[question]".sample" command in hspice to cal. samp

[sethtalk]

hi, dear all:
i do noise analysis for fully OTA with cap. feedback,
because the output signal/noise is sampled at freq=30MHz,
i write the commend in hspice code:

.nose v(outp,outn) vd 1
.ac dec 100 1 300Mhz
.sample fs=30x

the summary about noise power shows that the total output noise power is 53uV-rms,but the total output sampled noise power is zero,

why the .sample commend is not work ?

 

[erikl]

Re: ".sample" command in hspice

.ac dec 100 1 300Mhz
why the .sample commend is not work ?

Are you sure the .ac command actually ran?
Just a guess: I think HSPICE understands "300Mhz" as "300milli Hz".
If so, write: .ac dec 100 1 300Meg (or x , as in your .sample command).

 

[sethtalk]

hi,erikl:
thanks for you reply,in fact,i do write 300meg, and the AC simulation is sucessful and the .ac file generated , but in the .lis file, it shows the sampled output noise power =0,

 

[erikl]

Re: [question]".sample" command in hspice to cal.

... in the .lis file, it shows the sampled output noise power=0

For help you should show your hspice netlist - only those parts which are responsible for the simulation and its output (.print, .plot, .measure).
PS: did you write ".nose" ?

 

[sethtalk]

Re: [question]".sample" command in hspice to cal.

hi erikl:
i upload the complete netlist/spice/model. tks!!

 

[erikl]

Re: [question]".sample" command in hspice to cal.

i upload the complete netlist/spice/model. tks!!

Thank you! I can't find anything wrong in your netlists. Your ac.lis reports all the noise info you need - output noise, input referred noise, all the devices' contributions, finally the integrated noise, given as output as well as input referred noise voltage:

**** total output noise voltage = 91.5628m volts
**** total equivalent input noise = 14.8012u

The .sample simulation doesn't produce noise info. If you want noise info for a further frequency interval, you must run another .noise simulation using such an interval.
Have fun! erikl