cadence noise simulation results are different by two result display methods

Status
Not open for further replies.

hyleeinhit

Member level 3
Joined
Aug 11, 2008
Messages
58
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,286
Location
U.S.
Visit site
Activity points
1,697
gangs,
I am running .noise analysis in cadence's spectre. Two methods display the simulation results: method 1. plot equivalent input noise. This can be done by: locate spectre simulator menu-->results-->direct plot-->equivalent input noise. A figure will pop out. Then go to calculator's RMS function. The final result is the RMS noise over the bandwidth.
method 2. directly display input referred noise. This can be done by: locate spectre simulator menu-->results-->plot-->noise summary.

The problem is that the result from method 1 is much smaller than method 2. So, in which method the result is accurate.

Attached figure 1 is the oTA simulation circuit
Attached figure 2 is the plot result from method 1. The MAX value is ~4.2uV/sqrt(Hz) at 1Hz, the value is 120nV/sqrt(Hz)@100kHz.
The calculation result by caculator's RMS funciton is 127.9nV/sqrt(Hz) over 1Hz to 100kHz

However, the simulation result from method 2 show that the input referred noise is 40.5uV/sqrt(Hz) over 1~100kHz bandwindth.
 

Attachments

  • simulator circuit.JPG
    64.7 KB · Views: 236
  • Input referred noise-VperSqrt(Hz).JPG
    89.7 KB · Views: 268

To clarify about the plot result from method 1. The plot attached as figure 1 is wrong. It is for bandwidth from 1Hz to 100MHz. I ran the same simulation from 1Hz to 100kHz when I did the calculation with RMS function of calculator, .
Both of the input referred noises by method 1 and method 2 are for 1Hz~100kHz bandwidth.
 

Status
Not open for further replies.
Cookies are required to use this site. You must accept them to continue using the site. Learn more…