Chuck Yung
EASA Senior Technical Support Specialist
When testing DC armatures, whether incoming for repair or after completing a rewind, one question I often hear involves interpreting the surge test (or the high-frequency bar-to-bar test) results. There is a lot to our interpretation of the bar-bar test or surge test. First, if the armature has odd turns, a dead coil, and/or equalizers, each of these will influence how we perform and interpret each test.
Even before testing an armature, the first step is to count the slots and commutator bars and determine whether it might have a dead coil or conjoined bars. In most cases, an even number of armature slots indicates a lap winding, and most lap windings more than 10 hp (7457 W) in size will have equalizers. An odd number of armature slots usually indicates a wave winding, which does not require equalizers.
If we assume a conventional armature winding with an equal number of turns per coil and no equalizers, the bar-bar test should be consistent, within 10 percent of the average value obtained. If the armature has unequal turns, expect a high-low pattern proportional to the odd turn sequence. If it has unequal turns and a dead coil, expect the pattern to change partway around the armature. In most cases, the combination of odd turns and a dead coil results in a high-low pattern halfway around the commutator, suddenly changing to steady values for the other half of the armature. While rare, certain combinations of turns and bars even result in three distinctly different patterns, each for 1/3 of the circumference.
When Armature is Equalized
If the armature is equalized, look at the face of the risers to determine the equalizer pattern (See Figure 1). Equalized bars have an additional conductor at the bottom of the riser slot. In many designs, the number of equalizers is equal to half the number of armature slots, resulting in one equalized coil segment per slot. If an armature has more than this number of equalizers, that may indicate that the designer had trouble achieving satisfactory commutation. (Note that some larger armature designs place the equalizers on the back knuckle of the coil, and a very small percentage have equalizers routed through the core from end-to-end.)
Set up the bar-bar tester so that the two sides always have the same number of bars (and of equalized bars) under comparison (See Figure 2). Otherwise, there will be a pattern that follows the equalizer pattern. If the equalizer pattern is such that there are two equalizers in adjacent bars at one place on the armature, expect a slight anomaly in the bar-bar test at that location.
Remember, the equalizer pitch can be determined by the formula:
Commutator bars / pole pairs
Let’s consider a couple of examples:
First is a lap wound armature with 324 bars and six poles
324 / 3 = 108, the equalizer PITCH is 1-109
Example 2: Lap wound armature with 104 bars and four poles
104 / 2 = 52; the equalizer PITCH is 1-53
If an equalizer pitch is incorrect, the equalizers will draw higher current and quickly fail. If placed into service, during the short time the motor runs, expect the armature circuit to draw higher magnetizing current and suffer from poor commutation. In other words, it is going to arc at the brushes.
Effects of Heavy Varnish
Are the risers covered with varnish, and if so, is the varnish cracked? That is likely to trap carbon, and the surge pattern (or bar-bar tester pattern) could appear to have “static” that might be mistaken for a short. To remove heavy varnish from the risers, use a cutting torch set as if you were going to cut a steel plate. Hold the oxygen lever down, so the flame is oxygen-rich and play the flame across the risers. The oxygenated flame will burn off the varnish without heating the risers. Do this before trying to bar-bar test the armature.
AVAILABLE IN SPANISH
ANSI/EASA AR100
More information on this topic can be found in ANSI/EASA AR100
Print