uV. From above result can I interprete that the noise across capacitor is 3.6uV when clock is high?
2. In "sources" all the noise sources are assumed non-correlated unlike "timedomain". So, if some noise sources are correlated then total integarted noise in "timedomain" is larger than in "sources". Does it mean that "timedomain" noise analysis is more accurate than "sources"? Is "timedomain analysis always more accurate than "sources"?

Thanks

Title: Re: pnoise time domain vs sources
Post by pancho_hideboo on May 22^nd, 2009, 5:22am

k_p610 wrote on May 19^th, 2009, 10:35pm:

I have following query in pnoise :

What vendor's simulator do you mean ?
There are many simulators which have analysis called as Pnoise.

I assume Pnoise of Cadence Spectre.

k_p610 wrote on May 19^th, 2009, 10:35pm:

From above result can I interprete that the noise across capacitor is 3.6uV when clock is high?

I don't think so.

k_p610 wrote on May 19^th, 2009, 10:35pm:

2. In "sources" all the noise sources are assumed non-correlated unlike "timedomain".

I can't understand what you mean.

"sources" is completely same as "timedomain" regarding correlations of noises.

Shaping of noise in time domain cause correlation of noises in frequency domain.
This is true for both "sources" and "timedomain" in Pnoise of Cadence Spectre.
Both "sources" and "timedomain" are no more than small signal analysis in frequency domain.

k_p610 wrote on May 19^th, 2009, 10:35pm:

So, if some noise sources are correlated then total integarted noise in "timedomain" is larger than in "sources".

I can't understand what you mean.

"timedomain" give you Ensemble_Averaged_Noise_Power_Density(time,freq),
while "sources" give you Time_Averaged_Noise_Power_Density(freq).
But meaning or interpretation of frequency spectrum is different between these two.

See the followings.
http://www.designers-guide.org/Forum/YaBB.pl?num=1220976977
http://www.designers-guide.org/Forum/YaBB.pl?num=1207830622
http://www.designers-guide.org/Forum/YaBB.pl?num=1232220194/5#5

k_p610 wrote on May 19^th, 2009, 10:35pm:

Does it mean that "timedomain" noise analysis is more accurate than "sources"?
Is "timedomain analysis always more accurate than "sources"?

I don't think so.