cst low frequency
I have seen a lot of opinions about CST or HFSS, but I need a new one about microstrip design for frequencies up to 10GHz and for wideband simulations
We plan to buy an EM software and I want to make the good choice for our design.
Our structures are often for antennas, filters, printed inductors (size of elements very little in compare to frequency)..., at 868MHz, 2.4GHz,...
I would like to use a lumped elements inside EM simuladtions (lumped element in CST).
In my opinion I find CST easier than HFSS and it uses less memory for simulation. In V5 the meshing become good, and the updating meshing is similar than HFSS.
Thank
Greg
For wideband microstrip structures you can use CST.
For high-selective filters HFSS is better.
Hi all,
with the new GUI of version 9
HFSS is now very easy to use
and with its frequency sweep algorithm can be easily used to
analyse broadband structure
In these days I analised an REF Mems
in a range from DC to 800 GHZ
in one shoot with no problem
If I were in you I would try on my own on a serious structure
( not just a microstrip)
and see what HFSS can do for you
Rugbyfun
For the wideband simulations , I suggest you can use the xfdtd and cst. These software quickly simulate and give you the wideband simulations,they are a full wave 3D electromagnetic solver based on the Finite Difference Time Domain method.
Hi,
just have a look at both programs yourself and try the structures you would like to model....
... and then choose CST
F.
Hi
After a lot of simulations with Ansoft team and CST team we have chosen CST for the very good accuracy with our benchmark.
The choose of the software depend on the problems.
Each software is good but for us CST is better.
For the moment I don't find a lot time improvement with MSS but perhap's with the futur version!
It was very interesting to see the possibilities of these soft.
Regards
Greg
Just curious, why aren't you considering a planar tool for the planar portion of your problems?
If you need accuracy for I2R loss and want to be sure your volume meshing tools are converged, check the current distribution. It must be smooth with a clear high edge current. I think you will find that CST generally converges much faster than HFSS for this kind of problem. (Full disclosure, I work for Sonnet and we represent CST.)
You should contact Ansoft reps in your area - as part of the "try before you buy". Raise the following questions - since they may be relevant to your topics:
1. HFSS has optimization for elements smaller than wavelength called "low order solution".
2. If skindepth is important ask about the accurate "solve inside" feature
3. Regarding filters, I would ask about the dynamic link between HFSS and Ansoft Designer.
4. Regarding planar structures I would ask a demonstration of the cosimulation between Planar-EM inside Ansoft Designer and parameterizable geometry export to HFSS.
Hope this helps
Itai
Hi
In fact this not the printed structure (antenna,...) who have decided our choice of the software. We need to simulate multilasyers PCB with 3D connector and mechanical constraints.
The more accuracy results were obtained with CST. We have spent a lot of time with Ansoft reps and CST reps.
For the simple printed strucuture, like the antennas, the results were the same. We need to have a 3D tool because we have a lot of mechanical constraints around the design.
The advantage of ansoft is the possibility to use the HFSS and Designer in the same time.
The problem with HFSS is the meshing. If you want to sweep a parameter you need to rebuild a each time the meshing.
CST can keep the same meshing and then the HFSS simulation time is higher than CST.
For only one analysis HFSS is often quicker than CST.
Hi greg
The other important think for us is the easy to use of CST. This is an important consideration when it's not always the same peaople who use it
Hello,
I have a followup question about that:
Does the parameter change - changes the geometry ? If so, wouldn't it be smarter to rebuild the mesh in order to increase accuracy ?
Itai
Hello
It's not necessary to rebuild the meshing for th accuracy with CST
If you want to have a good accuracy, you can use the template for started a design and use the adaptative meshing for the first process. When your meshing is ok you can use it for all your runs.
When you use the adaptative meshing under CST, the Delta S is calculate on all your frequency range. It is not the case with HFSS (Single frequency point). Then is not necessary to choose the same value with CST. With HFSS it' important to look where you choose the frequency point.
Regards
greg
Hi,
well, I am using CST a lot and I am satisfied with it. But I never got good results if the structure becomes very small to the wave length. We did some tests with a coil (inductor) and it never worked. Check it out before you buy the tool for these things.
But for all other things, especially wide band simulations at antennas, it is a really good programm.
Ciao, Bodo