1-long jumper 360-deg or 300deg... 2-ôskip groupö? 2

[electricpete]

http://www.uiitraining.com/b51a/100/15104ac_motor_rewind_intro4.htm

SHORT JUMPER: This method is defined as when PPGs are connected finish-to-finish and start-to-start. It is the most common connection method and is used extensively in single-speed motors and in all instances of unequal coil grouping. Connections of each phase are identical. This method interconnects adjacent groups of alternate polarity within each phase. Each jumper spans 180 electrical degrees.

Refer to the figure at right during the following discussion.
Typical short jumper connections for a four-pole motor winding would be as follows:

Begin at the start of A1 (As), go finish-to-finish, start-to-start, finish-to-finish, start-to-start, etc., to A10 (Af). C phase and B phase follow the same way. The connection sequence will be the same, regardless of the number of poles.

LONG JUMPER: This method is used when connections are made finish-to-start and start-to-finish. It is used mainly for connecting consequent pole, single-winding, dual-speed motors. In some cases, it is also used for connecting one or both of the individual windings in two-speed, two-winding motors. Connections for each phase will be identical. This method interconnects alternate PPGs of the same polarity within each phase. Each jumper spans 360 electrical degrees. Alternate polarities are achieved by using a single short jumper as the center connection of each phase.

Typical Short Jumper and Skip Group Connections


Refer to the figure at right during the following discussion.

Typical long jumper connections for a four-pole motor winding would be as follows:

Begin at A1 start (As), finish-to-start, finish-to-finish (short jumper), start-to-finish, to A10 (Af). Connection sequence will be the same regardless of number of poles.

SKIP GROUP: This term refers to the method of starting connections for individual phases. It can be used with either long or short jumper methods. It provides physical displacement between starts of phases to help eliminate potential short circuits

Using my terminology of sequence of groups A B’ C A’ B C’...

I think I understand the difference between 1-4 connection using short jumpers and 1-7 connection using long jumpers (plus one short jumper at the midpoint). It makes sense that the short jumper spans an arc corresponding to 180 electrical degrees (for example left side of A to left side of A’ with groups B’ C groups between). But they say long jumper spans 360 degrees... whereas I think it should be only 300 degrees (for example right side of A to left side of A with B’ C A’ B groups between.

Question 1: Do you agree that “long jumper” spans an arc corresponding to 300 electrical degrees instead of 360 degrees?

Another term they use is “skip group”, which they describe as connecting T leads 120 degrees apart (for example T1-A, T2-C, T3-B within A B’ C A’ B C’) instead of 60 degrees apart (for example (for example T1-A, T5-B’, T3-C within A B’ C A’ B C’) in an integral slot winding. But I thought skip group was also just another way to describe 1-7 interconnection of poles using long jumpers.
Question 2: “Skip group” also refers to any connection that involves 1-7 interconnection of poles, doesn’t it?


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

 

[electricpete]

Also I have drawn attached what I understand the meaning of "adjacent pole" vs "skip pole" connections and "short jumpers" vs "long jumpers". Does it look right?

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

 

[edison123]

Your diagram is right. Other than two speed motors, I also use long jumpers in motor windings having parallel circuits to ensure the current passing through the diametrically opposite pole groups is same and does not create magnetic unbalance.

Muthu

http://www.edison.co.in

 

[electricpete]

Thanks Muthu

Interesting comment about trying ensuring opposite groups have the same current. I’m still thinking about it, but at first glance, it seems backwards.

I was under the impression that parallel circuit are better than series for the very reason that they allow current to redistribute rather than forcing the same current. With parallel circuits, the one near the small gap tends to pull less magnetizing current (lower gap reluctance and higher magnetizing reactance) and the one near the large gap tends tends to pull more magnetizing current.... all of which tends to reduce the flux differential and reduce the unbalanced magnetic pull (ump) in presence of eccentricity. Further I have heard equalizing jumpers can be installed in parallel windings and I thought the purpose would be as much as possible to put 180-opposite coils directly in parallel (rather than series) to allow the redistribution which reduces ump in presence of eccentricity. It is also the reason squirrel cage rotor with all its parallel circuits offers damping which reduces ump in presence of eccentricity... which is why wound rotors are a little more susceptible to ump in presence of eccentricity and also why squirrel cage rotors with skewed bars tend to be more susceptible than non-skewed. We have one family of slow speed, single-circuit wye skewed-rotor motors that has a whole lot of vibration problems including rotor rub with what seems like mild eccentricity.

