Internal motor connection question 6

[ProWinder]

Can any one please explain to me why this motor is connected in this manner.
Thank you

 

[petronila]

Hello ProWinder,

This kind of connection is an special case, named Interspersed winding. This connection helps to solve design problems in two pole machines.
Looking from the design point of view, this connection make possible manipulate winding design variables like chord factor, distribution factor, etc. in order to reduce negative effects of harmonics reducing the distortion of sinusoidal wave form of two pole machines.

In the case interspected connection were ignored, the motor will experiment a high torque reduction.

Another benefit of this special connection is efficiency improving by counter-torque reduction.

Best regards

Carlos

 

[zlatkodo]

Hi, ProWinder,
Carlos is explained the reason for this kind of winding.
In fact, it is an effort to reduce the differential leakage coefficient of winding.
But, only if you know the exact value of this coefficient for both windings (standard and interspersed), you be able to evaluate the benefits. Sometimes these benefits are important but sometimes negligible.
More on non-standard and iregular windings can be found here.

I must commend neatness of your diagram. It is very rare that the service center has so proper documentation.
I hope you will honor us with more of your diagrams for similar, non-standard cases.

 

[electricpete]

I vote LPS to both of the previous responders who are more knowledgeable about windings than me. I am just a student of windings – and I try to take the opportunity to learn when things like this come up on the board. (I’m doing that now)

After reading above responses and reviewing some references, I’d like to try to restate it from my perspective (open to comment).

Two pole motor coils tend to have very long span which causes lots of problems:
difficult to install, uses more copper in endwindings, increases leakage reactance due to long endwindings, makes cooling of endwindings more difficult, harder to accommodate clearances etc.

We can try to cut down on the coil pitch, but then we pay a price in pitch factor (low fundamanetal pitch factor and high harmonic pitch factor and harmonic leakage reactance). For example:

For your 2-pole motor with 54 slots.
Full pitch would be 1-28 (spans 27 teeth)
Here we have gotten all the 15 way down to a coil pitch of 1-16 (spans 15 teeth).
It is 15 / 27 = 56% of full pitch.
So we don’t have such long pitch or endwinding and those problems are gone.

If this pitch 15 was used on traditional winding, the pitch factor would be:
Kp(nu) = SIN(nu*Tau_coil*PI/(2*Tau_fullpitch))
Kp(nu) = SIN(nu*15*PI/(2*27))
Kp(1) =0.766
Kp(5) = -0.940

The pitch factor for fundamental (1) is very low and pitch factor for 5th harmonic is higher than for the fundamental! Not good at all.

So what do they do to make the harmonics lower? I don’t think traditional pitch factor can be used to describe this oddball configuration (interspersed). But I think we can understand the same thing by focusing on the smoothness of transition between phase belts instead of harmonics (because they represent the same thing, an abrupt transition in pattern is a square wave in mmf which is associated with harmonics).


In traditional lap winding, the center coils of pole phase group A will have top coils associated with phase A and bottom coils associated with phase A. The outer coils of that phase group of only top or bottom coils from A phase, while the other coil in the same slots are from phase B on one end of the group and from phase C on the other end. This provides a smooth blending in transition of the time angle of the mmf: it is completely aligned with phase A in the center of the A-phase pole phase group, and it shifts by roughly 60/2 = 30 degrees the outer coils where the A phase is combined with another phase that is 60 degrees away.

What would be the minimum span to implement that lap winding strategy in this motor. Let’s say the top coils of phase A occupy slots 1 through 9 and 28 through 37. We want the bottom half of the A phase coils whose top is at 9 to overlap the A phase coil at position 28 (minimum possible overlap is one slot). The coil span would be 9 to 28 (similar to 1 to 20), encompassing 19 teeth. It’s a longer span that the 16, and it doesn’t do a great job with harmonics because there is only one overlap in the middle. More overlap would be required to improve the harmonics which would cause even longer span.

Assume we don’t want the long span 19+ for reasons discussed above. Then we still have to create the smoothly changing angle as we transition from one phase to the next, but we can’t do it by making the clump of center slots A/A and the outer slots A/B or C/A. That leads to a strategy of interspersed windings where all (almost all*) slots of the pattern have only one coil from phase A, and to accomplish the smooth blend, we include some slots with NO phase A toward the outside of the pole phase group. Since the coils tend to average with their neighbors for all harmonics (except those that are too high order to worry about) , it gives a smoother blend.

Here, you have A phase winding top layer in slots
TopA1: 07, 09, 11, 12, 13, 14, 15, 17, 19
TopA2: 34, 36, 38, 39, 40, 41, 42, 44, 46

With coil pitch 1-15, the bottom coils of above slots will lie in
BottomA1: 22, 24, 26, 27, 28, 29, 30, 32, 34
BottomA2: 49, 51, 53, 54, 01, 02, 03, 04, 06

Combining like-polarity physically-adjacent groups Top2 and Bottom1:, we see the following pattern for this polarity of A phase:

TopA2: 34, 36, 38, 39, 40, 41,42, 44, 46
union..
BottomA1: 22, 24, 26, 27, 28, 29, 30, 32, 34
=…
22, 24, 26, 27, 28, 29, 30, 32, 34, 34, 36, 38, 39, 40, 41,42, 44, 46

Empty spots in the pattern (no A): 23, 25, 31, 33, 35,37, 43, 45
Double Spots in pattern (A top and bottom): 34 (exact middle)

The double spot in the center of the pattern and moreso the empty spots toward the outside of the pattern help to create a sinusoidal pattern with gradual change towards the edges of the pattern, which is what we want to avoid harmonics. The empty spots toward middle of pattern are not particularly helpful but also not particularly harmful considering two things;
1 – the top of a sine wave is relatively flat, so surpressing the peak in the center with these empty spots is not particularly harmful.
2 – the harmonic which is excited by depressing the peak of a sinusoid is the 3rd harmonic, which can be cancelled by wye connection.

