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.
I feel very impressed by all the discussions. I agree that the best is to try out some benchmark samples to see which one runs better. Would like to thank all members involving in this discussion.
In my view, cst is based on FIT,but hfss is based on FEM.For broadband structure,CST is the best choice.But the macroof CST is not easier to use
than the one of Hfss.for example,the phase of high impendence surface is hard to solve in CST.Someone can help me?
Long discussion, many opinions ...
But what I want to say that the answer (which program is better) strongly depends on solving problem.
Both HFSS and MWS have a very wide spectrum of applications.
I develope mostly a waveguide devices (~1 - 20 Ghz) From my point of view, HFSS is evidently better now as a solver.
For example :
1. It's still not possible to define arbitrary port at MWS (only at main plains)
2. In case of highly multimoded port - MWS is much slower.
3. In case of high lossy material & resonanse struсture - MWS gives wrong result in attenuation even with precise mesh and long time of calc.
(cheked in experiment, HFSS is perfect as usual)
The same time, MWS has a much more convinient GUI. So, I use MWS only for
rough estimation or for sweeping parameters or for preparing geometry for HFSS.
And if you are worry only about s-params the tandem of HFSS+CST DS is very usefull !
Its possible to analize a very long and complicated structure just by splitting it on elementary parts.
3. In case of high lossy material & resonanse struсture - MWS gives wrong result in attenuation even with precise mesh and long time of calc.
(cheked in experiment, HFSS is perfect as usual
"3. In case of high lossy material & resonanse struсture - MWS gives wrong result in attenuation even with precise mesh and long time of calc.
(cheked in experiment, HFSS is perfect as usual
I was just calculating the behavior of a coupled cavities in a lossy circular waveguide when
I found that MWS v5.02 gives the wrong result in attenuation. I checked it more carefully and it was the real bug.
Now, in MWS v5.1 the results are pretty same as in HFSS and there is a notice about it in the "whats new" :
" - Fixed wrong calculation for JD(lossy)."
I got huge problems in running a simulation in HFSS.
I investigate conformal antennas on cylindrical tubes.(array of them)
HFSS manipulates the resonant freq. and needs more than 5GBYTE RAM to run the simulation for a parametric in distance between two antennas on this cylinder!!!
So could you help me out of this misery, I´m not a friend in times of mpg and mp3´s to have software with more than 1GByteRam.
hi, i´m using hfss for designing an inset feed antenna on a cylindrical tube and wanted to investigate the coupling effects mutually for severall distances, but i´m sure you´re experienced- hfss took 4GByte of RAM and 1GByte of Harddisk - the problem is I´ve already talked to people from Ansoft the mesh and its refinement. I hate this.
I´ve heard that CLEMENTINE which is not available, was used to specify this structure.
But with CST I have no experience....
Please help me out of my misery!
thanx huckenmeyer@hotmail.com
I am generally convinced that for the general user, there are "Best Practices" that should be observered no matter which software that is used. If either CST or HFSS is known well enough, either can probably do most or all of the job... Of course there are specific limits in the "tails" of each software... so it is absolutely possible to stump either software...
This being said there are specific things in each software that are easier/better/faster to use than the other... that always seems to generate a religious style style debate.
I have examples where CST has solved larger problems due potential of distributed computing 60 ports vs 36 ports.... but HFSS has solved problems (that stay in core) faster... 48 hours vs 6 hours. The trick here is to develop the best practices in you job such that you can use the tool(s) that you own the most effectively.
Taking this one step further, I also have been around long enough (15+ years) to know that each version of any software will "jump" over the competition in some way. As such, it is usually important to stay current with your investment. It is also helpful to state which version you are using when making comments... Especially the "qualitative" comments like "easier to use".
OK... I haven't added too much technical content... stay uned for future posts.
Just for your information here is some comparison between MWS and measurement.
The cassegrain antenna I was talking here about earlier. 1/4 part of structure simulated (PEC+PMC walls used), ~5mil. of cells, 10 cells/λ (memory...)
Now I'm presenting E, H farfield comparison @ 26.5GHz.
Resolution of farfield cuts in simulation was 1° - smallest that MWS offers
Gain difference is ~ 1dB
Make conclusion by yourself, I think agreement is very well even if you consider meshing and the fact that model has been simplified (no subreflector clips in simulation).
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