So again at first glance, I am thinking that connecting 180-opposite poles in series with each other would tend negate the u.m.p-reducing benefit of parallel windings described above. But not positive whether I am missing something here... any comments?


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

 

[edison123]

The logic of long jumpers in parallel circuits (where possible) is about the coil induced mmf. Assuming all coils have the same no. of turns, then ensuring the opposite pole groups carry the same current also ensures the equal mmf (amp-turns) spaced 180 mechanical degrees apart. The uneven air-gap has got nothing to do with it though I agree a parallel winding creates less on ump when compared to a series one.

Muthu

http://www.edison.co.in

 

[electricpete]

The uneven air-gap has got nothing to do with it

I don’t understand why we would care about wiring of 180-opposite coils (series or parallel) unless there is some kind of asymmetry like an air-gap problem. Why is it important if we assume symmetric motor?

So I will assume we are talking about an airgap assymetry. In that case the parallel configuration tends to encourage redistribution of magnetizing current in a non-uniform manner (higher at the location of large gap, less at location of small gap), which tends to even out the flux distribution and reduce the unbalanced magnetic pull. I don’t see any downside of the parallel exept for perhaps a miniscule increase in I^2*R due to unbalance of magnetizing component. For example instead of 10A going through 1 ohm in two parallel paths creating 10^2 + 10^2 = 200 watts loss, we might have 8A in one path and 12A in the other path creating 8^2+12^2 = 208 watts. But it is only the magnetizing component of the current that becomes unbalanced. Is that small increase in I^2*R the concern?


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

 

[edison123]

Instead of power, think of amp-turns (mmf). An 8:12 ratio in mmf will create an ump, whether the air-gap is even or not.

Muthu

http://www.edison.co.in

 

[electricpete]

The only reason I talked about power was to try to figure out why anyone would want to try to intentionally put opposite coils in series rather than parallel, which is what I understood you to say ("to ensure the current passing through the diametrically opposite pole groups is same").

If all we care about is u.m.p., then more parallel paths is preferred, and having the physically-opposite poles in parallel (rather than series) is the ideal case.

Any change in magnetizing current which occurs as a result of eccentricity acts to decrease (not increase) the unbalanced magnetic pull. Again the reason is simple: If everything is in series, we the same magnetizing current all around, the same magnetizing mmf all around, and therefore higher flux at the location of the smallest gap (flux ~ mmf / gap distance). If we put physically-opposite coils in parallel, then the location of the smallest airgap has the lower magnetizing reluctance, higher magnetizing inductance, lower magnetizing current, lower mmf, which when applied to the lower gap tends to even out the flux distribution (flux ~ mmf/gap where minimum mmf occurs at location of minimum gap).

Some references:

http://lib.tkk.fi/Diss/2007/isbn9789512290062/article7.pdf

"Abstract—The eccentric rotor causes an electromagnetic force acting between the rotor and stator of an electrical machine. This force tries to further increase the rotor eccentricity and may severely degrade the performance of the machine causing acoustic noise, vibration, excessive wear of bearing, rotor and stator rubbing and etc. Parallel connections are long known as being a simple and yet effective remedy for the problems associated with rotor eccentricity. In this work, two common types of electrical machine running with eccentric rotor are investigated. Operation in a wide whirling frequency range is considered. The effects of parallel connections in the stator and rotor windings on the eccentricity force are studied numerically and compared to each other. Results of this project reveal that the parallel stator windings can be more effective in mitigating the unbalanced magnetic pull than the rotor cage (or damper winding), which, normally, has many more parallel circuits.

