Related articles at
Balun Test contains model of "perfect" dipole currents.
Sleeve Balun shows how a sleeve adds impedance, useful for VHF and higher baluns
Receiving Common Mode Noise shows how lack of a balun can contribute to system noise (it applies to transmitting antennas as well)
Longwires, Verticals, and Baluns shows how unbalanced antennas can have similar problems
Balun and Core selection for transformers and baluns
Testing Transmitting Baluns
One of the most popular Ham-lore rumors is a balun's performance can be tested or evaluated by grabbing the coax and watching for an SWR change. This is probably one of the worse test-rumors circulating!
An observable SWR change when touching, grounding, or altering length of a feedline (with the same Zo at the test meter) indicates severe common-mode current problems.
The opposite response, however, means nothing. Lack of SWR change does NOT prove the balun adequate and the system free from common-mode currents.
I was first exposed to this wild idea when asked to evaluate a problem with a commercially manufactured antenna. The antenna's design engineer decided a bead balun was adequate, based on this "grab and touch test". The antenna/balun combination passed the "grab test" with flying colors, even though the system had gross common-mode current levels. Despite extending the bead balun from the original foot to almost three-feet of beads, I couldn't remove common mode excitation of the feedline by the grossly unbalanced antenna.
If we think about it, we would never consider it likely that our hand would have significantly lower impedance than the shield on a long length of coaxial cable! We also would never expect, just by dumb luck, to always grab the high impedance point of a random wire like the feedline.
A hand-grab-test will cause an SWR change only when the impedance of your hand is very low compared to the common-mode impedance of the cable at the point where you are "grabbing" the cable. The portion of impedance attributed to feedline radiation (compared to overall antenna impedance) must be a significant portion of antenna terminal impedance or the SWR will not change.
It is possible, if we alter antenna feed-cable lengths, to observe SWR changes as an indicator of common mode currents from poor system design. But such methods are never nearly as accurate as an absolute current measurement, and are actually only a little bit better than the useless "hand test". They also can give false answers if the feedline impedance is not the same as the SWR measurement device impedance.
An Accurate Test
The most simple and accurate way to test the effectiveness of a balun is to actually measure the common mode currents on the feedline with a suitable RF current meter that does not perturb the system. Such meters are inexpensive to purchase and are also easily manufactured at home. The meter would consist of a closed core that snaps around the cable, a terminated winding on that core, and a meter that measures RF voltage across that winding. The coil's load resistor should be low enough in value that the impedance of the cable shield is not perturbed.
By sliding the meter along the feedline, we could get a good idea of the common mode currents. The meter should be moved along at least 1/4wl of feedline. It normally would be sufficient to sample just four points along that distance.
Feedline radiation, like antenna radiation, is a function of the linear ampere-feet of conductor. The important distance is the spatial or linear distance, not the length of a coiled conductor. It is the charge acceleration over a linear distance of space that accounts for the radiation. A one foot box packed with 500 feet of cable is still just a one-foot antenna!
There are many cases where adding a balun will actually INCREASE common mode currents. Be sure you either fully understand the system or have modeled the feedline and antenna correctly.
A choke or current balun can be tested by inverting the ground on either end, and using a dummy load. Actual operation may produce more stress than this test, or less stress, but such a test gives a reasonable idea of the power rating under reasonable conditions.
The balun should not overheat or change impedance with application of full power for the expected duty cycle in the expected maximum operating temperature (be sure to include sunlight heating).
Using the above test configuration, SWR can be measured. SWR should be flat, and should not change when the ground by the dummy load is moved from side-to-side. This should be true at full sustained operating power.
Testing Voltage Baluns
While voltage baluns are the least desirable types in most applications, they still have some useful applications. Using this configuration:
By moving the tap point along the dummy resistance, an idea of balance can be obtained. The balun should show perfect performance with the tap-point at the load center, and should show an SWR increase as the tap moves to the top or bottom.
Crossed Split Windings on Toroids
Some articles and handbooks show a Split winding method. This method is supposed to reduce winding capacitance by moving the ends of windings further apart. The proposed theory is by reducing shunt capacitance that "leaks RF around the balun", balun performance is enhanced.
This method very often hurts the performance! Check this article