Re: [SARA] Frontend filter for 1420 MHz ?

[ARNO...@aol.com]

Hello  ,

maybe someone from the list can help.

I need a design  for a filter ahead of the LNA for my 1420 MHz reception.

The LNA  I use seem to be to sensetive outside the design frequency.I have very very strong signals here around 1750 MHz and 800-900 MHz witch get into the frontend and cause problems.

I know that it is not a good idea to put a filter ahead of an LNA becaus of the extra noise but it looks like I have no other choise.

What would be a good design ?

 

thanks for Your help

Wolfgang , Dj3qd

[Marcus D. Leech]

It'll be hard to find one that is both selective enough, and has
low-enough insertion loss.

I know that David Halley has some combline filters that he's been using
(and has spares to sell). I think the insertion
loss is likely around 1.0dB, though. Still, if the RFI is *that*
strong, you might have no choice.

You can easily build a single-element coaxial filter using copper water
pipe and a couple of connectors. Tuning it will
be a challenge without an RF lab, however.

Radio Astronomy Supplies also sell a couple of different filters--so you
might look into that also.

The other possibility is to put a filter *after* the first LNA--in which
case, the loss isn't nearly as important.
As long as your front-end isn't being driven into compression, this
approach usually works well, and you have
a much larger choice of filters available.

I'm having a test run of filters made in China with an Fc=1420.41Mhz.
If the test samples work out, I'll be
ordering 1000 pieces, if I can get enough interest together to share
the financial burden (about US$500.00 plus shipping).
These are ceramic dielectric filters which will have an insertion loss
of about 3.5dB, which is negligible when used
after the first LNA.

Cheers

--
Marcus Leech
Principal Investigator, Shirleys Bay Radio Astronomy Consortium
http://www.sbrac.org

[Malcolm Mallette]

We used the Radio Astronomy Supplies hydrogen line filter after the Radio
Astronomy Supplies LNA and before a sat. tv type line amp. The LNA, filter
and line amp were at the feedhorn. Another line amp was at the receiver. We
needed 60 dB gain before the receiver. We had a DC block between the filter
and the line amp.

Malcolm
WA9BVS

[oxl...@att.net]

Wolfgang

What type of antenna are you using? If it has a waveguide feed, the easiest way to eliminate the 800-900 is by sizing the waveguide with a low frequency cutoff above 900 MHz. This acts as a filter without the insertion loss penalty unless you set the cutoff too close to 1420. The loss goes up near the cutoff frequency. 

 

Are you located near a mobile phone cell site? If so, this may be the source of your interference at 800-900 MHz. In addition, the 1750 MHz might be a second harmonic.  

 

I agree with the other respondents that you should avoid a filter in front of the LNA if you want to maintain a good noise figure and sensitivity.

Paul Oxley

-------------- Original message from ARNO...@aol.com: --------------

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
-~----------~----~----~----~------~----~------~--~---

[Charles S. Osborne]

Wolfgang,

 

The lowest loss approach I've seen is to use the filter as the input match for the preamp. But I don't have a design readily at hand for 1420. Seems like I did see something for 1296 EME that might be scalable from DuBus magazine a few years ago. The idea is to use a tuned line single resonator to get a bandpass response. The gate match is the tricky part since it would need to be tapped typically at a higher impedance point close to the resonator end.

 

The trick is doing this sort of thing without causing out of band oscillations since any type filter looks like 0 dB return loss in the rejection band.

 

In some respects notching the cellular bands out with a shorted stub might also have merit. But that too is tempting faite since that's exactly how an oscillator is made on purpose.

 

Another point worth making is that sacrificing a tenth dB of NF by increasing the output intercept of the preamp via more drain current thru rebiasing, may also help with intermod from the cellular site.

 

73,

Charles, K4CSO

----- Original Message -----

From: ARNO...@aol.com

To: sara...@googlegroups.com

Sent: Thursday, August 21, 2008 9:49 AM

Subject: [SARA] Re: Frontend filter for 1420 MHz ?

