On Sep 15, 4:09*pm, dave wrote:
Ian Jackson wrote:

And I can confirm that, with a transmitter-side (shack end)
matcher/tuner, and good, low-loss coax, you can do away with the
transformer.

All that does is match the 50 Ohm transmitter to the 50 Ohm (or less)
transmission line. It doesn't address the mismatch between the
transmission line and the antenna at all.

True, but not necessarily a problem. The battlefield is kept out, on
antenna and transmission line. On the radio side of the coax you do
see a function of the remote mismatch both in terms of wrong impedance
and in terms of a partly reflected wave delayed by the travel time
back and forth and with I/V out of phase due to the reactance at the
other end. What an ATU does is
1- compensating the reactive mismatch so that the power source sees a
resistive load
2- rephasing the incoming reflected wave to match the outgoing
3- providing impedance transformation at the transmitter end.
The result is maximized power transfer, and a clean load from the
p.o.v. of the transmitter, w/ minimized apparent reflection. I think 1
and 2 are mathematicaly dual to each other (please someone confirm!).
Not so sure how 3 fits in but it prolly does.

The price you pay for this is a nice mess along the transmission line,
with power being reflected back and forth in an infinite convergent
series of quickly dropping factors until radiated by the antenna, or
lost en route to heat and line radiation, while minimizing the stress
on the transmitter.

Nodes with higher current and voltage along the way also increase
Joule losses - but, again, this may be the dual of expressing the loss
due to reflection, I am not sure if there is a nonlinearity at work
here.

Having everything matched with the minimum use of concentrated L and C
is beneficial (inside the radio, in the ATU, in the transmission line,
and appended to the antenna), but if the transmisson line is very
efficient and not overly stressed (foam or air insulated, decently
sized conductors and insulator, not overheating nor breaking
insulation) a matching network close to the radio side is bearable.
Quite characteristically, as frequencies rise losses mount. Losses
along the same line at a similar mismatch may be negligible at LF,
acceptable at HF, excessive at V-UHF. This is why this is so often
done in professional applications.

In general. an antenna side matching device is better than radio side,
and no matching network is better than with, but perfection is not
always indidpensable. For example, the advantages of radiating in a
certain geometry (e.g. a Yagi, a nonresonant vertical...) or with
greater radiation resistance, may more than compensate for additional
losses due to a necessary concentrated-reactance matching device.