Cyndi Nyberg
Former EASA Technical Support Specialist

The surge test is used to detect winding faults in AC and DC windings. If there is a turn-to-turn short, the surge test will show that. The surge test is an important step in the initial inspection of a machine, as well as a final test to ensure a proper rewind. 

For this article, we will only focus on three-phase windings. The surge test is typically run by applying a high voltage across each of two phases of a three-phase motor. The decaying resonance patterns of the two phases are superimposed upon one another on an oscilloscope. If the two phases are identical, as they should be, then the patterns will be identical. A perfect match will yield only one apparent pattern (Figure 1a) while a variance or difference, as shown in Figure 1b, represents an apparent problem. Testing continues until all phases have been compared to one another – 1 to 2, 2 to 3 and 3 to 1. 

While turn-to-turn variances are identifiable, not all discrepancies indicate a potential problem. We will cover several examples where an ap­parently bad surge test pattern may not indicate a winding fault.

For three-phase motor windings, a magnetic field is established when the motor is energized. If all the coils in all the groups in each phase are identi­cal, then the magnetic field will be perfectly balanced, and therefore the surge test pattern will show a single trace regardless of which two phases are being compared. 

Lazy lap coils 
Occasionally we will see a winding where the coils are “wound over the span,” resulting in several coils (the number equal to the span) where both sides occupy the bottom of the two slots they span. In addition, the coils laid over the span only occupy the top Figure 1. surge test patterns showing (a) good trace sides of their slots. 

Using this method makes the coils easier to insert, espe­cially on some 2-pole machines with relatively small bores. There is a reason we call this type of winding a “lazy lap”; it’s to convey a negative tone about this method of winding. The main problem with this method is that the coils occupy­ing only the bottom of the slot do not get adequate airflow. It’s certainly not advis­able to insert the coils this way. 

Lazy-lapped coils may also show a bad surge pattern. The reason is because those “span coils” do not occupy the same relative position as the other coils. Proximity to the larger back iron for the bottoms, and proximity of the tops to the open stator bore, result in a different pat­tern on the oscilloscope. If the winder makes the span coils smaller, that may compound the problem, although a surge tester does not reliably detect differences in resistance. 

Part-winding start motors 
To surge test a standard part-winding start motor, you will need to connect leads 1 & 7, 2 & 8 and 3 & 9 in the run configuration. If you surge test the half of the winding with leads 1, 2 and 3 or leads 7, 8, and 9 only, then you will not get a good pattern on two of the three comparisons. Since the A-B-C phase groups are sequential, the “A” to “C” comparison will result in a good pattern, while comparing phase “B” to either “A” or “C” will result in a faulty pattern. The reason is this:  When the surge tester energizes the coils, if B phase is energized, it has an energized group on one side, and an inactive group on the other side. When A and C are energized, both groups are flanked by inactive groups. Depending on the amount of overlay of adjacent phase groups, it can get even more complicated. 

This is a similar effect that is seen when we energize only half the winding to test run – the magnetic field set up in the winding is not even, so the current is very high and unbalanced. 

If the surge test patterns are super­imposed when the motor is hooked up for running, then you know that the winding is good. 

In each phase, all the coils are equally spaced, but when we look at how the coils are spaced in relation to each other, we can see where there will be a difference in the surge patterns. 

Full slot lap windings 
An unusual case that has come up several times recently is that of a full slot lap winding with groups of 1. The way the coils lay in the slots can affect the surge test patterns for two of the comparisons. Figure 2 shows such a winding, which has 36 slots, 12 poles, consequent, with 18 groups of 1 coil and a 1-4 span. The coils appear as 6 sets of three, equally spaced around the bore. 

In each phase, all the coils are equally spaced, but when we look at how the coils are spaced in relation to each other, we can see where there will be a difference in the surge patterns. 

When the surge test is performed on this type of winding, there will be one good pattern and two bad. The good pattern is the one comparing phases A and C. When A to B or B to C phases are compared, the pattern appears to indicate a fault. However, it is only because of the relative postition of the coils that it shows this pattern. 

As with the part-winding-start winding, the pattern depends on whether or not the B phase, sandwiched between A and C phase, is energized. As long as both of the phase-groups being compared are flanked by energized OR de-energized groups, the resulting pattern is good. If either phase under comparison is flanked by an energized phase on only one side, the pattern will be bad. 

2-speed, 2-winding motors 
With 2-speed, 2-winding motors, if either winding has parallel circuits, the choice of jumpers used to make the connection is important. If wound using the incorrect jumpers, there will be circulating currents. When a stator has more than one winding, there is always a transformer effect with the other winding. 

When the wrong jumpers are used for a 2-winding motor with parallel circuits in either winding the incor­rectly connected winding affects the properly connected winding. Let’s assume that both windings have a 2-wye connection, and both windings are connected with the wrong jumpers. If you surge test both windings, neither will show a good pattern on any com­bination of leads. If one is connected with the wrong jumpers, the other winding will show the bad surge pattern. That is, the good winding will have a bad surge pattern, and the winding that has the wrong jumpers will have a good surge pattern. If both are connected properly, then both should show good patterns. 

Unequal grouping 
Some unequal grouping combina­tions do not result in a symmetrical winding. The connection is balanced in that there are the same number of coils in each circuit of the winding, but the placement of the coils does not give a satisfactory surge pattern when phases are compared to one another. As a re­sult the surge test patterns may show some separation. 

An example of this type of winding would be a 6-pole with 24 slots. There are 12 groups of 1 and 6 groups of 2 coils with a sequence of: 211,121, 1 1 2, repeated once, as shown in Figure 3. 

The physical spacing of the groups of coils is not symmetric around the bore, so there will be one good and two bad surge patterns. The 6-pole spacing cannot uniformly overlay the 4 places where there are adjacent groups of two coils. Since the winding is asymmetrical, a slight separation here does not indi­cate a fault. 

There are other cases where unequal grouping will result in bad surge patterns, where it does validly indicate a fault. One case would be when the grouping sequence is right for an adjacent-pole connection but is connected skip-pole, or vice versa. 

There are grouping combinations where the sequence is different depending on the jumpers. In those cases, using the wrong jumpers may result in each phase not having the same number of coils, or if there are multiple circuits, each parallel circuit may not have the same number of coils. This will likely show up as 54 two or three bad surge patterns. 6 Another case where a bad surge pattern indicates a problem is when there is unequal grouping, but a less-than-favor­able grouping se­quence is selected. A specific case for this goes back to early versions of the “EASA Coil Grouping” chart. Take for example of 2 coils. For 1-4 jumpers, the cor­rect sequence is 1 2 1, 2 1 2, 1 2 1, and repeat. If the coils are instead inserted as112,211,221,122,1 
12,211, 2 2 1, 1 2 2, the winding is electrically okay, but the surge pattern will indicate a fault. This again is an issue with the symmetry between the pole-phase groups. 

Concentric windings 
For some concentric winding configurations, there will be similar issues with the placement and symmetry of the groups. One example would be a 4-pole, 36-slot winding with 12 groups of 2 coils. 

Surge testing is a valuable tool to ensure that there are no turn-to-turn, phase-to-phase or coil-to-coil shorts in new or existing windings. Howev­er, knowing what can cause anomalies in the surge test patterns is also a very useful tool, which may prevent an un­necessary rewind! 

Categories: AC motors, DC motors, Testing, Motor testing

Tags: Motor testing Surge testing

Rate this article:

4.3

Print