ml,

I think the other posters have covered the fact that you just can't get
an order of magnitude or two over all bands just by lengthening the
wire. Two orders of magnitude, or 20dB gain over a dipole requires
really heroic effort in antenna building at HF. If you put a lot of
time, money and effort into it, you might be able to get 20dB gain over
a dipole on 10 meters. You would have to use stacked long-boom beams
or something of that sort.

Even 10dB gain over a dipole is going to require a big beam. Think 5
element yagi or so. So, unfortunately, you're not likely to get a gain
of more than a few dB over a dipole... maybe 5 or 6dB, like Cecil's
antenna on 17m, which is a substantial gain, but falls short of an
order of magnitude.

Cecil's point is valid. I went a bit overboard on saying that the
pattern of a long dipole was BAD. The multiple lobes can be useful if
they point in the directions you want to work, it's true.

It IS bad if there happens to be a null on a station you want to work.
It's good to be able to put the energy you're radiating in the
direction you want.

For a long multiband antenna you can't rotate, this could be hard
because the direction of the peaks and nulls changes with frequency.
This could be good, it could be bad, but without pictures, you just
don't know which way your signal is going. (you can find some pictures
of long doublet patterns here, by the way:
http://www.cebik.com/wire/abd.html.)

- - - - - -

Something to think about regarding massive signal improvement from
changing your antenna: Having a high gain antenna means that you get
response in the direction you want *at the expense* of other
directions.

This is why high gain ham antennas are made rotatable. You can send
all your transmitted power in a narrow *beam* in the direction you want
to work.

For a very long doublet, the lobes may have gain over a dipole, and
that could certainly be useful for some contacts. It could be, though,
that you have a null in the direction you want to work. In the end,
that probably all averages out.

- - - - - -

If the multiple-lobed pattern works for you, just make the antenna 1/2
wavelength long on the lowest frequency of operation. This will give
you a big EFFICIENCY boost on the low frequencies, because less of your
power will go to heat in the tuner.

I wonder, also, how high your antenna is above ground. Putting your
antenna higher will improve the signal on all bands, possibly
dramatically on the lower frequencies if it's not up at least an eighth
to a quarter of a wavelength on the lowest band.

You might just try to build something for a single band where you want
improvement. That's what I did when I was starting out. I had a 150
foot wire that I used on all bands, but then I started building single
band antennas for my favorite bands to improve performance there.

73,
Dan