[R-390] Preselector

Sat Aug 21 17:26:33 EDT 2004

Hi

That's a couple of different questions on a couple of different topics. 
Yell if I miss one of them.

Back when I worked for Motorola one of the standard products we made 
was a crystal filter that you stuck in the antenna line of your two way 
FM radio. Since the filter was in front of the entire receiver I 
suppose that's about as far forward in the process as you can possibly 
push the filter. I have seen the same thing done at HF both by Ham's 
and in military settings. NASA uses the same idea on their command 
destruct receivers.

One problem with  putting a poor little filter way up by the antenna is 
that filters have overload problems and produce intermodulation 
products. You can have a situation where the little narrow  crystal 
filter actually produces a worse result than a big helical LC filter 
with a wider pass band. It all depends on how far away the overloading 
signals are from your channel.

Another problem with any kind of narrow front end filtering is that 
they generally have measurable insertion loss. If you have a six db 
loss in the front end filter you drop the noise figure of the radio by 
at least six db. If you are trying to pick a weak signal this is 
probably not a real good idea. To use your analogy of whispering, it 
only works in a quiet room. Making the room more noisy isn't going to 
help things much ....

Cascading filters to get better selectivity is a workable idea. You 
have to be careful about pass band ripple. Depending on just where each 
filter peaks and dips you can wind up with some odd results. One 
solution is to use filters that have been designed from scratch to be 
cascaded. Another solution is to make one filter much wider than the 
other and pick a wide filter with very little ripple in the middle of 
it's pass band.

Using the minimum amount of power to get a message through is one way 
to reduce the probability of intercept.  One issue is the relative 
location of the intercept station If my R-390 has a better path to the 
transmitter than the intended receiver then my R-390 is going to get 
more power than the guy the message is intended for. Since propagation 
can be very strange stuff a better path may or may not be related to 
being closer to the transmitter. There are a number of stories about HF 
and even VHF intercept taking place at significant distances from the 
transmitter. Of course they may just be stories ....

Like it or not there is natural noise out there. It's a fact of life. 
As long as you are operating on a terrestrial path it's going to be a 
very significant limiting factor on what you can or can not do. Even 
with space communication they get to a noise limited situation, they 
just have to work harder to get there. Different modulation techniques 
result in different relationships between channel noise and noise after 
demodulation. They all run out of steam at some point though.

A guy named Shannon came up with a relationship about all this back in 
the 1940's. Still seems to hold true today ..... You can go real slow 
and real narrow and use low transmit power. You can go nice and fast 
and wide and use lots of transmit power.

One interesting experiment with your R-390. Next time there's a 
lightning storm running a couple of towns over try to listen to a 
distant station through a narrow filter and through a wide one. 
Narrower is not *always* better .....

You can design a system to perform optimally under a stated set of 
conditions. That often means it will be non-optimum under a much wider 
set of conditions. No free lunch .....

The R-390 is an amazing compromise design that works awfully well under 
a wide range of conditions. It's a few db shy of being optimum on a 
number of measures. It is a good example of covering a lot of bases 
well rather then just one base and throwing the rest of it away. Even 
so it's only a few db away from perfect ....

	Take Care!

		Bob Camp
		KB8TQ

On Aug 20, 2004, at 10:01 PM, Dan Merz wrote:

