Chuck Yung
EASA Senior Technical Support Specialist
There are benefits and drawbacks to the use of multiple circuits in a 3-phase winding. Whether discussing a random winding or form coil winding, some of the considerations are shared. Let’s start with the basics: The higher the power rating, and/or the lower the voltage rating, the fewer turns/coil used. Because a 3-phase winding has pole-phase groups alternating ABC, ABC, ABC, etc., the intra-phase jumpers could be 1-4, 1-7, 1-10, 1-13, etc., or any combination of these so long as the alternating polarity of the groups is maintained and the phases are not cross-connected.
Consider a 500 hp (375 kW), 4-pole motor designed for 460V operation.We know that the number of circuits should be divisors of the number of poles, so 1, 2 or 4 circuits could be used. Further, the turns and circuits can be changed proportionally to produce the same magnetic flux densities. The following options are electrically equivalent, but I’m sure you can guess which one the winder will prefer:
| 2 turns/coil |
1-delta using |
44 wires in hand |
| 4 turns |
2-delta |
22 wires in hand |
| 8 turns |
4-delta |
11 wires in hand |
For a random winding, it’s necessary to weigh the advantage of using fewer wires in hand against the disadvantage of the volts/coil stresses increasing in proportion to the number of circuits. Higher volts/coil stresses can be managed by inserting additional phase insulation midway through each coil group. But the random winding cannot manage the voltage stresses as well as the form winding. That’s because the first and last turns in a random-wound coil could make contact. When the calculated volts/coil exceeds 80 volts, the recommendation is to either reduce the circuits and turns, or to add additional “phase” insulation as mentioned.
Increasing circuits, turns
With a form coil winding, the voltage potential between adjacent turns is equal to the volts/coil divided by the turns in that coil. To keep the coil construction practical, the designer must sometimes increase the number of circuits in order to increase the turns.
Unlike a random winding, the space between coil knuckles of a form coil winding can limit the conductor size, and thereby limit the winding design to a higher number of circuits (fewer circuits > fewer turns > larger wire), whereas the preference in a random winding is for fewer circuits (to reduce the volts/coil).
The options can be further complicated by odd grouping. For example, if the motor described above had 54 slots, the fact that the coil grouping is unequal (6 groups of 4 coils and 6 groups of 5 coils) further limits the designer to only two circuits. The use of odd turns, in combination with odd grouping, should also be avoided.
Rotor pull-over during starting
Specific to two-pole designs, the manufacturer will often favor the use of a 2-circuit connection to prevent or minimize rotor pull-over during starting. Redesigning a motor from 2-delta to 1-delta, for example, could result in the rotor striking the stator during an across-the-line start. This is rarely a concern for motors with 4 or more poles, as the shaft is normally stiff enough to prevent such deflection. Shaft size is determined by torque, not the power rating, and for a given horsepower (kilowatt) rating, the lower the RPM, the greater the torque. Therefore, the larger the shaft diameter required.
The more poles a motor has, the more circuits are possible. For larger low-speed machines, it is not uncommon to encounter designs with quite a few circuits. EASA’s Technical Support Department routinely helps members with special connections not found in the Internal Connection Diagrams book. Perusing my connection folder, I have a 72-pole connection, and many connections for 10 through 40 poles. The larger the power rating, the more likely a winding is to have more circuits. Examples would include a 13-wye connection on a 26-pole, 5- or 6- circuit connections for a 30-pole, etc.
The more circuits a winding has, the greater the time required in making the connection – and the greater the chance of error.
There are many reasons a winder might consider changing the circuits: A 4-delta with 11 wires in hand might look attractive to someone wanting to avoid buying 40 more buckets of a particular wire size. Such a decision should always include consideration of the voltage stresses. Not only that, the “legal” number of circuits should always be reviewed.
For example, you cannot double the circuits of a 6-pole motor with an existing 2-delta connection, although many have tried.
Increased risk
When a winding has half as many circuits as poles, there is an increased risk of circulating currents, with increased heating and electrical noise. If the air gap is perfect, it really should not matter if the winder uses an adjacent-pole (1-4 jumpers) [see Figure 1] or skip-pole (1-7 jumpers) [see Figure 2] connection. But if the air gap is not perfect – the use of “extra-long” jumpers connecting pole-phase groups directly opposite one another avoids the issue. The reason for this is simple. It takes more current to drive flux across the airgap than through steel. So a coil group on the side with a larger air gap draws more current than a group on the opposite side of the stator (with a smaller air gap). By connecting those two groups in series, we force them to draw the same current, avoiding circulating currents which would otherwise be present. One manufacturer who was long known for using equalized connections has adopted the EASA recommendation to use the extra-long jumpers, recognizing that it reduces complexity of the connection – especially when a winding is complicated by having an odd number of groups per path.
EASA’s Internal Connection Diagrams book includes the recommended longer jumper connections as summarized in Table 1.
Because we are often rewinding motors that have been repaired before, there is always a question as to how concentric the assembled motor is. Has someone bored and sleeved a bearing housing off-center? Did the customer lift the motor at two diagonally opposite corners, warping the frame? Have the babbitt bearings been rebuilt by a bearing vendor who was unaware of the special high-lift bearing design used in bygone days by a couple of motor manufacturers? What about rotor runout? Any of these situations results in an air gap that is no longer uniform. Depending on the turns/coil, it may be prudent to change a winding design away from the known cases which call for special longer jumpers. A six-pole motor with a 2-delta connection and 10 turns is electrically equivalent to the same motor with 15 turns and a 3-delta. And the conventional jumpers reduce the material needed for the 1-10 jumpers.
The jumper options in Table 1 make for a cumbersome winding, with considerably more bulk than the conventional 1-4 or 1-7 jumpers.
One manufacturer built a large 14-pole motor, then called me when the new motor was extremely noisy. They were skeptical about the 1-22 jumper solution (yes, it solved the noise issue), and it was bulky, but reconnecting it was certainly less costly than rewinding it using a different design. There are frame/end bracket designs where the additional bulk might be impractical – even impossible.
Factors beyond our control
While we can try to make the best decisions about which jumpers to use, there are factors beyond our control which can exacerbate things. A customer facility with unbalanced voltage, for example, might experience higher winding losses and electrical noise that were not present during the final test run. When a motor has half as many circuits as poles, using the special longer jumper solution is prudent. Better to use it “just in case” and be paid for it, than to use it when a disgruntled customer sends a noisy motor back in to your shop.
AVAILABLE IN SPANISH
ANSI/EASA AR100
More information on this topic can be found in ANSI/EASA AR100
Print