Continue to Site

# SNR of transient signal in Cadence

Status
Not open for further replies.

#### Ali263

##### Member level 1
Dear all,[

I have a transient signal as shown in attachment. I run transient noise simulation too to check the effect of the noise

I dont get how to find the SNR in cadence.
I have searched alot and i dont understand how should i do it.
Is it ok to find the rms value of clean signal and rms value of (noisy signal - clean signal) and find the SNR from that or use some other method>?

#### Attachments

• CSPTRans.PNG
11.6 KB · Views: 79

Hi,

I don´t understand what this transient signal has to do with noise and SNR.

Calaculation of noise values:
If you have independent RMS values then they add with the squares:
N-total = sqrt(N_a^2 + N_b^2)

Klaus

What are you detecting in your signal? The peak or something else? That would be your signal information. The distance from that information to the noise floor would probably be a reasonable definition of SNR.

If the transient noise does its job and your circuit is reasonably linear over the range, the AC noise simulation will yield a result close to the transient noise. Transient w/o noise and then noise simulation would probably be faster in the end - but requires linearity.

SNR needs a good value for "N" and this is GIGO in simulation-land.

If noise is a known model then you can apply it. My advice, based on serial line interface work decades ago, is to characterize the input sensitivity and min signal for that ratio. If you don't know or can't control the noise then make it somebody else's responsibility to meet "conditions".

I guess you are working on some radiation detector front-end. What people are usually do, is to simulate impulse response in transient, measure a peak amplitude for given charge (to get charge gain), and performing ac noise analysis to get noise rms in frequency spectrum. Then ENC (equivalent noise charge) is calculated as noise rms voltage at the output divided by charge gain.

Status
Not open for further replies.