Correction:

Roy Lewallen wrote:
. . .
Connect the input terminals together and the output terminals together.
Connect it to your antenna analyzer as a plain series load -- that is,
connect the balun input terminals to the analyzer center conductor and
the balun output terminals to the analyzer connector shell (or reverse
the two -- it doesn't matter). Measure the impedance at frequencies of
interest. You generally need a minimum of around 500 - 1000 ohms for an
effective balun. The angle of the impedance doesn't matter unless you're
running a lot of power, in which case a low angle (that is, a primarily
resistive impedance) might result in objectionable balun heating.
. . .

This test works only with a current balun. If done with a voltage balun,
the impedance should be very low.

Although there are ways to test for the proper functioning of a voltage
balun, that is, to test for the voltages at the output terminals being
equal and opposite with respect to the "cold" side of the input, they
don't tell you how effective it'll be in stopping common mode currents
on feedlines. In fact, certain imperfections might improve the
performance of a voltage balun in that application. I don't recommend
using a voltage balun for feeding antennas for the reasons I gave in
http://eznec.com/Amateur/Articles/Baluns.pdf.

I apologize for the error. Thanks very much to Owen Duffy for pointing
it out to me.

Roy Lewallen, W7EL