Front end filter design problem

[atlantis7]

Hi

I was trying to design front end bandpass filters for the popular ham bands of 6m, 2m and 70cm. Enclosed you find the circuits I used and the resulting curves. In my opinion the curve for 6m is acceptable but for 2m it's too wide and the 70cm curve is no good at all, however these are the best results I could achieve.

Can somebody give me some hints about what's no good about it and what can be done better? The circuit should be the same for all bands only differing in values and it shouldn't be much different from the one I have at the moment.

Many thanks for all ideas.

Regards
Martin

 

[keith1200rs]

How did you do the design? How did you scale from one band to the next?

The 6m one doesn't look too bad so I scaled it to 70cm and attach the results. They aren't the same values as yours.

Keith.

 

[atlantis7]

Many thanks, you're right, like this the result is much much better, and I don't know what happened. Basically I did try to scale it and then tune the values to optimize but I must have made an error right at the beginning so that the tuning made the values even worse...

Do you think these results are ok for a front end filter for an fm receiver?

Regards
Martin

 

[keith1200rs]

I am not sure if they are good enough. I scaled the values from the 6m design but the out of band rejection seems poor at high frequencies which makes me think it is not a good design. I don't know where you got the original design from but I think I would start again if I were you.

Try the Nuhertz "Filterfree" program. It will only do a 3 pole filter but that will produce a decent filter for your purposes.

**broken link removed**

Keith.

 

[atlantis7]

Hi Keith

This is great piece of software, thanks for the link! The only problem I see are the pretty unrealistic values it calculates. Capacitors in the femtofarad range (for UHF I got values in that range) might be available from some sellers but the adjustable coils I'm going to use only go up to about 250 nH, not 4 uH... I need to fire up MWO and see what happens if the values are changed to more realistic ones.

Martin

 

[keith1200rs]

What are the design parameters you are using? Usually extreme values are caused by extreme requirements. For example, if you try to make a bandpass very narrow with a small number of poles I think the solution will be impractical values. Try increasing the bandwidth.

Keith.