Continue to Site

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.

Verification of a 3D simulation of a bandpass filter

Status
Not open for further replies.

niki

Member level 4
Joined
Feb 20, 2002
Messages
70
Helped
20
Reputation
40
Reaction score
20
Trophy points
1,288
Location
Switzerland
Activity points
418
A classic bandpass filter (Microstrip Interdigital) was designed and optimized on the substrate RO4350B.
The filter was checked with 3D simulations (CST).
I would be very interested in verifying the simulations with other tools (HFSS, ADS, Sonnet,...).
If someone has the time and the desire, I will provide the layout and the necessary information
for the simulations (Gerberfile). After manufacturing and measuring the filter I will present
the results in comparison. Who wants to try?
bandpass.png
Substrate: h=0.254 mm,
Frequency range: 1-12 GHz
 

Having worked in EM simulation for 20 years, I have seen many measured-vs-simulated comparisons.

You get a bunch of curves, but these are usually difficult to interpret because of manufacturing effects (etching tolerance etc), measurement tolerances (port calibration), uncertainties in the material modelling (rolled vs. electroplated copper, inhomogenities and anisotropy in woven substrate) and user related simulation inaccuracies. In addition, the boundary conditions (side walls and top cover vs. open) matter.

So what is the purpose of your testcase? Public benchmarking is usually prohibited by the software license agreement. If you just need a second opinion on your simulation, we need geometries and more complete material data.
 

Here are some more infos and a gerber file with the top-layer data. I know the problems of comparing measurements and simulations from my own experience. Nevertheless, I would be interested to know how far away the results are.
 

Attachments

  • bandpass_info.pdf
    44.6 KB · Views: 109
  • bandpass.zip
    733 bytes · Views: 93

Your gerber file is not correct. It seems that you have all layers in one file (or just the ground plane width drills?). Gerber requires one file for each layer. You better use DXF instead, and please import yourself to check that it's complete.
 

Thanks for trying. Now only the top layer in the Gerber file.
 

Attachments

  • bp.zip
    1.3 KB · Views: 96

Ok, I can add some FDTD results from Empire XPU. At the moment, I use that for RFIC on-chip antenna simulation mostly (240GHz on-chip antenna etc) so hopefully there's no user mistakes for this PCB use.

Model details:
Substrate er=3.66, tand=4e-3
Conductor copper for top metal, PEC boundary for bottom side
Side walls absorbing sheet, top cover PML6
Microstrip ports
Mesh setting "very fine", 1.5MCells

bpf_open_empire.PNG

bpf_2d.PNG

bpf_3d.PNG

bpf_mslport.PNG

For comparison to measured, we should check the influence of via position offset. What do you think,is +/-100 µm reasonable?
 
Last edited:
For comparison to measured, we should check the influence of via position offset. What do you think,is +/-100 µm reasonable?

A quick test shows that via hole placement is indeed very sensitive. This is what we get with via displacement (all vias same offset) of x=50µm, y=50µm and x=100µm, y=100µm:

bpf_viaoffset.PNG
 

Thank you for the simulations! The results are not far from my simulations. If possible and allowed, please save the S-parameter file (Touchstone format xx.s2p), so I will summarize all curves in one plot.
 

Thank you for the simulations! The results are not far from my simulations. If possible and allowed, please save the S-parameter file (Touchstone format xx.s2p), so I will summarize all curves in one plot.

I prefer not to do that, because everyone here uses slightly different model settings and such a comparison can be misunderstood by readers as a tool accuracy benchmark.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top