On Thu, 4 Jul 2013 00:33:18 +0100, Channel Jumper
wrote:

You cannot use a tuner with a tuner.

Nope. I've done that for fun. I just happen to have two identical
MFJ tuners available and thought it might be amusing to put them back
to back and measure the losses at the 50 ohm output. One tuner was
set to be capacitive, while the other was matched to have the
conjugate inductive reactance. It worked nicely until I tried 80
meters, where I heard some internal arcing. Measured losses were
fairly high on 40 and 75 meters.

If the matching network is the ladder line and you connect a tuner to it
- yes you can trick the transceiver into believing that is is seeing a
50 ohm matched load - but all you are going to create is heat.

Baloney. The losses come from the limited Q and high resistive losses
of the inductors used in the antenna tuner. That's why really good
antenna tuners use big fat silver plated coils. Try it yourself with
this Java app:
http://www.rsq-info.net/PSK-modelling.html
You'll start to see substantial losses on 80 meters with the default
values. The example uses Q=100 for the inductor, which might be a bit
optimistic for 80 meters. (I haven't done a tuner in 30 years so I
forget the typical Q values). If you plug in real values extracted
from your favorite MFJ antenna tuner, you'll see losses at higher
frequencies.

On the other side of the coin, I hear all the time - I can work
everything that I can hear - with my G5RV - the problem is - what can
you hear?

Unless you have a real 80 meter dipole and you compare them side by side
- within one hour of each other, at the same height and in the same
neighborhood - you cannot compare the two.

In the end - you will realize that the efficiency is so low - you are
not hearing much - just the strongest of signals - when the band is
open, and not much of anything when the bands are no cooperating.

Sigh. In the 1970's, I did some work with diversity reception on HF.
In order for diversity to work, the reception between the two antennas
needed to be different presumably via a different skywave path. The
tests were on WWV at 2.5, 5.0, 10.0, and 15.0Mhz with a simple dipole
and balun tuned to 5.0Mhz. We started with the antennas on opposite
sides of the parking lot. The signal levels tracked each other. I
ran 1000ft of RG-58c/u down the roadway and the signal still tracked.
I ran another 1000ft down the roadway in the opposite direction, and
the signals still tracked. I moved one of the receivers about 10,000
ft away and ran twisted pair audio back to the factory. Finally, with
11,000ft of separation, I was able to see frequency selective fading
at HF frequencies suitable for diversity reception. (Incidentally,
this was adjacent to SJO airport, which added a political layer to
such testing).

The real problem with comparing antennas closely located is that they
interact with each other. Ideally, I would want to see 2-3
wavelengths separation between antennas to prevent interaction. Well,
at 80 meters, that's 500 to 750 ft separation, which is difficult to
achieve.

For added amusement and confusion, there's the commonly ignored
problem of takeoff angle. The usual drawings in the books show a
signal bouncing between the ground and the ionosphere several time
with the angle of incidence equal to the angle of reflection. We'll
it doesn't quite work like that. There was an article in QST last
year demonstrating that the signal comes from directly overhead. While
DX'er try to optimize the takeoff angle to match the equal angles of
incidence and reflection, perhaps it would more interesting to try
maximizing the gain straight up? I'll see if I can find the issue and
article.

How the G5RV fits into the picture is beyond my limited imagination.

The thing that tricks people into thinking that they are doing something
is the fact that they see 100 watts into the meter and they think that
they are modulating all 100 watts - when in fact a single side splatter
signal is only fully modulated part of the time - most of the time - we
aren't really using more then maybe 15 or 20 watts out of 100.

Well, you can set the % modulation to 100% and get 100% modulation.
The problem is that it can easily splatter as you describe. 25% of CW
power is the recommended maximum.

Note that none of this diversion has anything to do with antennas.

Only the digital modes and CW - which is the original digital modes -
dots and dah's - is 100% fully modulated.

Wrong. Percent modulation is the radio of the peak-to-peak voltage at
the waveform peaks, divided into the peak-to-peak voltage in the
modulation troughs, as shown on an oscilloscope. 100% is very common
and easily obtained. Please look at the RF on a scope and see for
yourself.
http://electriciantraining.tpub.com/14193/css/14193_146.htm

That is the reason why we turn down the power when we work digital
modes.

Nope. The reason we turn down the percent modulation is to reduce
splatter, not because the transmitter is somehow inherently unable to
produce 100% modulation.

Most transceivers do not have a 100% duty cycle - hence if you operate
at 100 watts for very long - your transceiver will not take it!

Wrong again. The reason for the low percentage of modulation for most
digital modes is to keep the occupied bandwidth fairly reasonable. As
you approach 100% modulation, the signal starts to become wide and
begins to splatter. Beyond 100%, it's really wide and ugly. Here's
the math for PSK31:
http://www.rsq-info.net/PSK-modelling.html
Compare the occupied bandwidth and spurious junk at 25% modulation
(Fig 3) with the others showing various anomalies.



--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558