[R-390] Input impedance of R390A
Tom Warren
[email protected]
Sat, 7 Sep 2002 14:19:02 -0400
Ah, in the great spirit of stirring the pot on this reflector and putting
in my $0.02 worth, here goes. (Disclaimer -- I haven't done any RF
engineering for about 37 years and though I have too many EE degrees, I'm
still learning about these 390 beasts and trying to remember things that are
supposed to be at my finger tips. Dr. Jerry, where are you when we need
you????)
1. Bob, I think you're right on about your statement "If your antenna
> does not behave like a matched source who knows what will happen when you
> hook antenna A up to system B.
because there will be some impedance combination looking into the receiver
that results in the lowest noise figure (noise factor) and when looking back
to the generator (antenna or not) should result in the highest SNR. I
remember seeing curves of noise figure (noise factor) versus driving
impedance (generator or antenna) for transistors, but I don't think I've
ever seen the same thing for tubes. Could be that the physics are so simple
and weren't taught when I came along and the bottom line answer is that if
you're running Class A (which an RF stage ought to be), then the optimum
impedance for lowest noise figure is an infinite resistance (when then
corresponds to placing the grid at the top of a parallel tuned tank
circuit).
2. Conjugate matching (i.e., making the driving impedance the conjugate of
the load impedance) is much over touted in my mind. First off, the
guarantee is that half the power is lost in the driving impedance -- not
what you want, as an example, in a power grid or for an audio amplifier
(oh, I can hear the EE gods clammering to straighted me out -- genuinely I
ask that you do so!!). Secondly, conjugate matching into the front end of a
390A, IMHO, doesn't mean squat in terms of best SNR. (So there, EE gods!!!)
and
3. (from a private note I sent Roy this morning) "Actually, I believe the
technique on pp. 134-135
of TM 11-5820-358-35 is a
pretty robust technique of comparing the 390A receiver to a standard where
the technique yields good comparison results to a receiver set up properly
in
a lab environment yet the technique can be applied readily in the field.
What I mean by that is using the 50 ohm to 125 ohm pad gives a pretty good
impedance match to the URM-25D (depending on the input impedance of the
actual receiver at the frequency in question, but the pad the ameliorates
those variations as seen by the signal generator). The receiver is
definitely seeing a 125 ohm driving resistance (note that I didn't say
impedance) looking through the pad back to the signal generator. Using the
16 kHz bandwidth has two implications: 1. these receivers were actually
used in the 16 kHz bandwidth for multi-channel RTTY, so that's one good
reason to measure with that BW, and 2. in the 16 Khz BW mode, measurements
are easier in that with the other filters, at least I fuss with tuning to
recover maximum audio (thus SNR) and since the 16 kHz filter is broad and
very flat compared to the other filters, one doesn't have to fuss with the
tuning much if at all (which is my experience). Thus the technique is
robust (although it gives "bad" numbers like 4 microvolts up to 14 mHz and
then 5 microvolts after that) because it's hard for some young E-2 through
E-5 to mess the technique up, yet it yields a valid comparison to the
measurements made
back in the lab where they really knew what they were doing."
So there, EE gods, let the fun begin.
Dr. Tom, W4PG
----- Original Message -----
From: "Bob Camp" <[email protected]>
To: "Tom Warren" <[email protected]>
Cc: <[email protected]>
Sent: Saturday, September 07, 2002 9:37 AM
Subject: Re: [R-390] Input impedance of R390A
> Hi,
>
> This gets a little weird but here it goes:
>
> Let's say you have two systems set up -
>
> System one - a 50 ohm generator running into a 50 ohm matched radio
> input
>
> System two - a 50 ohm generator running into a high impedance radio
> input
>
>
> In both cases the generator is set to the same *indicated* signal level.
In
> the first case since you are matched half the voltage out of the generator
> is dumped by the divider action of the source and the load. In the second
> case you do not dump half of the voltage. The result is that you have
twice
> as much voltage on the input of the radio in system two.
>
> Since signal generators measure a high level voltage (at the input to a
big
> attenuator) even though the second system has twice the voltage on it the
> signal generator meter reads the same both times. Seems a bit odd. The
> reason it's odd is that most antenna's work the same way. As a matter of
> fact any 50 ohm source works this way provided it's a matched output.
>
> Still with me so far I hope.
>
> Now for the fun - noise in a resistor is a constant power function. The
> voltage goes up as the square root of the resistance. The noise voltage on
> the second system is 3 db higher than the noise voltage on the first
system.
> Since the signal voltage is 6 db higher (twice the voltage) the best case
> signal to noise ratio is 3 db better on the second system. Of course the
> actual signal to noise depends on a lot of things, but the best you can do
> is still 3 db better on system 2.
>
> What does this all mean about radios ? If you have a really hot receiver
> the input impedance may not be anything like 50 or 125 ohms. If your
antenna
> does not behave like a matched source who knows what will happen when you
> hook antenna A up to system B.
>
> End of strange but true tale ....
>
> Enjoy!
>
> Bob Camp
>
> KB8TQ
>
>
>
>
>
> ----- Original Message -----
> From: "Tom Warren" <[email protected]>
> To: <[email protected]>
> Sent: Tuesday, September 03, 2002 9:52 PM
> Subject: [R-390] Input impedance of R390A
>
>
> > Has anyone actually measured the input impedance at the balanced
feedline
> > connector of the R390A? Yes, I know they specify 125 ohms as the
working
> > impedance for the antenna/feedline, but that doesn't mean that the
actual
> > input impedance looking into that port is 125 ohms. I'm pretty sure
they
> > don't specify a conjugate match (then meaning that the input impedance
is
> > 125 ohms). So the question stands, has anyone measured the input
> impedance
> > at a number (even one) frequencies.
> >
> > Curiosity still gets to this cat!!
> >
> > Tom, W4PG
> >
> > _______________________________________________
> > R-390 mailing list
> > [email protected]
> > http://mailman.qth.net/mailman/listinfo/r-390
> >
> >
> >
>
>
> _______________________________________________
> R-390 mailing list
> [email protected]
> http://mailman.qth.net/mailman/listinfo/r-390