That's an excellent summary by Dave. However, I didn't see any mention
of common mode current due to mutual coupling between the feedline and
the antenna. Even if you perfectly choke the common mode current at the
feedpoint, considerable current can be induced onto the feedline. This
won't happen with a symmetrical dipole if the feedline is oriented at
right angles to the antenna. but where the antenna and feedline are
collinear, as they are in a J-Pole or ground plane antenna, coupling can
be substantial. The criterion for maximum current in that case is
whether the decoupled section of the feedline is approximately resonant.
For example, a quarter wave feedline grounded at the bottom and
decoupled at the top (i.e., with a current balun - common mode choke -
at the feedpoint) can have considerable current induced.

It turns out that the conditions for maximum induced current can be
opposite those for conducted current. For example, a half wavelength
feedline that's not grounded at the bottom end won't have a great deal
of conducted current. However, it can have quite a bit of induced
current *if decoupled at the feedpoint*. If there's a good balun at the
feedpoint, the isolated feedline becomes approximately self-resonant.

Here's an example, for those who have EZNEC:
Begin with example model VHFGP.EZ. First, in the main window, select
Units, change to any units other than Wavelengths, and click Ok. Then
select Units again, change back to Wavelengths, and click Ok. (This is
necessary because of an EZNEC bug I discovered while preparing this. It
affects only old files with Wavelength units, which include example
files VHFGP and W8JK. When opened, these files will show the wire
diameter units as being in Inches in the Units selection, but
Wavelengths in the Wires Window. This bug will be fixed in the next
update release, v. 4.0.17. In the meantime, you can re-save the files
after changing the units and changing back, and they'll be ok from then on.)

Add a wire with End 1 at 0, 0, 5 and End 2 at 0, 0, 4.5, 10
segments. Make the diameter 0.25 inch. This represents the outside
of a feedline. Click the Currents button at the left of the main window
and look at the current on the new wire, Wire 6. You should see that
it's about 0.19 amp at End 1 (the top). (You'll see a different value if
you've set a power level in the Options menu.) Now add a decent balun by
adding a load with R = 1000 ohms to End 1 of Wire 6 (0% of the way from
End 1). Click Currents again. Note that the current is now maximum at
the middle of Wire 6, and it's more than twice what it was before --
about 0.42 amp.

This is a situation that I'd frankly never considered before -- where a
feedpoint balun can actually *increase* the common mode current! The
current can, of course, be lowered to a small value by adding a second
current balun (common mode choke) about a quarter wavelength down the
wire (Wire 6, 50% from End 1).

To stop the conducted current, you need a choke at the feedpoint or,
alternatively, an even number of half wavelengths from it. To stop the
induced current, you need a break up a resonant line by adding a choke
about a quarter wavelength or an odd number of quarter wavelengths from
an open end, or an even number of half wavelengths from a grounded end.
Of course, you can alter the feedline length in such a way that both are
minimized without needing a choke.

Roy Lewallen, W7EL