Hi chaps,

I've decided to bite the bullet and try to build an RF filter for
40Mhz. This filter will ideally have a very, very sharp characteristic
at one single spot frequency +-20Khz and attenuate the crap out of
anything either side of this. It'll need to be tunable over a range of
say 200Khz. Can anyone give me a steer on what type of arrangement
would be best suited to fit this purpose?
Thanks,

p.
--

"I believe history will be kind to me, since I intend
to write it." - Winston Churchill

On Sat, 26 Jul 2003 12:35:11 -0400, Joseph Legris
wrote:


What is the intended application?

I want to increase the sensitivity of a radio control rx. To this end,
I'm going to build an RF pre-amp, which needs to be highly selective
to avoid the probablility of increased ajacent channel interference.
Hence the need for a sharp filter in conjunction with the pre-amp.

--

"I believe history will be kind to me, since I intend
to write it." - Winston Churchill

On Sat, 26 Jul 2003 12:35:11 -0400, Joseph Legris
wrote:


What is the intended application?

I want to increase the sensitivity of a radio control rx. To this end,
I'm going to build an RF pre-amp, which needs to be highly selective
to avoid the probablility of increased ajacent channel interference.
Hence the need for a sharp filter in conjunction with the pre-amp.

--

"I believe history will be kind to me, since I intend
to write it." - Winston Churchill

On Sat, 26 Jul 2003 17:42:19 +0100, Paul Burridge
wrote:

On Sat, 26 Jul 2003 12:35:11 -0400, Joseph Legris
wrote:


What is the intended application?

I want to increase the sensitivity of a radio control rx. To this end,
I'm going to build an RF pre-amp, which needs to be highly selective
to avoid the probablility of increased ajacent channel interference.
Hence the need for a sharp filter in conjunction with the pre-amp.

This is a strange way of achieving your aim. You should be going for
adjacent channel selectivity in the IF, not the front end. That way
you can use a fixed frequency filter, and it can be as lossy as you
like - and it will be lossy to achieve these kinds of selectivity with
achievable unloaded Q.

If you put this filter at the RF stage where it will help with
adjacent channel selectivity,m it must be before the first amplifier,
and that will have killed your hoped-for sensitivity increase. If you
put it after that amplifier, then you may as well leave it where it
belongs - in the IF.

Radio design is always a compromise between sensitivity, selectivity
and large-signal performance. The standard topologies haven't emerged
by accident, I'm afraid.

d

_____________________________

http://www.pearce.uk.com

On Sat, 26 Jul 2003 17:42:19 +0100, Paul Burridge
wrote:

On Sat, 26 Jul 2003 12:35:11 -0400, Joseph Legris
wrote:


What is the intended application?

I want to increase the sensitivity of a radio control rx. To this end,
I'm going to build an RF pre-amp, which needs to be highly selective
to avoid the probablility of increased ajacent channel interference.
Hence the need for a sharp filter in conjunction with the pre-amp.

This is a strange way of achieving your aim. You should be going for
adjacent channel selectivity in the IF, not the front end. That way
you can use a fixed frequency filter, and it can be as lossy as you
like - and it will be lossy to achieve these kinds of selectivity with
achievable unloaded Q.

If you put this filter at the RF stage where it will help with
adjacent channel selectivity,m it must be before the first amplifier,
and that will have killed your hoped-for sensitivity increase. If you
put it after that amplifier, then you may as well leave it where it
belongs - in the IF.

Radio design is always a compromise between sensitivity, selectivity
and large-signal performance. The standard topologies haven't emerged
by accident, I'm afraid.

d

_____________________________

http://www.pearce.uk.com

In sci.electronics.design, Paul Burridge
wrote:


Hi chaps,

I've decided to bite the bullet and try to build an RF filter for
40Mhz. This filter will ideally have a very, very sharp characteristic
at one single spot frequency +-20Khz and attenuate the crap out of
anything either side of this. It'll need to be tunable over a range of
say 200Khz. Can anyone give me a steer on what type of arrangement
would be best suited to fit this purpose?

