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.
Hi.
I have a question about Simulating by microwave office.(MWO)
MWO has two simulator.(EM and Circuit)
which of them is the best for microwave filter design.
bests.
Circuit Simulation is a faster way to predict the performance. EM will give you closer result as compared to the actual measurement. I would suggest you to start filter design with circuit simulation and run EM to verify its performance.
You can use the advanced EM models (X-models) in the linear simulator. Once you filled the database for your specific substrate (if you are using common substrates, they are allready in the EM_Models directory).
The X-models are based on EM simulations, are very accurate, and the simulation take less time in the linear simulator.
Take a look at the filter examples in the examples directory of MWOffice.
After tuning and optimization, you can simulate your whole structure in the EM simulator.
there is an old method for filter design, it is especially well suited for band pass filters having center frequency Fc and bandwidth BW. It starts with circuit simulator models and then after manufacturing the filter is measured. Typically, the measured center frequency would be shifted down for dFc and bandwidth narrower for dBW. Then a filter specifications would be changed, center frequency to Fc+dFc and bandwidth to BW+dBW and filter redesigned in circuit simulator to achieve changed specifications. Redesigned filter will have characteristics very close to target ones.
Instead of manufacturing and measuring a filter, you can switch to EM simulations, it should serve you as a good indicator how you should change your specifications and what to target in redesigning. (There is an optimization procedure called space mapping that is underlying this approach, if you ask me, space maping is the only successful method for EM optimizaton) Anyway, there is not an optimizaton option within MWO EM simulator, so you will have to have some criteria how to change dimensions of filter elements.
I did some filters in that way using @DS-Momentum pair, it was extremly successful. It took only 2 iteration of the described method to get the performance spot on. The agreement with measurements was also very good.
The only difference I can see regarding the use of MWO will be that the EM structure must be snaped to the fixed grid, it might introduce some discrepancies and limitations. The X-models will most certanly get you closer to the EM results.
There was an article by Peter Martin
"Designing Edge-coupled Microstrip Band-Pass Filters Using in Microwave Office"
which illustrates the method described by Flyhigh.
unfortunatelly the download section of his website is temporarly suspended.
To flyhigh: MAy I know what is the substrate used for your previous design and the centre frequency?
I used RT6002 to design BPF at X-band and simulated with momentum. The centre frequency measured shifted by 500 MHz from the desired centre frequency....The same problem occured when I use Taconic softboard. Until i changed my design to hard board (RO4003), the frequency measured agree perfectly with simulation. I just want to find out whether softboard is not suitable for filter design.
After etching process, moisture may remain in the board which will lower the center frequency of e.g. a side coupled bandpass filter. Insertion loss will increase as well
To reduce remaining moisture, it's important to dry-out (bake) soft boards in high temperature after etching. See
**broken link removed**
I used RT6002 to design BPF at X-band and simulated with momentum. The centre frequency measured shifted by 500 MHz from the desired centre frequency....The same problem occured when I use Taconic softboard. Until i changed my design to hard board (RO4003), the frequency measured agree perfectly with simulation. I just want to find out whether softboard is not suitable for filter design
it was hard, ceramic board, inhouse material. Unfortunately I don't have any experience on soft boards. The center frequency for BPF was 22 GHz, bandwidth 5 GHz.
To GaAs FET: actually I mill the PCB instead of chemical etching. After milling, I use abrasive paper to remove the debris. Under 10x magnifier, the milling is quite good.
So what do you think the problem will be since it is not moisture absorption? Appreciate your reply.....
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.
