What is the purpose/function of having multiple Noise/Fred pairs in independent sources?

The noise can be coloured (it doesn't have to just be white noise) - which is why you can specify multiple noise/frequency points. It's also used in noise analysis as well as transient noise.

The best thing would be if you could share the input.scs from the two examples - so I can see the setup, the values, and the analysis statements. Then I can probably tell more easily what's right and what's wrong.

Regards,

Andrew

Hello Andrew,

Andrew Beckett said:

The noise can be coloured (it doesn't have to just be white noise) - which is why you can specify multiple noise/frequency points. It's also used in noise analysis as well as transient noise.

Thanks for the confirmation.

The idc I used, I use the same schematic, just connected/disconnected them:
I4 --> 4 Noise/Freq pairs
I6 --> 1 Noise/Freq pairs

Here is the netlist.

simulator lang=spectre
global 0
parameters Fmax=10G CSH=1p

// Library name: Regulator
// Cell name: TEST_integ_noise
// View name: schematic_ori

I4 (net2 0) isource dc=0 type=dc noisevec=[ 10 (1p*1p) 10K (1p*1p) 10M \
        (1p*1p) 100M (1p*1p) ]

I6 (net4 0) isource dc=0 type=dc noisevec=[ 10 (1p*1p) ]

C0 (VOUT 0) capacitor c=CSH
R0 (VOUT 0) resistor r=1G isnoisy=no
IPRB0 (net4 VOUT) iprobe
simulatorOptions options psfversion="1.4.0" reltol=100e-6 vabstol=1e-6 \
    iabstol=1e-12 temp=27 tnom=27 homotopy=all limit=delta multithread=on \
    scalem=1.0 scale=1.0 compatible=spice2 gmin=1e-12 rforce=1 maxnotes=5 \
    maxwarns=5 digits=5 cols=80 pivrel=1e-3 sensfile="../psf/sens.output" \
    checklimitdest=psf
parameters Iteration=1
dcOp dc write="spectre.dc" maxiters=150 maxsteps=10000 annotate=status
dcOpInfo info what=oppoint where=rawfile
sweep sweep param=Iteration start=1 stop=100 step=1 {
tran tran stop=300n errpreset=conservative noisefmax=Fmax \
    noiseseed=Iteration write="spectre.ic" writefinal="spectre.fc" \
    annotate=status maxiters=5
}
finalTimeOP info what=oppoint where=rawfile
modelParameter info what=models where=rawfile
element info what=inst where=rawfile
outputParameter info what=output where=rawfile
designParamVals info what=parameters where=rawfile
primitives info what=primitives where=rawfile
subckts info what=subckts where=rawfile
save IBIAS1:sink
saveOptions options save=allpub

Dear delgsy,

delgsy said:

I am trying to find the rms noise at VOUT from transient noise simulation and got the same result regardless of the number of Noise/Pair.

I do not currently have access to spectre, so I cannot verify my hypothesis delgsy, but it appears the reason you are measuring a similar rms value when noise source I4 is used as when noise source I6 is that both have the same noise profile over frequency. Although you only specified a single frequency pair for I6 (10, 1e-24 V^2/Hz), that value will extend to your value of Fmax. The resulting noise profile is therefore identical to that you specified for I4 where you specified 4 frequencies each of which has the same noise density of 1e-24 V^2/Hz [(10 Hz, 1e-24), (10 kHz,1e-24), (10 MHz,1e-24), (100 MHz,1e-24)]. Thisnoise profile will also extend to your value of Fmax.

Another item you might consider, if you have not already, is your transient simulation time. With a 300 ns simulation time, you have effectively applied a high-pass filter function to the measured rms noise. In short, with a 300 ns simulation time, you cannot accurately assess frequency components whose period is greater than 300 ns (or less than 1/300ns = 3.33 MHz). Hence, the rms noise measurement will not show the impact of your noise frequency breakpoints below about 3 MHz.

Shawn

Hello Shawn,

Thank you.
I was aware already when I post my other reply yesterday regarding the noise level at other frequency.
Regarding the simulation time, I made new simulation summarized in the picture below.
In the transient analysis setup, I used Fmax=100GHz.
You can see that the VOUT at the transient result are exactly the same even though the noise spectrum are different.
From software point of view, do you think it is valid?
Is there anything that i can try?

Hello Shawn,

Thank you.
I was aware already when I post my other reply yesterday regarding the noise level at other frequency.
Regarding the simulation time, I made new simulation summarized in the picture below.
In the transient analysis setup, I used Fmax=100GHz.
You can see that the VOUT at the transient result are exactly the same even though the noise spectrum are different.
From software point of view, do you think it is valid?
Is there anything that i can try?