Stats

  • Locked Locked
  • Replies 3
  • Subscribers 64
  • Views 3184
  • Members are here 0
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Noise Figure in N-path band pass filter using PSS/Pnoise

Binhngo1210

Hi everybody,

I am using Pss/Pnoise to simulate NF of a 4-path band pass filter. In term of theory, this NF is medium around 5 to 10 dB depending on the switching frequency, which is 

contributed by thermal noise of source resistance, switch resistance and noise folding from other harmonics of switching frequency., flicker noise is supposed to be neglected because there is no DC current from port 1 to port 2

However, the simulation is a little weird to me (it is around 27 dB at 2.41 GHz - center frequency of filter). 

Is there any wrong here, could you help to figure out some possible reasons? (any culprit that is able to fold the flicker noise into the desired band, because there is a peak in simulation result) 

my circuit infor: it has a divider/4 from 9.64 GHz to 2.41 GHz controls cmos switches 

input: 2.41gHz 

output 2.41 GHz 

insertion loss: around 5 dB 

source resistance: 50 ohm at two port  

here is my setting for pss/pnoise: 

           

thanks for any advice.

Binh Ngo 

  • Dear Binhngo1210,

    I am very late on a response, but just noticed your Forum post, and would like to comment if I may.

    Binhngo1210 said:

    In term of theory, this NF is medium around 5 to 10 dB depending on the switching frequency, which is 

    contributed by thermal noise of source resistance, switch resistance and noise folding from other harmonics of switching frequency.,

    and

    Binhngo1210 said:

    my circuit infor: it has a divider/4 from 9.64 GHz to 2.41 GHz controls cmos switches 

    input: 2.41gHz 

    output 2.41 GHz 

    insertion loss: around 5 dB 

    You did mention that the NF is "medium and around 5 to 10 dB depending on the switching frequency". However, I am thinking that with a switching frequency of only 4 times the desired center frequency, your Noise Figure will be significantly degraded from that of a filter with a much greater switching/center frequency ratio.  Hence, I was not that surprised to see your simulated value of NF as about 27 dB.

    1. Did you examine the transient response of your circuit nodes? I see your settings suggest you saved the transient simulation data prior to the PSS phase of the simulation.

    2. Have you examined the transient response of the switch control signals?

    3. Have you explored the impact of changing your switch device sizes on NF?

    Shawn

  • in reply to ShawnLogan

    I'm answering this very quick without really diving into things (too busy at my day job :-) ). 

    • Do you wanting to look at time-average noise?  (noise averaged over one cycle of the pss_fund) 
    • Or do you want to look at sampled noise (noise at a transition) and measure Edge Phase noise?   
    • What are the waveform shapes?  Square wave-ish or sinusoidal?   

     If you want to look at sampled noise (most common for switching circuits), then I'd use sampled(jitter), rather than timeaverage noise.  Your pnoise setup would be incorrect for sampled(jitter) noise.   The pnoise stop time would be 1/2*pssfund.   

    This sort of question is a great one for Customer support https://support.cadence.com .  

    best regards,

    Tawna

  • One more thing... the Pnoise Summary is great for finding all sorts of pesky noise sources which cause problems.

Community Guidelines

The Cadence Design Communities support Cadence users and technologists interacting to exchange ideas, news, technical information, and best practices to solve problems and get the most from Cadence technology. The community is open to everyone, and to provide the most value, we require participants to follow our Community Guidelines that facilitate a quality exchange of ideas and information. By accessing, contributing, using or downloading any materials from the site, you agree to be bound by the full Community Guidelines.