[TowerTalk] Stub Filters

[Jim Lux]

At 01:40 AM 3/24/2006, Peter Chadwick wrote:
>As Gary said, L-C filters are a better way to go. I'd design bandpass 
>filters for each band. There are various programs around that will design 
>a filter for you, although they can come up with awkward values, and more 
>advanced design techniques are useful. A good introductory book (although 
>pretty ancient, but the principles haven't changed) is 'Simplified Modern 
>Filter Design' by Geffe. There are a fair number available on Abe books.
>73
>Peter G3RZP


Depending on what your fabrication capability and budget is, you might also 
consider ordering a filter from a commercial filter 
manufacturer.  Companies like TTE (http://www.tte.com/) and others: get a 
copy of an issue of "RF Design" or "Microwave Journal" and look at the ads.

These folks can crank out almost any filter characteristic you want.  I've 
bought lots of lumped LC filters for HF frequencies at work (for breadboard 
IF strips in microwave systems, etc.) and they tend to be a couple or three 
hundred dollars each, with a 4-6 week delivery.

Yes, a couple hundred bucks is a lot more than the cost of the parts if you 
build it yourself, but, on the other hand, you don't spend hours and hours 
trying to find which standard components will give you the response you 
want, then building one and seeing if it works (because you didn't model 
all the parasitics), etc.


On the other hand, there is a satisfaction involved in learning something 
new (filter design) and developing a skill that relatively few people have 
(tuning filters). If you DO build one yourself, you should look into 
adjustable inductors and capacitors for the filter, and figure out how to 
tune it up yourself.  (A radio with a computer interface can serve as a 
poor man's spectrum analyzer)