[Help] Microstrip Filters

[neferet]

Hi all,
I have to make a simulation and then the fabrication of a microstrip filter (at 1, 1.5 or 2GHz). I have decided to make an edge coupled bandpass filter(it may be something else) by using Ansoft Designer SV. However, my substrate is not in the Designer. So, I first make a filter by using the filter tool and then I change the permitivity. However, I cannot add the loss tangent when I am inside the filter tool. When I add it afterwards, the response does not look as before(as usual). This time I try to optimize it by changing the dimensions by myself (because I think SV does'nt have tuning property), but I end up with 400-500MHZ bandwidth and 2.5dB loss at the center freq. Is this so unsual at this frequency or am I able to cope it by another way. Any decision?


Thanks...

 

[sunnySan]

Parallel coupled line could also be a nice way to create microstip filter, following the thebyshev or butterworth theories.

 

[beginer]

Re: [Help-40pts] Microstrip Filters

Hi, in Designer is possible deffine your concrete substrate and sumulate your concrete fiter on this material.
This procedure is explaned in manual.

Hi E.

 

[borich_03]

Re: [Help-40pts] Microstrip Filters

Hi all,
....This time I try to optimize it by changing the dimensions by myself (because I think SV does'nt have tuning property), but I end up with 400-500MHZ bandwidth and 2.5dB loss at the center freq. Is this so unsual at this frequency or am I able to cope it by another way. Any decision?
40 pts for a reasonable and working advice.
Thanks...

"2.5dB loss" is a common specification for microstrip filters.
If you want to have good insertion loss, you need use other types of microwave filters with high-Q resonators: combline, interdigital, helical ....

 

[birdone]

you can select a high Q resonator,

 

[mwmmboy]

hi, in my opinon an interdigital filter and a hairpin filter can have better performance at those frequencies.

bye

 

[fredy]

hi

I think 2.5 dB is unusual.

the main loss in combline filter is reffered to lump capacitances and dielectric loss have a little effect in this filter.

but I must know that what is your substrate and I think you use ideal capacitances, so if your loss tangent is less than .003 (such as RF4003) then your answer is wrong or your design is not true.

I wrote software for this type of filter and if you want I can check it with my program.

send me your dimensions (If you want)

best regards

fredy

 

[id_guy]

Ha,how about cavity filter

 

[Kit-the-great]

On my experience, the main loss in combline filters is resistive loss in lines and box walls and jointings. A chip capacitors not good for this filter type due to high parameters dispersion; topological caps more good.

 

[neferet]

Thank you all,
I gave up making a filter with this material. (Electric Permittivity = 4.7, Loss Tangent = 0.019, Thickness = 1mm). I have made a branch-line coupler instead.