It was a bust. :-(

I used copper pipe -- 1-inch for the bottom quarter-wave element and
1/2-inch for the upper quarter-wave element. The elements are 8 feet, 1 5/8
inches long and at the feed point, the space between the elements is about
three inches.

I put a piece of thickwall PVC inside the lower element as an insulator (per
Cecil's advice) and slid RG-8 through the PVC to the feed point. I attached
the coax shield to the 1-inch pipe and the center conductor to the 1/2-inch
pipe.

This afternoon, a ham pal of mine used his MFJ antenna analyzer (model?) to
test the sleeve. The VSWR was never better than 2.8:1 and exceeded 3:1 over
much of the intended range of operation. He suggested than we pull the coax
out and reconnect it as a more conventional dipole. Good idea. The VSWR
dropped to a 1.3:1 best and didn't exceed 1.5:1 over the entire 10m band.
As long as I keep the coax a few inches from the lower dipole element,
everything's fine. Oh ... don't lean it up against the orange tree,
either.

I think we might have extracted more information with his analyzer had we
not been getting close to being in the dark. If the antenna is inductive,
feeding it through a cap might pull the VSWR down to 1:1 at best. I know
from my J-poles that this is the case -- not sure if it applies to copper
pipe dipoles.

I have very few ideas for correcting the original sleeve design. Feeding one
or both elements through capacitors comes to mind. Resizing would not seem
to be appropriate, since the poor VSWR of the sleeve was still better at 10m
than anywhere else.

73,
"Sal"
(really KD6VKW, anxiously awaiting Saturday morning, to present my CSCE's at
the local VE session.)