cst mws coaxial cable design

when i simulate with CST MWS 5.0 i was warned that "maximum number of pulse widths simulated,Solver stopped.please note that the steady state energy criterion has not been reached." I found that i have chosen "energy based" option for the "refinement strategy" in the refinement setting of adaptive refinement and i guess it is the reason leading to the warning, can i avoid this error if i choose "expert system based" option? Thank you !
I am a beginner, i have some other question to ask:
1 which is more accurate when using waveguide port and discrete port ? here my problem is to simulate a printed dipole.
2 Generally, i use the default exciting signal when simulating, is it proper for most problems? if not, when should we change the exciting signal ?
That is all. Any help will be appreciated ! Best regards!

Hi,
the choice of excitation depends on the structure you would like to simulate. Whenever possible you should use the WG Port. This is especially true if the dimension of the excitation source (Coax, M-Strip e.g.) are comparable to the dimension of the full structure (like Coax dimensions are lager the 1/100 of the total structure).
If your source is much smaller (like 1/1000 of the total structure size) it might be better to replace the WG excitation by a discrete port.
In some other simulation you do not have a modal excitation in the first place (packaging, dipole feeding gap e.g.). In this cases the Discrete source would be the more natural excitation as well. I guess is more a general comments and should hold for HFSS as well
F.
PS:
some more comments :
1. If you are interested in Frequncy Domain Data you should use the default excitation.

2. In the time domain solver of CST it usually makes more sense to use some manual meshing to refine the mesh inside the substrate and the M-Strip before you run the simualiton. This is very simple and save you a lot of simulation runs and very offent there is not additionl mesh adaption required. This is the reason why the adaptive mesh refinement is not switch one by default in the TD Solver.
In the FD Solver (which since Version 5.1 uses a TET mesh as well) it is always recommeded to use the adaptive mesh refinement - just like in HFSS.).

Hi, RFsimulator, thank you for your help!
I am simulating a bowtie antenna with a full size of 228*182*50 mm, it is fed with coax cable which cross over the gap between the two triangles , the cable's inner and outer radius are 0.44 and 1 mm respectively. It seems that the Coax dimensions are comparable to the dimension of the full structure, and in fact, a waveguide port should be used in this case .If the feeding coax cable is too small ,you suggested that a discrete port should be used,i want to ask should we set a discrete port pointing from the outer conductor to the inner conductor,like in HFSS.
as an another choice, can i take out the feeding coax cable and just set a discrete port across the gap between the two triangles (connecting the apex of the two triangles ) ?
you mentioned that In the time domain solver of CST it usually makes more sense to use some manual meshing to refine the mesh inside the substrate and the M-Strip , then i want to ask how to refine the mesh manually?
Best regards!

Hi,
this is the idea. Take out the complete coax and just use a discrete port. If you leave the coax in the model and use it together with the discrete port this does not help at all to reduce the calculation time.
You can refine the mesh inside any object by specifiing the minimum mesh density. Change to the mesh view and select the object right mouse click Mesh properties....

F.