Fred W4JLE wrote:
I think Roy is referring to a T configuration rather than an upside-down L.
The currents will balance in the T so wire size is limited by physical
considerations rather than electrical. This is just another form of a
capacity hat. The net result is to raise the radiation resistance.

In a tee type antenna, there will be considerable current at the
junction of the horizontal and vertical wires. While it's unlikely that
any wire strong enough to be used won't be able to handle the current
from a heating standpoint, it is possible that using a wire on the small
end of the range might result in noticeable loss. A quick run with a
modeling program would show whether or not that might happen with a
given set of dimensions.

One thing I should mention. If the horizontal portion is higher than
about 0.2 wavelength, MININEC-type ground can be used for modeling
either a T or L. The vertical wire is connected directly to ground, and
ground loss can be inserted at the base as a resistive load. If the
horizontal wire is much less than 0.2 wavelength high, the MININEC-type
ground can still be used with reasonable accuracy only for the T type
antenna. For an L type antenna where the horizontal wire is less than
0.2 wavelength high, a model has to use the High Accuracy ground model,
with the ground system modeled as radial wires just above the ground.

Roy Lewallen, W7EL