question about CST MWS simulation, help needed

[tenderne]

cst mws view bug

Dear all:

I encounted some problems when simulate a simple antenna structure; the model, zoomed port&lumped element and the s11 resultsas, are all shown in the attached file. The simulation conditions are:

1. discrete port, Gaussian excitation
2. one lumped element, serial RLC ( R = 2.4, L = 2.4e-8, C = 0)
3. frequency range: 0.5~3.5GHz
4. mesh properties:
line per wavelength = 30, lower mesh limit = 30, smallest mesh step = 0.1
5. open boundary, add space = wavelength / 8
6. background matrial: normal
7. simulation accuracy = -50dB, PML = 5 layer

seems everything is ok, but the s11 result is quite strange. I tried a few times, still cannot get some reasonable results.

So, anybody, with experiences of CST MWS simulation, please take a look of my model and s11; kindly give me some suggestion or hint. thanks a lot in advance.

regards,
tenderne

 

[shameemkabir]

cst mws lumped element port

From S11 it is clear that all the powers are reflected. It seems that your port is not correctly setup. First try to setup the discrete port without the lumped components. If it works, then put your lumped components.

 

[tenderne]

cst mws frequency range s11

shameemkabir, thanks a lot for your reply.

actually I tried the case of discrete port only, but it doesn't work. I also tried to replaced the lumped element with another discrete port, still not work. so seems not the problem of lumped element

when I insert a port, the port type is: s-parameter, impedance = 50.

please give me some suggestion. thanks.

regards,
tenderne

 

[shameemkabir]

You have an extra metal part between your antenna and ground. Remove that first (which is between the lumped component and discrete port) to simulate without lumped component. Take a look at the attachment what I refer to.

 

[tenderne]

shameemkabir, thanks a lot for your suggestion.

I cannot change the structure of the whole model. The structure is just given like that. The problem is in simulation, not in structure design, that's no doubt. thanks.

regards,
tenderne

 

[snkhan]

This error happen most of the cases when using the cracked version of CST 2006/2006 beta. Its better if u can upload ur design files

 

[fmoney]

Hi guys,
i think that the problem is in the values of RLC elments. In this situation the values of R may be a fraction of ohms, while the value of L may be of 1-2 nH (1e-12).

However, i think that the use of RLC in these simulation it's not necessary and the results without RLC element are good and close to the reality.

Best regards

 

[mogwai]

I know that this might sound really strange, but a while back using mws2006 for the first time on my core duo cpu, I remember seeing similar S-param results...

This is obviously a bug on the specific version but I managed to see "normal" S-params by altering the option for multiple-processor simulation in Transient Solver options (Use parallelization for up to : x processors).. When I disabled this option my S-params where calculated correctly !

Hope this helps !

cheers!
mog.