Single Stub Matching Network in HFSS

[mprathap]

Hello All,

I am trying to emulate the single stub matching network example given in Pozar's Microwave Engineering using a microstrip realization on a 14mil FR4 (dielectric constant = 4.4) board. The 50ohm thickness works out to be 0.675mm for this board. However I am not getting anywhere close to 50 ohms at the port in my HFSS design. I have attached the design file and an extract of the problem I am trying to emulate.

Could someone take a look at this design file and let me know if I am doing something wrong?

For your information, the parallel RC equivalent of the series RC circuit in the example (60ohms, 0.995pF) is 167ohms and 0.6367pF which I have used in the HFSS design file.

And please remove the .txt extension from the hfss design file since the forum won't let me upload a .hfss file.

Thanks a million in advance.

--

 

[tyassin]

Hi

Please use WinRar when attaching files.

Thanks

 

[shlomo22]

The port is wrong. A lumped port should go between two conductors, with the integration line running from one edge to the other edge, touching the conductors. Your port goes way up into space, it does not end on the second conductor (the signal line). It is shaped like a waveport, not like a lumped port.

 

[mprathap]

The port is wrong. A lumped port should go between two conductors, with the integration line running from one edge to the other edge, touching the conductors. Your port goes way up into space, it does not end on the second conductor (the signal line). It is shaped like a waveport, not like a lumped port.

Thanks. I corrected the problem. The only difference I see now is that the matching has improved very close to 50 ohms at 2.54 GHz though in theory it should be at 2 GHz. Any ideas what I should change?

 

[de_prince]

Thanks. I corrected the problem. The only difference I see now is that the matching has improved very close to 50 ohms at 2.54 GHz though in theory it should be at 2 GHz. Any ideas what I should change?

Hi, how did you get calculation for stub geometry in microstrip and also the location?

 

[reidintransit]

The parameters that will effect matching and frequency are:
- material dielectric constant (not one you typically adjust)
- stub position (distance from load)
- stub length

 

[de_prince]

The parameters that will effect matching and frequency are:
- material dielectric constant (not one you typically adjust)
- stub position (distance from load)
- stub length

Hi Thanks!

I got some idea after I read "Microwave Engineering" by David M Pozar, page 239.

Other questions that still remain are, what should be the dimension of the stub? and what would be dimension of the short to GND connection? or should I just follow trace's dimension?

 

[reidintransit]

Pozar's method uses a 50 ohm line for the stub dimensions. For a shorted stub you are typically using a via that is determined by the manufacturing process, a cylinder. There are many papers on different types of stub geometries. Start with 50 ohms and then play with the width to get a feel for the impact.