Has anyone had difficulties reproduce Table 2 in the article? The total noise results from my simulation are significantly lower than the results in the paper. I simply execute "spectre -E tah.ckt" from a command prompt. From the pnoise result, I calculated integrated noise power from 0 to 4MHz then I took square root of the result to get the total noise.

What settings can affect my result this much?

maxsideband      maxacfreq      Ken's total noise      My total noise
5      6 MHz      9.28 uV      10.35 uV
10      8 MHz      12.44 uV      12.15 uV
20      12 MHz      15.49 uV      14.09 uV
50      24 MHz      18.14 uV      14.14 uV
100      44 MHz      19.36 uV      14.17 uV
200      84 MHz      20.13 uV      14.16 uV
500      204 MHz      20.51 uV      14.19 uV
1000      404 MHz      20.48 uV      14.18 uV

Thanks,