Welcome to EDAboard.com

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.

Very bad results of HFSS for an X-band Waveguide Filter

Status
Not open for further replies.

Matthew_Dawson

Newbie level 2
Joined
Jun 7, 2009
Messages
2
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,301
x-band filters

We have analyzed, simulated and built an X-band waveguide filter. we performed a simple circuit analysis of the filter and an analysis using HFSS and later with CST microwave studio. The discrepancy between measured and circuit analysis results is small however simulation results show significant frequency. I have double checked everything yet this problem was not resolved. I'd be grateful to hear your suggestions. Here I attached simulation files and results.
 

ansonjimli

Junior Member level 3
Joined
Apr 29, 2006
Messages
25
Helped
2
Reputation
4
Reaction score
2
Trophy points
1,283
Activity points
1,382
mesh seeding,hfss

you should mesh more densely to get an accurate results. HFSS is good for this waveguide structure.
 

burton

Full Member level 3
Joined
Dec 12, 2003
Messages
177
Helped
15
Reputation
30
Reaction score
10
Trophy points
1,298
Activity points
1,077
cst studio mesh

HFSS is better for high Q filter, but mesh seeding may be used, especially for the coupling section with evanscent waveguide.
 

Matthew_Dawson

Newbie level 2
Joined
Jun 7, 2009
Messages
2
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,301
microwave filter with waveguide hfss

Thank you for your replies.
The Q-factor of the filter is 20. There is 1.8% frequency shift in the response of the filter. this problem was not resolved even when I used very dense mesh. (1.5GB of RAM was used for an x-band filter)
I ran the simulation with CST microwave studio, as well and used very dense mesh. It took 6 hours to do the simulation. However, results of CST Microwave studio is similar to HFSS and exhibits a frequency shift of 1.8% which results in considerable error due to narrow band behavior of the filter.
I'd appreciate to hear your comments.
Matthew
 

ansonjimli

Junior Member level 3
Joined
Apr 29, 2006
Messages
25
Helped
2
Reputation
4
Reaction score
2
Trophy points
1,283
Activity points
1,382
results in hfss

as burton said , the coupling part is very important and lmain for the load effect .you could use dummy project to improve the dense of this part.
 

chenqihit

Newbie level 1
Joined
Jan 20, 2010
Messages
1
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,285
Hi, Mattew. I am also doing a project designing a waveguide filter in HFSS and my result shows that the passband frequency shifts from the required frequency. It is similiar to your problem. So did you figure out a way to fix this problem?
 

PeterMartin

Member level 1
Joined
Mar 28, 2010
Messages
37
Helped
5
Reputation
10
Reaction score
2
Trophy points
1,288
Location
UK
Activity points
1,501
HFSS and CST/MWStudio are equally good tools for simulating WG filters.
MWS has a choice of several solvers. Their FD solver is good for filters , their transient solver is not so good , it is more suited to less resonant but broadband structures like antennas and couplers, but it is still usable and useful as final check .

It sounds like you are doing a check anyway between HFSS and MWS and are getting close agreement but in reality you are just over 1.5% off in your centre frequency.

For a filter that's not too bad. They can be very sensitive to machining tolerances, or imperfections in soldered joints etc, -which is why tuning screws are often needed. Thermal expansion can be another problem for the same reason. You might want to calculate what machining inaccuracy will produce the error you are measuring.

If you are slightly too high in frequency adding a tuning screw in the cavity should fix the problem.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Top