MGFoster wrote:

Thanks to everyone who gave advice; very helpful.

The person who stated that "high ohmic connections" were deleterious to
QRP ops was describing his construction of a linear loaded dipole and
the connections to the inductance section of the antenna (see N5ESE's
article on the "Notebook antenna"
http://www.io.com/~n5fc/notebk_ant.htm). This is a small radiation
resistance antenna and could be affected by bad connections.
. . .

Yes, it does have a low radiation resistance:

21 MHz: 31.7 ohms
14 MHz: 11.6 ohms
7 MHz: 2.6 ohms
3.5 MHz: 0.64 ohms
1.8 MHz: 0.17 ohms

And the tuner has a lot of reactance to take care of, from about 180
ohms at 21 MHz to nearly 8000 at 1.8 MHz. If a 1.8 MHz tuner had a coil
with Q = 400, the tuner loss alone would be 20 dB.

My model doesn't show the claimed spectacular rise in radiation
resistance from the meander lines. Replacing them with straight 14 inch
wires just about halves the radiation resistance. That's nearly 3 dB at
the lower frequencies, so nothing to sniff at, but nowhere near the
factor of up to 14 claimed. The meander lines cut the reactance by a
factor of 1-1/2 to 2, which helps the tuner efficiency.

The fact that QSOs can be made with this antenna and QRP is yet another
illustration of just how little radiated power is needed in order to
communicate.

Roy Lewallen, W7EL