[ARNO...@aol.com]

Hello Charles,

thanks for the input .I will look into your suggestions.

 

73 Wolfgang Dj3qd

[ARNO...@aol.com]

Hello Paul,

I use a septum feed designed for 1420 MHz .I will calculate wher the low cutoff is with the present dimensions.

Yes the next mobile phone cell site is about about 500 m away.I even see the 800 and 1750 signals if I

cover the fron end of the feed with aluminum foil ,so it must be very strong.If I replace the feed with a 50 ohm input resistor (a 30 db attennuator ) no signal is detectable  ,only the noise of the amplifier.

 

thanks for your input 

Wolfgang , Dj3qd 

 

 

 

 

What type of antenna are you using? If it has a waveguide feed, the easiest way to eliminate the 800-900 is by sizing the waveguide with a low frequency cutoff above 900 MHz. This acts as a filter without the insertion loss penalty unless you set the cutoff too close to 1420. The loss goes up near the cutoff frequency. 

 

Are you located near a mobile phone cell site? If so, this may be the source of your interference at 800-900 MHz. In addition, the 1750 MHz might be a second harmonic.  

 

I agree with the other respondents that you should avoid a filter in front of the LNA if you want to maintain a good noise figure and sensitivity.

Paul Oxley

-------------- Original message from ARNO...@aol.com: --------------

[ARNO...@aol.com]

Hello Malcom,

I will try Your approach.

 

thanks for the input

Wolfgang Dj3qd 

 

 

We used the Radio Astronomy Supplies hydrogen line filter after the Radio
Astronomy Supplies LNA and before a sat. tv type line amp.  The LNA, filter
and line amp were at the feedhorn. Another line amp was at the receiver. We
needed 60 dB gain before the receiver. We had a DC block between the filter
and the line amp.

Malcolm
WA9BVS

----- Original Message -----
From: "Marcus D. Leech" <mar...@propulsionpolymers.com>
To: <sara...@googlegroups.com>
Sent: Thursday, August 21, 2008 11:44 AM
Subject: [SARA] Re: Frontend filter for 1420 MHz ?


>

> ARNO...@aol.com wrote:
>> Hello  ,
>> maybe someone from the list can help.
>> I need a design  for a filter ahead of the LNA for my 1420 MHz reception.
>> The LNA  I use seem to be to sensetive outside the design frequency.I
>> have very very strong signals here around 1750 MHz and 800-900 MHz
>> witch get into the frontend and cause problems.
>> I know that it is not a good idea to put a filter ahead of an LNA
>> becaus of the extra noise but it looks like I have no other choise.
>> What would be a good design ?
>>
>> thanks for Your help
>> Wolfgang , Dj3qd
>>

[ARNO...@aol.com]

Hello Marcus ,

are thos filters designed for 50 ohm ?

If yes I would be interessed .

best regards

 

Wolfgang ,Dj3qd

[Marcus D. Leech]

ARNO...@aol.com wrote:
> Hello Marcus ,
> are thos filters designed for 50 ohm ?
> If yes I would be interessed .
> best regards
>
> Wolfgang ,Dj3qd
>
>

Yes, 50 Ohm.

I expect to take deliver of the first 10 samples sometime in the coming
week. I'll connectorize one or two of them, then
test them in the lab at work.

[Kimmo]

Hi

If you are in an urban area then a bandpass filter before a LNA may be
a necessity.
I was reading an article about the 5m Argos telescope, used for 21cm
HI observations. The article is at

http://e-collection.ethbib.ethz.ch/eserv/eth:28917/eth-28917-01.pdf

They are using a filter from ID-elektronik, see their www-page at

http://www.id-elektronik.de/en/index.htm

The filter has only -0.5dB loss and is relatively cheap, about 100
euros two years ago.

cheers, Kimmo

----------------------------------------------------------------------------------------