Trouble Understanding .noisetran simulation results in eldo

Hi Folks,

Problem statement:

Circuit: 1.5v DC ideal voltage source connected to a 1pF capacitor through a CMOS (transmission gate) switch.

Switching frequency 100nS.

Transient sim run for 50 clock cycles and the average of the hold state noise voltage
taken over these 50 cycles.

Run eldo noisetran statement as follows,

.noisetran fmin=1 fmax=1e9 nbrun=50 nmod=2

(also disabled flicker noise components in my Bsim3v3 model files)

The idea is to observe only the thermal noise.

One would expect the output noise on the capacitor to be
KT/c of 1pF at room temp which would be 64uV or close.

The rms out of the simulator gives me 52uV.

Dosent make sense.

Can anyone who knows how to interpret .noisetran sims comment on this.

Thanks,
Mowgly

Re: Trouble Understanding .noisetran simulation results in e

Hi
Plz run the simulation for more time.
u will get KT/C for 1 pf ~ 64uv.