Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

fft analysie in Hspice Vs. PSS in SpecterRF

Status
Not open for further replies.
F

flamingo

Member level 4
Member level 4
Joined
Jan 10, 2005
Messages
72
Helped
6
Reputation
12
Reaction score
5
Trophy points
1,288
Activity points
717
To get the performance of IM3 while keeping the output of amplifier 1Vpp, I used two methods: one is PSS single point analysis, using two tone signal with same amplitude which keeps output 1Vpp, the other is using FFT analysis in Hspice with the same stimuli. But the results from two simulators are so different. IM3 from PSS is much much better than the counterpart in Hspice, about 20dB higher. So which result shoud I believe??

And the fft results are inconstant with different simulation configurations, i.e., for different duration or step in .tran analysis, and different NP in .fft, get different results, variation is about 3~6dB, but they are yet incomparable to the result in PSS. what is reason??
 

Each of them has its own tricky points.
When using fft look for windowing and spectrum leakage phenomena. To get good accuracy in your case one has to use flat-top window. I'm not sure if hspice provides this as a standard feature or you should resort to 3rd party postprocessing.

Talking about pss, occasioanly it converges to some strange states. Not sure how and when this happens, seems to happen at random. This might explain your discreptancy.
 

thank your for your tips. I have noticed the spectrum leakage phenomenon, but so little help occured.

another phenomenon in FFT analysis (Hspice): for a high gain amplifier (about 60dB), the input signal amplitude is very small. when doing fft analysis, the results indicate that the harmonic distortion will be improved with the increasing input signal strength in some range. But this violates the reality. because for a linear amplifier, Y(t)=a0+a1*x(t)+a2*x(t)²+a3*x(t)³, then HD3 (defined as the ratio of 3ord HD to fundamental) can be expressed as ¼*|a3/a1|*x(t)² , so with increased x(t), HD3 is lower. How to explain this?
 
flamingo said:
thank your for your tips. I have noticed the spectrum leakage phenomenon, but so little help occured.
Click to expand...

Do you use window when doing FFT? Which one?

flamingo said:
another phenomenon in FFT analysis (Hspice): for a high gain amplifier (about 60dB), the input signal amplitude is very small. when doing fft analysis, the results indicate that the harmonic distortion will be improved with the increasing input signal strength in some range. But this violates the reality. because for a linear amplifier, Y(t)=a0+a1*x(t)+a2*x(t)²+a3*x(t)³, then HD3 (defined as the ratio of 3ord HD to fundamental) can be expressed as ¼*|a3/a1|*x(t)² , so with increased x(t), HD3 is lower. How to explain this?
Click to expand...

This commonly occures in AB or B-class amplifiers. They have a "dead" range or non-linear range at very small signals. Another possibility is improper or imprecise work of common-mode circuit in a fully differential amplifier - this can cause small signal dead range as well.
 

    F

    flamingo

    Points: 2
    Helpful Answer Positive Rating
I have tried the hann, and haris. there are some differences in wave form but not the performance of HD3 or IM3.

Acturally, there is a CMFB in my ckt, maybe that is the problem. But how to scale the preciseness of a CMFB, how to simulate, or is there any spec for this performance?

Similar phenomena are not dectected in PSS simulator (spectreRF),i.e, smaller output amplitude, the better IM3. Then how we explain the CMFB instability or impreciseness in SpectreRF?
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top