A little more explanation about what I posted earlier about a single
stub not in general being a very good filter...
Let's say you want to operate on 146MHz and want to get rid of a
strong signal from a nearby 156MHz transmitter. So, you think, let's
put a stub across the line which is shorted at the far end and is an
electrical half-wave on 156MHz. Fine, it'll get rid of the 156MHz
signal. -- And that's true enough. But what does it do on 146MHz??
If it's 50 ohm line, it will look like about -j10 ohms at 146! That's
not a very friendly thing to drop across the line on our operating
frequency. But there's a way around it. If you think of that
180-electrical-degrees on 156 as 180*146/156 = 168.5 degrees on 146,
then just add another stub there, also shorted, and 180-168.5 = 11.5
degrees long on 146MHz. That stub cancels out the effect, on 146MHz,
of the other stub, so there's essentially no effect at 146, but still
the dead short at 156. All this is idealized, for lossless line. In
such a case, where the rejection notch is close to the frequency you
want to pass, be sure to use low loss line. For example, if the
longer stub has 0.1dB loss and the shorter has 0.01dB loss, you'll
reflect not a short but about 0.6 ohms on 156, and not an open but
about 158 ohms resistive, representing considerable loss. That might
be something like RG174 (didn't check the numbers, but ballpark). If
you instead use some LMR600 line that has about 1/5 the loss, the
resistances will also be in that 5:1 ratio: about 0.12 ohms for the
short and a much nicer 780 for the open. Or you can make a really low
loss stub (a resonator) from moderately large diameter copper pipe.
Also, you can put trimmer caps across the ends of stubs to make them
much easier to tune than by using the "cut and re-solder" method.
If you thrive on just numbers, you can get freeware programs that will
calculate all that sort of thing for you, or just do it yourself on a
calculator or in a calculating program like Excel or Scilab. I much
prefer graphical representations, and use a Smith chart program that
does calcs that are just as good but displays the results graphically.
I use WinSmith, but there are freeware ones that will do similar
things. RFSim99 is just one of several.
Cheers,
Tom
(PS to Jim: NN7N is about a hundred feet away at the moment...)
"Jim" wrote in message ...
--(Corrected per Jim's updated posting...)--
Shorted 1/4 wave coax (twin lead, strip line, ect) produces a HI -Z at
resonance
OPEN 1/4 wave coax (ect) produces a dead short at resonance
SHORTED 1/2 wave coax (ect) produces dead short
OPEN 1/2 wave coax (ect ) produces HI-Z at resonance
This phonominom repeats every 1/4 wave (ad nausium)
If can find, look at July, 1965 edition of 73 magazine-- shows how to
calculate coaxial
striplines, ect for vhf/uhf circuits . if need (and these BIG files , 11
pgs) have in jpg, and some of it in word.
Jim NN7K
--
No trees were killed in the sending of this message. However, a large number
of electrons were terribly inconvenienced !
" People who never get carried away, should be! " --- Malcom Forbes
"Tom Bruhns" wrote in message
m...
If you just want to pass your signal, what's wrong with just
connecting the coax from the source to the load?
Or do you also want some filtering? If that's what you want, consider
two stubs. (Actually, a single stub is generally not a very good
filter...) These two will be attached to your feedline at the same
place. One will be shorted on the other end, and the other will be
open on the other end. Their total length will be 1/4 wave
(accounting for the line's propagation velocity). So you're feeding
your line through a 1/4-wave resonator, tuned to your signal
frequency. But the shorted part will reflect back a short at some
frequency, and knock that one out. And the open part will do
likewise. And you can so a similar thing, but with different notch
frequencies, if you use a half-wave resonator open on both ends, or
one shorted on both ends. The key is all these will pass the
frequency the resonator is tuned to. Beware of resonator losses if
you tap way down on it.
RFSim99 is a good tool for playing with this sort of thing. You can
readily see the inevitable harmonic responses that occur with stubs:
things tend to repeat at even and odd multiples of lowest resonance
frequencies. But you can get more accurate results by including the
line loss in your calcs. There are generally only a few "intersting"
frequencies to check, easy to do in a spreadsheet for example, using
accurate line equations.
There are lots of other configurations that can do interesting things,
too...you can make a harmonic filter out of series pieces which
alternate impedances above and below the main line's, for example.
It's not very easy to insert a stub in series with a piece of coax,
but I've seen examples of that with balanced line.
Cheers,
Tom
(PDRUNEN) wrote in message
...
Hi Group,
Many thanks to the folks that posted a message for my FM trap question.
I been studying the traps made out of coax for several years, they all
form a
parallel circuit.
Is there any way I could connect the inner conductor and braid such that
I form
a series circuit. I want to pass my signal not keep it in.
de KJ4UO