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.
CST Microwave Studio uses
the FIT method (finite integration technique) to discretize the computed
volume.
This mathematical technique, among other advantages, guarantees that
NON-physical solutions cannot exist. Other numerical techniques that solve
Maxwell equations, may produce results which do not come to close proximity with
measurements, i.e. filter modes predicted by software may not exist in reality
etc. This problem may appear in particular, for example, on numerical techniques that
solve high order differential equations and their solution roots do not necesserilly
correspond to the reality or cause instability problems etc..
So, from my experience someone who requires accuracy and reduction in the uncertainty factor, may choose MWS. However, the selection of
THE optimum simulation program is always an area of endless discussion..
CST Microwave Studio uses
the FIT method (finite integration technique) to discretize the computed
volume.
This mathematical technique, among other advantages, guarantees that
NON-physical solutions cannot exist. Other numerical techniques that solve
Maxwell equations, may produce results which do not come to close proximity with
measurements, i.e. filter modes predicted by software may not exist in reality
etc. This problem may appear in particular, for example, on numerical techniques that
solve high order differential equations and their solution roots do not necesserilly
correspond to the reality or cause instability problems etc..
So, from my experience someone who requires accuracy and reduction in the uncertainty factor, may choose MWS. However, the selection of
THE optimum simulation program is always an area of endless discussion..
The implication is that it is impossible to get non-physical soluitions to exist w/ FIT...HHHhhhmmmm... What do I do with the non-passive / non-causal sparameter results that show gain when looking at passive waveguide structures from CST simulations?
Or... how about this one... I have a small gap between structures (0.05 mm) CST has shown the results to be continous (i.e. connected)...
Both of these situations are easily solvable... as generally, they can also be described as "user errors"... Of course that also carries some self implecation that does not show up very well with one's yearly performance review.
The point is not that CST (or any other tool) is bad... instead, focus on how to use the tool... and the present limitation of a specific version for a specific application when setting up your problems. Thus to get away from the "endless" debate about which is better... and focus on "how to get the job done as best as possible with a specific tool".
The above mentioned issues, correctly stated, are user errors.
However, you should be careful, because for example, the
positive S11 in the response is NOT because of the FIT method
but because of the Fast Fourier Transformation between time and
frequency domain. Such, effect is a fundamental phenomenon
in signal processing, due to the trancated time response. MWS
can easily rectify this issue through the available options. It is
quite interesting to note also, that a resonant frequency is not
affected from this.
As far as the 0.5mm gap is concerned... this is a much easier
task to be solved, by applying manual meshing locally, if
required.
I quote the above information, so that people who are unware with
the principles of MWS have a better picture of solver's abilities.
However, as an ordinary used, I simply exchange ideas within
this forum for the benefit of everyone.
For planar structures like Baluns, spiral inductors, transformers, @DS M@omentum seems pretty alright.... I compared it to #fss and it is simpler and quicker....
I must agree with all of them who has suggested @DS. I do not agree that it is difficult to learn, on the other hand it's interface is splendid. Keep in mind that this comparison is done regarding to the RF simulations tools. No RF simluation tools can, in my opinion be considered easy. I know it is expensive but still it is the best. Cadence is good for layout of some specific components but i think it somtime is difficult to use.
The fastest harmonic balance and most stable one today in the market is EldoRF from Mentor Gr@phics. SpectreRF has a market lead in sales though but its technology is a bit old now and needs update. I know for a fact that C@dence is now building HB into its simulation environment
SOME SUGGEST:
There is not the best RF simulator.So you need to give some patient to study the detail of the RF design skill.
In my work,I favorite to use the MWO and design.,and zealand and hfss.
Ofcourse there is not a best RF simulator you could talk of. It all depends on the end product of your application. For filter designs at RF, I can Ansoft is the better. But it boils to the end product of your application.
It all depends on what you are simulating. Is it antennas, filters, amplifiers, complete modulated systems? I generally find that the simplest for your use is best. One of my friends who runs a business has all of his needs met with a ten year old version of Eagleware Genesys.
Added after 41 seconds:
besides,
DS is good for planar simulations but nothing beats HFSS in 3D simulations.
Added after 2 minutes:
Ansoft HFSS is very good and powerful for simulating antennas.
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.
