Dear Mal, LF Group,
At 09:52 20/11/2001 +0000, you wrote:
...What method and where in the system is the RF being measured,
ie top of coil, cold end of coil, at the tx output etc...
It seems to be conventional to measure RF current at the current maximum
point of the antenna which, in the case of virtually all LF antennas tuned
against ground and less than lambda/4 long, will be the point where
downleads are connected to the loading coil. I have my ammeter there, and
this is where my quoted antenna currents have all been measured - the
current is also almost the same at the cold end of the loading coil in most
cases. If there is any impedance matching between TX output and antenna,
one would not generally expect the TX output current to be the same as the
antenna current.
...I have seen rf measured at different points in a system and all the
readings
are different with a thermo couple RF ammeter...
This is to be expected - the current will vary from maximum at some point
to zero at the un-connected ends of the antenna wires.
.....Some systems are not designed for 50 ohms, and the antenna tapped
up and down a PA coil for maximum SMOKE regardless of SWR...
There is nothing magic about 50ohms, it is just convenient for making
cables and things. There is also nothing magic about 1:1 SWR apart from the
transmitter liking it, and transmission line losses being minimised. Most
antennas do not have (50 + j0) ohms impedance, but it is how much power you
get into a given antenna which governs the amount of signal radiated and
lost. This is effectively what measuring the current determines, together
with knowing the radiation resistance and loss resistance of the antenna.
...I have also seen matched 50 ohm systems indicate more RF current output
when detuned from resonance...
No reason why they shouldn't - it has been a common misconception for
several decades that an amplifier designed to deliver its rated output into
50 ohms actually has a source impedance of 50 ohms. This generally is not
even approximately true. For example, Class D designs usually have a very
low source impedance, perhaps a few ohms, and so reducing the load
impedance seen by the PA will increase the output - which is why many
designs have multiple taps on the output transformer to vary output power.
There may also be a significant reactive component in the PA source
impedance, and so de-tuning the antenna from resonance can cancel this, and
lead to increased output. The downside is of course that increasing output
in these ways may well overstress output devices and PSUs, and will indeed
lead to maximum SMOKE!
...It would appear that the thermo couple ammeter must be getting
enhancement current (more heat) from the higher SWR...
Provided it is screened from the effects of RF fields and protected from
other sources of error, an ammeter measures current and nothing else. SWR
is a function of currents and voltages; so an ammeter will not be affected
by it.
...Those using other devices other than thermo-couple rf ammeters
might not see this effect...
Thermocouple ammeters have the advantage of being true RMS responding. But
in this application, where the signal is a good approximation to a sine
wave (you've got real problems if it isn't), any RF current measuring
device that is appropriately calibrated will give the same result. This
includes rectifier-based instruments, and shunt resistors and current
transformers in conjunction with RF voltmeters, power meters, scopes,
analysers etc.
The efficiency of an LF antenna is essentially the ratio of the radiation
resistance to the loss resistance of the antenna - you may also wish to
include the matching components. The radiation resistance Rrad is set by
the geometry of the antenna, primarily described by the "effective height".
Loss resistance Rloss is determined by resistive, dielectric, and magnetic
losses in the materials of the antenna and all it's surroundings. For LF
amateurs, the loss resistance is invariably much larger than the radiation
resistance, and so the antenna current I which can be achieved with
transmitter power P is just I = sqrt(P/Rloss). In my case, P is about
1200W, Rloss on a good day is about 40ohms in total, so I should get about
5.5A, which is about right.
If someone is running the same TX power as you, but getting higher antenna
current, there is no tricks or magic, it just means the Rloss of their
antenna is less than yours. This does not mean that their antenna is more
efficient than yours, since this depends also on what the radiation
resistance of the other ant is. So for example they might be getting double
the antenna current, but have less than 1/4 the radiation resistance, in
which case their ERP would be less than yours, and in the end this is what
counts.
But once you have got the maximum practical effective height for your
antenna, improving efficiency can only be achieved by reducing losses. The
discussions on this reflector show many factors determine losses, and so
experimentation is needed to find the causes of loss and devise solutions.
In the light of mine and other's experience with smaller antennas than the
one at G3KEV, I would be disappointed if I were Mal on finding that my
antenna had a loss resistance of 50ohms. Experience shows that lower losses
than this can be achieved with relatively tiny antennas, and making an
antenna bigger will usually reduce its losses, so there must be some major
source of loss that Mal has not identified. Since Mal does not give a
complete description of his antenna, transmitter power, method of measuring
Rloss, any matching devices apart from the loading coil, and so on, so it
is difficult to make suggestions, but I imagine major improvement must be
possible.
...so you never know really where you are with measurements, they are just
relative...
Everything is relative, as physicists have understood for some time. There
is a maxim in engineering which says "any measurement is better than no
measurement at all"; certainly I find measurements more reliable than tarot
cards, asking a bloke in the pub, or discussions on 80m SSB. No measurement
is ever perfect, but without some sort of measurements, we know nothing.
Cheers, Jim Moritz
73 de M0BMU
|
