Help to design a 1GHz to 2GHz BPF using FR4 in ADS software

[spaceboy5241]

I have one project using FR4 to design a center frequency 1.5GHz and range 1Ghz to 2Ghz Band pass filter in ADS software.
Any one know how to design.. i already simulate alot of pattern using ridge and stub . but can't get the desired result.

 

[biff44]

FR4 = bad for filters! Anisotropic, poor er control, which you need to make a bandpass filter. So UNLESS you are using only lumped elements soldered onto the board, use another type of board. FR4 would make poor, uncontrolled, transmission line resonators!

 

[Phytech]

You can design this filter with a cascaded LC low pass and LC high pass. You can use FR4 as the PCB, but you cam also use any other low loss PCB, such as Rogers PCB.

 

[Dr.Drew]

You can use HPF+LPF structure. For example HFCN-800 and LFCN-2150 Minicircuits. This filter will be smaller than any other.

 

[mwmmboy]

I think the better solution is a microstrip combline filter, but anyway FR4 is bad for filtering...

 

[lurchman]

FR4 is not the best way to go but it will work (sort of).
There are various boards out there like RT Duroid / Arlon but they can be expensive
and difficult to work with for various reasons.

You don't say what you are using this filter for. So it is a little difficult to
give advice.

As a rough rule of thumb if the bandwidth is greater than +-12% of the center
frequency you are better off designing a low pass followed by a high pass.

You do however need to seriously consider termination as connecting any type
of filter to a double balanced mixer for instance effects the mixers performance.

 

[razor28]

I also agree! FR4 is not the best way !

 

[biff44]

FR4 is perfectly fine IF all the reactive components of the filter are lumped elements.

If the transmission line lenghts and widths are used as filtering elements, then FR4 is a poor choice of substrates.

 

[biff44]

Just when you think you know an answer...you find something new! Here is a guy using FR4 up to 6 Ghz:

**broken link removed**

 

[volker_muehlhaus]

FR4 is perfectly fine IF all the reactive components of the filter are lumped elements.

I do not agree with that. FR4 is a lossy dielectric, so avoid building distributed capacitive elements. However, using metal traces as inductors works just fine on FR-4.