I've got that deja-google feeling all over again:


http://groups.google.com/groups?hl=e...com%26rnum%3D6

I like the idea of downconverting to an IF, filtering using
standard IF technology, and (if you want the output to be the same
frequency band as the input) upconverting using the same local
oscillator as the downconverter. Just change the LO frequency (maybe
use a frequency synthesizer for stability) to do tuning.

Thanks,

p.
--

"I believe history will be kind to me, since I intend
to write it." - Winston Churchill

In sci.electronics.design, Paul Burridge
wrote:


Hi chaps,

I've decided to bite the bullet and try to build an RF filter for
40Mhz. This filter will ideally have a very, very sharp characteristic
at one single spot frequency +-20Khz and attenuate the crap out of
anything either side of this. It'll need to be tunable over a range of
say 200Khz. Can anyone give me a steer on what type of arrangement
would be best suited to fit this purpose?

I've got that deja-google feeling all over again:


http://groups.google.com/groups?hl=e...com%26rnum%3D6

I like the idea of downconverting to an IF, filtering using
standard IF technology, and (if you want the output to be the same
frequency band as the input) upconverting using the same local
oscillator as the downconverter. Just change the LO frequency (maybe
use a frequency synthesizer for stability) to do tuning.

Thanks,

p.
--

"I believe history will be kind to me, since I intend
to write it." - Winston Churchill

Take a look at lowband filters and duplexers from CellWave, Telewave,
etc.
Your filter is going to be a coaxial cavity about 2mx20cm with piston
for tuning.
I am sure there are some better approaches to your task


Vladimir Vassilevsky, Ph.D.

DSP and Mixed Signal Design Consultant

http://www.abvolt.com


Paul Burridge wrote:

Hi chaps,

I've decided to bite the bullet and try to build an RF filter for
40Mhz. This filter will ideally have a very, very sharp characteristic
at one single spot frequency +-20Khz and attenuate the crap out of
anything either side of this. It'll need to be tunable over a range of
say 200Khz. Can anyone give me a steer on what type of arrangement
would be best suited to fit this purpose?
Thanks,

p.
--

"I believe history will be kind to me, since I intend
to write it." - Winston Churchill

Take a look at lowband filters and duplexers from CellWave, Telewave,
etc.
Your filter is going to be a coaxial cavity about 2mx20cm with piston
for tuning.
I am sure there are some better approaches to your task


Vladimir Vassilevsky, Ph.D.

DSP and Mixed Signal Design Consultant

http://www.abvolt.com


Paul Burridge wrote:

Hi chaps,

I've decided to bite the bullet and try to build an RF filter for
40Mhz. This filter will ideally have a very, very sharp characteristic
at one single spot frequency +-20Khz and attenuate the crap out of
anything either side of this. It'll need to be tunable over a range of
say 200Khz. Can anyone give me a steer on what type of arrangement
would be best suited to fit this purpose?
Thanks,

p.
--

"I believe history will be kind to me, since I intend
to write it." - Winston Churchill

Paul Burridge wrote in message . ..
I've decided to bite the bullet and try to build an RF filter for
40Mhz. This filter will ideally have a very, very sharp characteristic
at one single spot frequency +-20Khz and attenuate the crap out of
anything either side of this. It'll need to be tunable over a range of
say 200Khz. Can anyone give me a steer on what type of arrangement
would be best suited to fit this purpose?

It'll need to be a crystal filter, and your requirement that it must
be tunable means that you will convert down/up to an IF frequency and back
up/down again. (Well, you may not have to convert back up again but you
don't tell us your application).

Problem with the IF and conversion is the production of images. Images
won't be a killer problem because your tuning range is really quite narrow.

If you wanted to really cheapskate out some ceramic IF filters also seem
to meet your stated needs.

Tim.