i have trouble in vivaldi simulate. i make a vivaldi array model in hfss.and now the result of fast and discrete is very different. the band is 1-3GHz. and my fast sweep use broadband sweep,i dont know how to do
the result which s11 below -10dB is fast ,the another is discrete
The discrete sweep is more accurate than the fast sweep, so I would trust the results from the discrete sweep more.
Mohamed K. Emara now i know the reason. because i use the 150ohm port rather than 50ohm. and i dont change the renormailed impedence. so now i want to know how the renormailzed impendence influence s parameters? in my design, use the 50 or 150ohms setting,the fast sweep result not change,but the discrete sweep resutl change as i change the renormailzed impedence.
Make use of discreet sweep instead of fast.
The discreet sweep results most accurate computations.
Ss Aaaa the renormalization impedance is a port post processing step that should always influence the the calculation of the S-parameters as they would be normalised with respect to that impedance. The renormalization impedance would also influence the fields if you check the "Include Post Processing Effects" in the "Edit Sources" (Right-click on Excitations > Edit Sources in the project panel). The fact that discrete sweep results change as you change the renormalization impedance is expected. The fast sweep results not changing is strange and I don't have an explanation for that. I would personally suggest that you use discrete sweep at specific frequencies for any simulations in which need the fields results (e.g., gain) and the interpolating sweep for other results (e.g., S-parameters).
Dear Ss Aaaa,
Let us first understand the basic difference between discrete sweep and fast sweep in computation process. In discreet sweep total sweep bandwidth is divided in to predefined interval point and at each point averaging is applied. Thus we get better result on each computation point point due to averaging . In fast sweep mode, sweep bandwidth is linearly varied from lower frequency to upper frequency in a single sweep. obviously discrete sweep will give batter/accurate result but it takes more computation time due to averaging. Pl. note a point hear that result in both type of sweep depends upon formulation of computation sequence. Assignment of port impedance value is comes under "formulation of computation sequence" . As all of us know in calculation of S- parameter if terminating impedance changed it influence full result. Hence when you change the terminating impedance value 50 Ohm or 150 Ohm result are bound to be deferent. In order to get particular/intended frequency point in to discrete sweep, we must divide total sweep band width in such way that particular/intended frequency fall at one of the discrete sweep point. With this explanation now we should take decision if we are not so particular about accuracy we can go for fast sweep mode or if our aim is to improve the accuracy we should got for discrete sweep mode but penalty we pay for is higher computational time. Fast sweep mode result are also good and acceptable. Hope This may clarify your point and may be use full to you.
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators