Mike and others,

OK on the elevated loading coil.  I'm not convinced that hard results
support your claims:

Yesterday, I reported that I had increased my mast-head
inductance so that the antennna was almost self-resonant. The new
coil introduced resistive losses and the current reduced, but reports
were better than before.

If I recall, you said that the changed current distribution was an
explanation for the reduced value of current at the feedpoint, which was
incorrect.  Current is dependent on "loop resistance" at resonance.

Last night, Jim, M0BMU measured my field strength compared to a
measurement taken two weeks ago. It was 0.9dB up.

That is very similar to some tests I did some time back, at 181 kHz, with a
temporary top loading.

This does not account for the greatly improved reports,

Quite so.  0.9 dB hardly shifts the S meter.

 but it does
show that with less current and more resistive losses, the only way
the ERP could have increased is if the 'effective height' had
increased.

Yes.  The net result of decreased current because of increased loop
resistance (a top loading coil necessarily needs higher inductance (more
wire) than a base loading coil) and the improved current distribution in the
vertical wire.  The effective height factor is just winning out over current
reduction, by some 0.9 dB.

Field strength is of course directly proportional to current
squared multiplied by effective height (h) squared.

True for radiated power.  Field strength is directly proportional to the
product of current and effective height (amp metres).

At last here is real evidence that the elevated coil really does
increase the 'h' part of the equation. Several of us were sure that it
did, and several have noticed improvements in our signals when
using elevated coils, but the evidence has always been anecdotal.

And still appears to be the case.  The same "gain" should be observed on
transmit and receive (being wary that local noise could be different at each
end of a path).

There is still the matter of adequately housing the elevated loading coil
and sustaining good insulation at high RF voltages.  The weight of a coil
usually means it needs to be supported by a tower or mast.  Antenna
modelling shows that if the vertical feed to a T antenna is near metalwork,
it caused shunt capacitance that pulls down the gain by about 1 dB.  Tuning
(and re-tuning) of the loading coil is somewhat inconvenient when it up in
the air.

One situation not amenable to modelling (with NEC-2) is when there is
clutter around the vertical wire.  It seems that clutter causes
disproportionate losses from "soakage" as the high field strength part of
the wire rises above ground.  Some experimenters have reported significant
improvement by relocating the "upwire" to being in a clearer environment.  A
similar effect would arise with an elevated loading coil, as the potential
on the "upwire" is considerably lower than for a base loading coil
installation.  However, all of these environmental factors should show up in
impedance data and gain measurements.

I have observed the construction technique a number of LF NDB stations, from
visits and photographs.  None have elevated loading coils.  All have clear
sites.

I'm not against experimentation and developing better antennas that fit in
our back yards.

73, Bob ZL2CA