> Bob,  ok,  I am listening and becoming informed.  My idea probably has 
> no
> bearing on making a 390a work better.  I do remember an article by D.
> Langford where he put a 6 khz filter in a 390a in front of all the 
> other
> filters to improve the set,  closer to the front end.  My thinking 
> started
> extending to the idea that if you didn't want to tune the radio and 
> were
> happy with a fixed channel,  how close to the antenna could you put a
> crystal or mechanical filter,  and derive much improved performance.  
> Have
> such receivers been built?  It would seem that whispering is one of the
> earliest forms of hiding your conversation from another listener.   
> Has this
> principle been extended to radio transmission by making the receiver 
> capable
> of hearing such weak signals that no one else,  except someone with 
> the same
> type of receiver,  can hear the deliberately weak signal.  Or is this
> wishful thinking that such a concept would work?  I suppose it's 
> simpler to
> just encode the information by other means.    Dan
>
>
> ----- Original Message -----
> From: "Bob Camp" <ham at cq.nu>
> To: "Dan Merz" <djmerz at 3-cities.com>; "R-390 HF Receiver List"
> <r-390 at mailman.qth.net>
> Cc: "Charles B" <ka4prf at us-it.net>
> Sent: Friday, August 20, 2004 7:35 AM
> Subject: Re: [R-390] Preselector
>
>
>> Hi
>>
>> The balanced input to the R-390 has some fairly good filters on it
>> already. In order to do better you would have to go to a fairly 
>> complex
>> filter. The exception to this would be a notch filter for something
>> like broadcast band overload.
>>
>> So far pretty straightforward ..... now off the deep end .....
>>
>> Passive LC filters are not quite as simple as they look. As you add
>> sections of a filter together they interact with each other. A simple
>> example is  to whip up two identical three element pi section 1 MHz 
>> low
>> pass filters. Check them out to make sure they 3 db right at 1 MHz.
>> Next put them in series and check out the result. For most filters you
>> now have something that peaks significantly and may or may not cut of
>> anywhere near 1 MHz. If you do the math for a proper five element
>> filter you will find that the parts values are not the same as for the
>> two three sections put together. The two filters interacted in a
>> fashion that is predictable (the math works) but not intuitive  (1 + 1
>> does not = 2). Filters can be cascaded but you can't design them to 
>> run
>> into and out of a resistive load and then go and run them into
>> something way different ....
>>
>> Now back to reality .... sorry for the drop off into theory land.
>>
>> If you have a passive filter on the front end of the R-390 *and* its
>> narrower than the filter that's already in the R-390 front end then
>> they probably are going to interact. The result may be a filter with
>> more loss or a wider pass band than you expected.
>>
>> A solution to the problem is to isolate the two filters. That way they
>> won't interact. The good old way to do this was to slap a tube in
>> between the two filter sections. That keeps each filter so it runs 
>> like
>> it should. Another equally good solution is to slap a 10 db attenuator
>> in between the two filters. Either way the filters each see a proper
>> resistive load and they work right.
>>
>> If you go with gain to isolate the two filters then you have a real
>> possibility of overload and distortion. If you put in a pad then you
>> have cut your sensitivity. Either way you are trading off one thing 
>> for
>> another. I'm not suggesting that you can't do a better job than was
>> done on the radio originally. All I'm saying is that it's a fairly
>> complex thing to do.
>>
>> Take Care!
>>
>> Bob Camp
>> KB8TQ
>>
>>
>> On Aug 19, 2004, at 11:03 PM, Dan Merz wrote:
>>
>>> Bob,  I've heard this said before - and I always delight when my
>>> receiver
>>> shows noise when the antenna is connected - a good sign.  But what if
>>> the
>>> preselector  that I add has narrower bandwidth than the front end of
>>> the
>>> receiver I'm using but is still wide enough for what I want to hear?
>>> Won't
>>> I see a benefit by adding this preselector?
>>>
>>> another misinformed listener ready to be informed,
>>>
>>> Dan.
>>> ----- Original Message -----
>>> From: "Bob Camp" <ham at cq.nu>
>>> To: "Charles B" <ka4prf at us-it.net>; "R-390 HF Receiver List"
>>> <r-390 at mailman.qth.net>
>>> Sent: Thursday, August 19, 2004 4:28 PM
>>> Subject: Re: [R-390] Preselector
>>>
>>>
>>>> Hi
>>>>
>>>> One place that a pre-amp might help an R-390 is up on 10 meters. The
>>>> radio is plenty sensitive enough on the lower bands but as frequency
>>>> goes up the noise level off a typical antenna goes down. The antenna
>>>> it's self gets smaller and both the man made and natural noise drop
>>>> off
>>>> some as frequency goes up. You can come up with a sub one db noise
>>>> figure pre amp for just about any band from 1 GHz on down. That has 
>>>> to
>>>> be significantly better than the front end of most HF radios.
>>>>
>>>> A simple way to check if it will help:
>>>>
>>>> Listen to the noise out of the radio as you attach the antenna. If 
>>>> it
>>>> goes up you don't need a pre-amp. Since you have an antenna trimmer
>>>> involved it's a little more complicated than with a rice box, but 
>>>> the
>>>> net result is the same.
>>>>
>>>> Take Care!
>>>>
>>>> Bob Camp
>>>> KB8TQ
>>>>
>>>>
>>>> On Aug 19, 2004, at 5:04 AM, Charles B wrote:
>>>>
>>>>> Hi all,
>>>>>
>>>>> What happens if you install a small preselector or preamplifier in
>>>>> front of an R-390A?  Does it help or hinder?
>>>>>
>>>>> Chuck
>>>>> ka4prf at us-it.net
>>>>>
>>>>> ______________________________________________________________
>>>>> R-390 mailing list
>>>>> Home: http://mailman.qth.net/mailman/listinfo/r-390
>>>>> Help: http://mailman.qth.net/faq.htm
>>>>> Post: mailto:R-390 at mailman.qth.net
>>>>> Unsubscribe: http://mailman.qth.net/mailman/options/r-390
>>>>>
>>>>
>>>> ______________________________________________________________
>>>> R-390 mailing list
>>>> Home: http://mailman.qth.net/mailman/listinfo/r-390
>>>> Help: http://mailman.qth.net/faq.htm
>>>> Post: mailto:R-390 at mailman.qth.net
>>>> Unsubscribe: http://mailman.qth.net/mailman/options/r-390
>>>
>>>
>>
>
>