Design of Microwave BPF

[Lokesh Kumar]

microwave bpf in ads

i want to design a BPF (butter-worth or chebyshev) with following specifications:
1)frequency band of operation 0.5Ghz-3Ghz.
2)center frequency 1.8Ghz
3)1db Bandwidth = 25Mhz
4)Pass-band insertion loss should be less than 2dB
5)pass-band VSWR = 1.5:1
6)rejection @ +50Mhz to -50Mhz away from center frequency should be 60dB min.
7)source and load impedence =50ohm

please give me rough idea how this filter look like(diagram) and how to calculate order of the filter???

 

[vfone]

bpf specs

Download one of these free (trial) software and it will give you an idea.
**broken link removed**
**broken link removed**

 

[radiohead]

waveguide wr770

This will be a very hard design - your relative bandwidth is approx. 1% meaning that you'll need good resonators:
* Tuned to the correct center frequency
* Very good Q-factor (>400)

And 60 dB at +/- 50 MHz is simply impossible; you would need a very high filter order.

The only technology that could result in this sort of suppression is SAW (don't know if it can be done at 1.8GHz). But then the insertion loss would be very high, up to 20dB or so.

I think there's something wrong with your specs. Why not use a superhet, mix down to a lower IF and filter there? The filter will be much easier to implement and on top you'll have frequency selectivity.

 

[Lokesh Kumar]

bpf 50mhz

This will be a very hard design - your relative bandwidth is approx. 1% meaning that you'll need good resonators:
* Tuned to the correct center frequency
* Very good Q-factor (>400)

And 60 dB at +/- 50 MHz is simply impossible; you would need a very high filter order.

The only technology that could result in this sort of suppression is SAW (don't know if it can be done at 1.8GHz). But then the insertion loss would be very high, up to 20dB or so.

I think there's something wrong with your specs. Why not use a superhet, mix down to a lower IF and filter there? The filter will be much easier to implement and on top you'll have frequency selectivity.

don't bother about the frequency range of operation.....

Added after 5 hours 4 minutes:

come on fellas ...gimme something to work on....

 

[radiohead]

comline　bpf　design

OK if you really want to get it done your way I suggest you to make a waveguide filter in WR-770 which is rectangular 7.700 x 3.385 inches. Very good Q, tunable with screws and you can implement way too many poles.

Good luck.

 

[Lokesh Kumar]

tunable waveguide combline filters design

OK if you really want to get it done your way I suggest you to make a waveguide filter in WR-770 which is rectangular 7.700 x 3.385 inches. Very good Q, tunable with screws and you can implement way too many poles.

Good luck.

Thaks for ur help.
It will be highly appreciated if u can provide any tutorial on waveguide filter.

 

[Manjunatha_hv]

Hello Lokesh Kumar,


Try this, which may help you to design Waveguide Band Pass Filter...
This site is dedicated to design and analysis of micro- and millimeter wave filters from 0.5 GHz to above 100 GHz.

https://www.guidedwavetech.com/index.htm


---manju---

 

[Phytech]

You may try combline filter. It looks like an easy filter for comline filter.

 

[Lokesh Kumar]

You may try combline filter. It looks like an easy filter for comline filter.

i want to make the filter with microstrip coupled line....
i m not able to determine passband frequency in it....i m doing it on paper using chebyshev and not using any simulation.....

 

[Phytech]

You may try combline filter. It looks like an easy filter for comline filter.

i want to make the filter with microstrip coupled line....
i m not able to determine passband frequency in it....i m doing it on paper using chebyshev and not using any simulation.....

Microstrip will not work due to its low unloaded Q. Your filter has a narrow bandwidth, high rejection, and low insertion loss, therefore requires high Q. Combline filter is the best I can see, you may even use cross coupled filter design if needed. You may also try ceramic filter, which gives you higher insertion loss but much lower cost.

 

[radiohead]

Combine filters are microstrip-based, so same problems with limited Q and tolerance . Am I missing something here?

 

[Lokesh Kumar]

Thanks for the help friends...

 

[micky74]

Combine filters are microstrip-based, so same problems with limited Q and tolerance . Am I missing something here?

The combline filters being mentioned here are the cavity type, so it has low insertion loss unlike microstrip.