bottom of page 5
In order to assess the influence of the rotor cage on the UMP magnitude, the stator winding was connected in series. The effects of the parallel stator windings were studied when there were no bars in the rotor cage (rotor cage material switched to “air” in the FEA). The stator winding with two parallel paths was assembled in such a way that two neighbouring poles were connected in series and the opposite poles in parallel. According to [8], motors with stator windings assembled in such a way would run more quietly than those with neighboring poles in parallel and opposite poles in series.

http://findarticles.com/p/articles/mi_qa3726/is_199912/ai_n8862160/

end of the page
...winding changes can sometimes be made to decrease the influence of unbalanced magnetic pull. The most common change is to increase the number of parallel circuits in each phase (often not possible with fractional-slot windings

http://lipo.ece.wisc.edu/1994pubs/94-07.PDF

[Bottom of page 5 and top of page 6]
Simulations were then performed for the same conditions as case (A) but with the stator windings connected to have two parallel paths with the adjacent coil groups (poles) in series. The six no-load stator currents are shown in Fig. 6. It is obvious that for each phase the currents in the parallel circuits are distorted and unequal. Figure 7 shows the steady state UMP in both the horizontal and vertical directions. As expected, the average magnetic pull in the direction of minimum air gap is greatly attenuated compared to previous results with a single circuit . It should be noted that computation of the effect of attentuation can be accurately obtained by the coupled circuit model of this paper but is only roughly estimated by the previous steady s t a t e approaches. The mechanism by which the average UMP is reduced in case of adjacent poles connected in series and opposite poles connected in parallel is explained in reference [7]. The inductance is lower in those circuits located where the air gap is larger than nominal , relative to the inductance of circuits located where the air gap is smaller than nominal. The currents in the circuits facing the larger air gap are slightly greater than those facing the smaller air gap resulting in reduction of the eccentricity-induced distortion of the air gap flux density. Correlating this result with Eq. (8) shows that due to currents being unequal in the parallel paths , the term containing [ K x ] is negative, resulting in a reduction of the total value of Fx given by that expression. Similar results are expected in the case where all coil groups are connected in parallel.

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

 

[edison123]

I don't see these citations contradicting my point. And I agree with their point about parallel circuits being better than series ones. I even did a redesign & rewind of large hydro generator (despite objections from the OEM) and proved that a parallel ckt reduced the ump.

Long jumpers connecting opposite poles in series in parallel ckts and (where feasible) is a standard industry practice and EASA recommended one.

Muthu

http://www.edison.co.in

 

[koizumi]

I agree with electricpete that "long jumper" connection spans an arc of 300 degrees.
In practical usage (not to mention in multi-speed windings), in normal single speed winding, I only have encounters with 2 path parallel circuits using long jumper connections. I've never seen any winding with 1-path, 3-path parallel or more using long jumpers.

 

[electricpete]

So, it is an agreement that for reducing u.m.p., parallel circuits are good and equalizing connections even better for reducing u.m.p. The only part I am questioning is the bolded part of the sentence: I also use long jumpers in motor windings having parallel circuits to ensure the current passing through the diametrically opposite pole groups is same and does not create magnetic unbalance.

The objective imo is not to ensure the same current through diametrically opposite pole groups, but to allow redistribution of current.... parallels allows redistribution.... more parallels is better..... equalizing lines even better... and the best possible redistribution is when we connect individual diametrically opposite pole groups directly in parallel via equalizing lines.... which seem to me the opposite of trying to ensure that diametrically opposite groups have the same current.

Attached is an article with example application of equalizing lines that is claimed to be theoretically and experimentally proven. On 2nd page is the description of the 4 pole motor. From the diagrams we can judge that it has 18 slots. q = 3/2.... in a given phase 2 groups have 1 coil and 2 groups have 2 coils. The coils in phase A are 1, 2 / 6 / 10, 11 / 15. Figure 1 shows that it is laid out to provide 1 in parallel with 10, 6 in parallel with 15, 2 in parallel with 11. The difference in coil numbers within a parallel pair is 9 which is one half of 18. The objective of this design was clearly to put diametrically opposite coils directly in parallel to allow them to carry unequal current.