Of course above only considered A phase and then we narrowed down to include only the coil sides of like polarity BottomA1 / Top A2. B and C are also present in any empty top/bottom/both positions in the above pattern and they will be 60 degrees different.

To fully analyse this winding, I’d say we cannot use the concept of pitch factor and distribution factors because those concepts have embedded assumptions about the standard winding which are not met. I think we would have to use a brute force spreadsheet which does vectorial sum of the contributions from all slots (including top and bottom), with each contribution’s angle adjusted based on
1 – position
2 – phase and polarity


=====================================
(2B)+(2B)' ?

 

[ProWinder]

Thank you for the information.
electricpete,
I would like to know what school you're attending to learn all this information. I have only got on the job learning (12 years) with zero schooling on theory. I am also apart of a joint apprenticeship committee with my employer. It would be great to know of some more text books to use for training purposes.

 

[electricpete]

From my perspective, the two most useful references for windings that I have are:

1 – EASA Tech Manual – tells you what to do
2 –“Design of Rotating Electrical Machines” (by the Finnish guys: Pyrhonen, Jokinen and Hrabovcova published 2008 by Wiley ISBN: 978-0-470-69516-6) – tells you why.

The 2nd one is very heavy on the math, but also has a lot of accompanying descriptions and figures which help make it slightly more understandable (still a challenge though).

Maybe Carlos or zlatkodo have other suggestions?



=====================================
(2B)+(2B)' ?

 

[electricpete]

I should also say, I took engineering BSEE and MSEE program where I studied a lot of math. I draw on that to try to understand the book #2. Without that background, I think it would be even more of a challenge to learn from that particular book.

I work in a plant and don't spend time in repair shops except when we have a motor there for repair/rewind. No matter how much I study the books, I still learn the most when I spend time with the guys that have the hands on experience in rewinding.



=====================================
(2B)+(2B)' ?

 

[EngRepair]

Hello, ProWinder,
If I'm right, it's about:
- double-layer winding (54 coils)
- lap,
- span 1-16
Please, correct me if I'm wrong.
What is an external wiring to supply? Why you have 6 leads if you have an internal star?
Does it span the same for all coils?
Is the number of turns / coil same for all coils?

 

[ProWinder]

EngRepair,
There are six leads out to be connected in parallel. (It is a 2 star connection) You could have one single lead but it would be a much larger conductor size. Span is the same all coils. The turns are all the same for each coil as well.

 

[zlatkodo]

Why you have 6 leads

L1 to 1-1x,
L2 to 2-2x,
L3 to 3-3x
Note that this winding has another peculiarity.
There are two "wires" between the two circuits in each phase that connect points of equal potential in two parallel paths.
It is often used in large machines in order to reduce the vibrations caused by uneven air gap.
Therefore, this diagram looks quite weird.

To fully analyse this winding, I’d say we cannot use the concept of pitch factor and distribution factors because those concepts have embedded assumptions about the standard winding which are not met. I think we would have to use a brute force spreadsheet which does vectorial sum of the contributions from all slots (including top and bottom), with each contribution’s angle adjusted based on
1 – position
2 – phase and polarity

You are absolutely right. This applies to any non-standard winding.
More on the winding analysis.

 

[electricpete]

That is one stylish website.

=====================================
(2B)+(2B)' ?

 

[Motorwinder]

I've wound two coil interpersed. Never four. Can someone explain the jumpers? Is this to control circulating currents?

Saving this connection for future reference.

 

[electricpete]

zlatkodo said the jumpers were equalizing jumpers to minimize unbalanced magnetic pull

=====================================
(2B)+(2B)' ?

 

[electricpete]

If I start with lead 1, current flows through:
42, 44, 46, (tee), 39, 40, 41, (tee), 34, 36, 38 without any change in polarity.
The tee goes to the opposite circuit connecting points of similar potential for vibratio reasons.
I don’t quite understand why it couldn’t have been:
34, 36, 38 (tee) 39, 40, 41, (tee) 42, 44, 46…. that would seem simpler

=====================================
(2B)+(2B)' ?

 

[electricpete]

I don’t quite understand why it couldn’t have been:
34, 36, 38 (tee) 39, 40, 41, (tee) 42, 44, 46…. that would seem simpler

I guess maybe there are physical considerations which make it easier to fit all those jumpers in this way

=====================================
(2B)+(2B)' ?

 

[zlatkodo]

Hi, electricpete,
Please, see attached file.
One phase is drawn in my way of drawing. Two mentioned connections (per phase) are presented with dashed lines.
Now an internal connection looks quite simple.