CST MWS - emission vs frequency 3D model
I have a question regarding the CST MWS program, and don't seem to find
the answer in the tutorials/ forums.
I want to test the imunity of a box (shield) - which will contain a PCB -
in some frequency area (1-3 GHz), so that to know where to place it.
My question is how can I make this frequency sweep (Emission vs. Frequecy 3D model)?
Thank you for your time.
To make it a little bit clearer:
I want to put a EM source at some distance (1 meter) from the box and make a frequency sweep to see the 3D model of the emission distribution in the box, in order to know where to place some sensitive components.
Use field probes in CST MWS. It will give you field vs Freq plot at specific points. CST MWS 2011 now comes with CST Microstripes which maybe useful to EMC problems- In the example folder (CST MS) there is an example which is rather similar to what you want.
It does not really help with my problem, since i need a specific surface, not points.
I have found a macro which automatically puts field monitors for each frequency I need. There will be about 2000 monitors, since these are the requirements of my project (between 1-3 GHz).
Afterwards I would like to evaluate the field on a specific surface which I have defined, but it really is pointless to do it manually for each of the 2000 results. How can i make this automatically?
Can I in any way export all the results at once? Not each results which is in a different folder manually?
The final goal is to evaluate the maximum E-field inside my box on a specified surface (plane) at each frequency and plot this graph. From thee graph I can see at what frequencies thee E-field is too large, and can analyze just these frequencies. Is there any way I can do this easier (automatically)? Do I have to write some macros?
Any examples related to this project are welcome.
Since your source is located in the farfield I would use a plane wave for your excitation source. You can use the EMI/EMC template that will define a plane wave for you.
You can use as many field monitor probes as you like because they are computationally free but as you mentioned... that's a lot of data to go though at the end. You can plot all the probes on one graph and then pick the one that reports the highest field and ignore the rest perhaps. If many curves are selected on one graph, you can copy and paste (or export) all those curves with one command as long as the graph is displayed.
If you are interested in a few frequencies within your band, you can set up E and H field monitors at certain frequencies that would give you the field distribution inside your box and maybe then you can put probes in locations that you might be interested in.
Just to be sure:
-do you mean the field monitors which evaluate the field in 3D (or 2D) or the field probes which measure the field in just one point (because i am interested in the field monitors)?
-how can i plot all the field monitors on the same graph? what command should I use?
Hmm.. I'm beginning to understand you question. Typically I use several field probes to plot field inside the enclosure -- this way it's easier to monitor trend over frequency. I may use field monitors to look at several frequencies across the band of interest but I typically space those points apart because field monitors aren't computationally free and are difficult to organize and present.
The way that I can think of to have all the 2D/3D frequency monitors on one plot would be to use a time domain monitor instead of frequency (problem here though is you can't determine frequency). Personally I like using the probes -- I believe you can get the data you are looking for much easier this way.
I have exactly the same problem as Corin.
The reason we are trying to use field monitors instead of probes is because we want to find the maximum emission point in the entire PCB plane (from inside the box). We could miss some maximums by testing only a few manually placed points.
Our purpose is to determine the frequencies which have the highest emission on a plane inside the box. A time domain monitor which would not allow us to determine frequency would thus be of no help.
I have just tried the field probes and it looks very close to what we need. Thanks!