I believe it is a similar principle expressed by Arkio: "The stator winding with two parallel paths was assembled in such a way that two neighboring poles were connected in series and the opposite poles in parallel. According to [8], motors with stator windings assembled in such a way would run more quietly than those with neighboring poles in parallel and opposite poles in series." And also expressed by Lipo "The mechanism by which the average UMP is reduced in case of adjacent poles connected in series and opposite poles connected in parallel is explained in reference [7]."

Those guys don't clarify what they mean by "opposite", but they both refer to the same reference which does clarify. Reference 8 of Arkio and reference 7 of Lipo are both referring the same article. "Effects of Rotor Eccentricity and Parallel Windings on Induction Machine Behavior: A Study Using Finite Element Analysis" by DeBortoli, Salon, Burow, and Slavik. I have a copy but cannot post it since it here. They look at various configurations for a 4-pole motor, including two contrasting parallel/series possibilities. Both consist of parallel combination of two circuits each having two series groups. They are called "opposite in series" and "adjacent in series" according to whether we place opposite or adjacent coils together in a series branch before paralleling. The conclusion is that the adjacent in series works better, which allows opposite coils to exist in parallel branches (not directly in parallel, but not in the same series). Since it is a 4-pole motor, in this case "opposite" also means diametrically opposite.

I am sure there can be many other parallel configurations and equalizing configurations that reduce u.m.p. without placing diametrically opposite coils in parallel. Particularly as the pole number goes up. But my bottom line is that I don't see how it would ever be a design objective (when designing to reduce u.m.p.) to make diametrically opposite coils carry the same current. It may be an attribute of designs that are successful for other reasons that they happen to have diametrically opposite coils in series (to carry same current), but it is not a goal, they succeed for other reason. Given a choice we'd prefer to allow the current in diametrically opposite coils to redistribute their currents unequally to the max extent possible by placing them directly in parallel as in the attached article.

Sorry for being longwinded. I am sure you are well familiar with these principles.




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

 

[zlatkodo]

Hi, Electricpete,
I must admit that I have not analyzed this issue in details, as you are, but I think it would be good if you (or anyone else) can make some short conclusions about this subject, which would be useful for everyone else.
What are the advantages and disadvantages of "long jumper" and "short jumper"?
When to use the first when the other?
What changes in motor performance ,if we change the internal connections from the "short" to "long" or vice versa?
By the way, in this diagram, you are writing here:

Attached is an article with example application of equalizing lines that is claimed to be theoretically and experimentally proven. On 2nd page is the description of the 4 pole motor. From the diagrams we can judge that it has 18 slots. q = 3/2.... in a given phase 2 groups have 1 coil and 2 groups have 2 coils.

it could be 36 slots, 4 poles, single-layer, because , this winding can thus be made, too. It is not for 18 slots ( pitch is to high).

Zlatkodo

 

[electricpete]

it could be 36 slots, 4 poles, single-layer, because , this winding can thus be made, too. It is not for 18 slots ( pitch is to high)

Good point – my terminology was wrong. It has 36 slots, 18 coils, single layer. The coil side numbering includes coil numbers 1 through 16 with two sides (a and b) per coil, so the parallel coils separated by 9 numbers (1/ 10, 6 / 15, and 2 / 11) are still diametrically opposite (same conclusion that the objective was to put diametrically opposite coils in parallel).

I think it would be good if you (or anyone else) can make some short conclusions about this subject, which would be useful for everyone else.What are the advantages and disadvantages of "long jumper" and "short jumper"?When to use the first when the other?What changes in motor performance ,if we change the internal connections from the "short" to "long" or vice versa?

All good questions. The only thing I can add is that I believe the general principle to minimize u.m.p. is that the winding should be arranged to allow redistribution of current (in response to eccentricity) to the max extent possible. If anyone is interested, I can provide theoretical analysis of simple geometry that shows why that is so.

