how to improve the accuracy of SNR in HSPICE??

[nickooeda]

hspice snr

dear all:

I found that the SNR simulated in HSPICE is very bad,how to improve it??
thanks first

 

[maxwellequ]

snr calculation from given accuracy

Can you be a little more specific ?

 

[nickooeda]

how to improve the snr

fft test ... sine
v1 1 0 sin(0 1 1k 0 0 90)
r1 1 0 1
.tran 10u 2.0m
.option post=2 fftout
.fft v(1) np=1024 fmax=10k freq=1k window=hann
.end

*******simulation result***********************


1.0000k hz 0. db at 39.2943u Deg ****Fundamental****
2.0000k hz -69.0041 db at 407.0327m Deg Harmonic # 2
3.0000k hz -81.8473 db at 838.3165m Deg Harmonic # 3
4.0000k hz -88.4335 db at 1.2383 Deg Harmonic # 4
5.0000k hz -92.9984 db at 1.6213 Deg Harmonic # 5
6.0000k hz -96.5294 db at 1.9946 Deg Harmonic # 6
7.0000k hz -99.4235 db at 2.3620 Deg Harmonic # 7
8.0000k hz -101.8827 db at 2.7256 Deg Harmonic # 8
9.0000k hz -104.0242 db at 3.0866 Deg Harmonic # 9
10.0000k hz -105.9228 db at 3.4458 Deg Harmonic # 10
11.0000k hz -107.6293 db at 3.8036 Deg Harmonic # 11
12.0000k hz -109.1798 db at 4.1604 Deg Harmonic # 12
13.0000k hz -110.6009 db at 4.5164 Deg Harmonic # 13
14.0000k hz -111.9130 db at 4.8718 Deg Harmonic # 14
15.0000k hz -113.1318 db at 5.2267 Deg Harmonic # 15
16.0000k hz -114.2699 db at 5.5811 Deg Harmonic # 16
17.0000k hz -115.3376 db at 5.9352 Deg Harmonic # 17
18.0000k hz -116.3430 db at 6.2891 Deg Harmonic # 18
19.0000k hz -117.2933 db at 6.6427 Deg Harmonic # 19
20.0000k hz -118.1941 db at 6.9961 Deg Harmonic # 20
21.0000k hz -119.0506 db at 7.3493 Deg Harmonic # 21
22.0000k hz -119.8668 db at 7.7024 Deg Harmonic # 22
23.0000k hz -120.6465 db at 8.0553 Deg Harmonic # 23
24.0000k hz -121.3928 db at 8.4081 Deg Harmonic # 24
25.0000k hz -122.1085 db at 8.7609 Deg Harmonic # 25
26.0000k hz -122.7961 db at 9.1135 Deg Harmonic # 26
27.0000k hz -123.4577 db at 9.4661 Deg Harmonic # 27
28.0000k hz -124.0953 db at 9.8185 Deg Harmonic # 28
29.0000k hz -124.7105 db at 10.1710 Deg Harmonic # 29
30.0000k hz -125.3049 db at 10.5233 Deg Harmonic # 30
THD = -28.7068 ( 36.6996m% ) SNR = 5.9991 db
SFDR = 6.0092 db at frequency 1.5000k hz

NOTE: SNR=5.991!

 

[sunking]

spice noise snr calculation

modified:
.tran 10u 2.01m

Try.

you should use integer times

 

[xuel]

fft improve snr

It seems u misunderstand fft analysis. The frequency bin in your sim file is 1/2m=0.5K, which is too large comparing with your signal freq 1K. Therefor u should increase signal frequency or increase sim time.

 

[nickooeda]

snr hann

I don't clearly understand about FFT.
when I compared RECT window with HANN,the result is very different with the same other parameter.
SNR(rect)>200db
SNR(HANN)<7db
(((

 

[mists]

how increase snr & signal

use three points as signal for hann window to calculate snr.

 

[piao]

if you want to improve the SNr, you should delete the DC component in that signal, it is hard. the SNR is related to the Dc bias of that signal.

maybe you can come out some ideal to delete the DC component in the signal.

Good Luck!

 

[nickooeda]

I think the best way to calcultate SNR of signals is to use MATLAB to calculate.))