In general I think more parallels help and equalizing lines help and equalizing lines accross 180-opposite coils would be the most efffective type equalizing lines, although not necessarily the easiest.

As far as practical and specific ways to implement that in motors, I don’t know. Maybe others can comment. Muthu mentioned EASA guidelines, I’m not familiar with what EASA says about parallel windings or equalizing jumpers.

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

 

[electricpete]

small clarification:
"equalizing lines accross 180-opposite coils would be the most efffective...."
should have been
"equalizing lines accross diametrically-opposite groups would be the most efffective..."
Because of course we are talking groups, not individual coils.... and the terminology "opposite groups" starts to get confusing when you don't clarify electrically opposite or physically opposite.


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

 

[edison123]

pete, zlatkodo - see attachment. Have a go at it.

Muthu
www.edison.co.in

 

[electricpete]

Thanks Muthu, I will have to study that one.

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

 

[edison123]

Just remember, I didn't write it.

Muthu

http://www.edison.co.in

 

[zlatkodo]

Thanks, Muthu,
It is very useful.
Zlatkodo

 

[electricpete]

I have been scatching my head for awhile thinking about that EASA article that Muthu posted. I think I might know what is going on. Maybe.

First, let me say I think everything I have said is correct in regard to unbalanced magnetic pull created by eccentricity. i.e. the best way to prevent that particular condition would be equalizers that place diametrically opposite poles directly in parallel. That is certainly the conclusion we get from all of the references mentioned 17 Sep 10 22:13, plus quite a few others that I have.

What was focused on above is the fundamental component of the mmf, which results in forces/vibration at time frequencies of 0 (static pull) and 2*LF (twice line frequency).

There is another weirder brand of vibration associated with rotor and stator slotting. The currents in the rotor create an airgap mmf at a time frequency of RBPF +/- 1*LF (RBPF = Rotor Bar Pass Frequency). Typically we see that can interact with the fundamental to create familiar vibration at RBPF +/- 2*LF. (force is proportional to flux squared, so the force fourier transform contains cross products of all different frequencies present in the mmf... and multiplying sinusoid at 1*LF with sinudoid at RBPF +/-1*LF gives sinusoid at RBPF and RBPF +/- 2*LF) . But that RBPF +/- 1*LF component of airgap mmf also has the potential to induce unwanted circulating currents in the stator if there are parallel paths present which would create more possibilities for unwanted slot-related vibration at RBPF+//2*LF vibration and possibly other strange frequencies. That would in general be most severe when the polarities of the voltages induced by the RBPF +/- 1*LF are in-phase when we add them around the parallel loop. That will depend on the spatial distribution of the particular RBPF +/- 1*LF mmf wave time frequency of interest as well as the spatial distribution of the two coil groups in the loop. The RBPF +/- 1*LF can come in a wide variety of spatial orders depending on rotor and stator slotting combination. The most prominent are |R+/-S| and |R+/-S+/-P| (where R is number of rotor bars, S is number of stator slots, and P is number of poles). This can be very easily be an even number. If we have even number >=4 for the spatial order of the mmf and even number >= 4 for the number of poles in the machine, then the associated induced voltages in a loop of diametrically-opposite pole groups would be additive. All of this has nothing to do with eccentricity (which tends to create new mmf spatial harmonics shifted by 1, generally odd) and can occur in a perfectly symmetric airgap. So in summary it may be wise to avoid diametrically opposite poles in parallel because they may tend to increase slot-related noise/vibration (even in perfectly symmetrical stator).

At least that's what I think is going on at the moment. I am open to suggestions or comments.

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

 

[electricpete]

I think I was wrong about the specific spatial order of concern. Assuming a motor with 4 or more poles, then diametrically-opposite poles are in-phase. We would need out-of-phase harmonic voltages induced to create circulating currents. The only spatial order mmf harmonic that would do that is 2. There should not be large mmf at spatial order of 2 since slot combinatinos are avoided that match |R+/-S|=2 or |R-S+/-P|=2. Let me think about that some more.

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