Induction motor winding questions

Paul E. Schoen wrote: (snip)

Wi9thout commenting on your specific experiments, I have some generalities for you. You are altering the number of poles on the stator without regard to the number of poles on the rotor. Rotoe poles are not fixed on the rotor, but slide along the circumference of the rotor as the motor drives a torque load. The rotor poles are a consequendce of currents circulating in the squirrel cage conductor pattern cast into the slots in the rotor laminations. The motor needs several rotor shorting bars per pole to allow the rotor magnitization to slip smoothly along the surface of the rotor, while maintainimng an essentially stable field to produce torque. I think you are trying to magnetize and react against areas on the rotor that are too small relative to the spacing of the rotor shorting bars.

A better solution might be to start with a magnetic design that is optimized for a large number of poles. Hybrid stepper motors are essentially high pole, permanent magnet multi phase (usually 2) synchronous AC motors, and are often low voltage designs, to start with. Since their rotors are permanently magnetized, they don't deal with rotor shorting bars. 50 poles is a very common number.

Otherwise, I suggest you go with a straight low voltage rewind of an induction motor (without changing the number of poles) and a gear box to lower the motor speed.

Take a look at what folks are doing to CD ROM motors for 3D flying. It generally involves regluing magnets, and rewinding. Here is a link: (in german)

You can also look around on

for information on this. Some of it is even in english... ;)

--
Regards,
  Bob Monsen

If a little knowledge is dangerous, where is the man who has
so much as to be out of danger?
                                  Thomas Henry Huxley, 1877

wire,

and

total

Thus

seems

source

number

for

and

phase,

in

the

31

rather

It

were

wire

B,

as

knowledge

motor,

essentially

exist

recommendations

to

RPM

phase

about

pole

why

I think you need to check the inductance of each winding as you finish it to check for shorted turns, also as someone pointed out you need to consider the rotor too, I think the problem might be not enough slots per pole in the rotor, or even in the stator with the non sinewave supply, maybe you might consider a hysteresis type rotor ? it has a flat torque/speed charecterstic.

I dont think you can have only positive poles on the stator unless i misunderstand how youve wound it, the negative ones will be there too, deep bar rotors are used for higher torque at lower speeds.

Colin =^.^=

For best efficiency, power rating and size its very hard to beat a geared 2 or 4 pole design. Your rotor and stator are already optimised for this solution and it's a straightforward rewind for your new operating voltage.

It's a real pain trying to thread heavy wire through stampings intended for use with relatively thin wire. It's much easier to use multiple parallel strands of wire of roughly the same thickness as the original high voltage wind.

Jim

I appreciate the comments and they raise valid points. However, the suggestions to use reduction gearing miss the objective, and I want to use an induction design rather than a PM rotor. What I am trying to accomplish is to make an inexpensive, rugged, and simple motor with maximum horsepower for a given frame size. I would like to use a standard frame for simplicity of mounting and coupling to sprockets and chains or other means to obtain the desired speed range. I want to make the basic motor with the slowest speed possible (at 60 Hz), which means maximizing the number of poles. I will then use PWM techniques to raise the frequency to a multiple of 60 Hz, to a maximum determined by the highest safe speed of the motor, as well as any limitations of the magnetic material. So far, the motors I have taken apart have 24 or 36 slots, and that appears to limit the number of poles to

8 or 12, but my experiments and analysis so far indicate that there may be problems with those numbers, so 4 or 6 poles may be a more reasonable limit.

I have tried to come up with various winding schemes, but it appears that when poles are closer together, the windings interfere with each other due to overlap and cancellation of effective number of turns. I'm not sure I am doing this correctly, but for each pole piece, I do a vector addition of the currents in the windings of the adjacent slots. Ideally, if all the windings (assume 4 in one slot and 4 out the other slot) are the same polarity (A), the total effective turns would be 8A. However, if the slots have overlapped windings from another polarity, the total effective turns will be less. The vector addition shows that crowding poles so close that adjacent pole pieces have a polarity 60 degrees apart (which creates a 12 pole motor with 36 slots), the effective winding is reduced to 1/4 of the ideal. I have not done all of the calculations, but the 2 pole and 4 pole designs should be close to ideal, but the 6 pole is probably reduced to about 50%, and it drops to 25% for 12 pole. This closely correlates to what I have seen on the size and weight of motors with slower speeds at the same HP. However, for some very large motors, this effect is much less, which indicates that they must have many more poles and will not have the same problems with overlap.

My conclusion is that I should probably use no more than 6 poles for a 36 slot stator. This will probably give me only a 50 % increase in HP when I run it at 3x (180 Hz), and perhaps three times rated HP if I dare run it at

7200 RPM. I may be able to find some commercially available motors with 48 or even 72 slots in the lower HP types I am looking at (up to 5 HP).

If any of you have any old motors you may wish to give (or sell) to me, it may help with my experimentation. I will of course pay for any reasonable shipping and handling. You could first remove the copper windings and sell them for scrap and reduce the weight. I have some new motors I will be stripping and rewinding, but I'd rather play with an old junker.

BTW, one of these motors is a three phase 1 HP unit which runs at 850 RPM, which implies 8 poles. However, it has a 36 slot stator, so I assume it is wound with considerable overlap and a 50% reduction of the torque it would have with 2 or 4 poles. Interestingly, it is the same size and weight as a 2 HP motor rated 3450 RPM.

< Snip - details of experimental motors>

recommendations

to

RPM

Well, for Good Grief's sakes, haven't they been working on that already for almost a hundred years? From the rest of your explanation, it sounds like you're trying to determine by experiment, stuff that somebody else already has the numbers on. What a waste of time!

Or are you trying to, in your garage, come up with the magical design that will blow the experienced motor makers out of the water? ?:-|

Good Luck! Rich

What I am trying to accomplish is a low cost and higher power version of the tiny but powerful 400 Hz aviation motors that I found on a website

I also want the motors to be powered directly (through a PWM inverter) from fairly low voltage battery supplies of 12-72 VDC or so, and see about what frequency limits may show up as reduced performance due to magnetic losses. The inexpensive implementations of the concept of Volts/Hz motor control are fairly recent, and as far as I know have been applied only to standard 240/480 VAC type motors. I have not seen anything about increasing HP by overdriving the frequency above 60 Hz, mostly because the DC bus is generated from the same original AC line voltage as the motor rating, and also because it can be unsafe to drive a motor much over its nameplate RPM and voltage.

I like the simplicity and ruggedness of AC induction motors (all hail Tesla!), as well as their well-documented speed and torque control. I am looking forward to making a motor that I can hold in my hand, connect to a

12 VDC battery, control speed with a pot or keypad, produce 1 HP, and cost under $100. If nothing else, I will learn a lot and be able to build a hell of an electric drill!

Paul

use

accomplish

horsepower

horsepower

simplicity

Hz,

to

limit.

am

the

windings

overlapped

The

pieces

the

they

overlap.

at

2

I know some washing machine motors used induction motors for quietness and they were rather large (compared to unversal ac/dc type), also they were dual wound for low speed high torque/high speed low torque. im not sure of the power in each mode.

so theres 36 slots in the stator but how many slots are there in the rotor of these motors ?

Usualy with 3 phase overlapping windings, when one winding is peak energised in the positive direction the other two phases wil be energised in the oposite direction becuase the sum of the all the 3 phase currents is 0.

energised in the negative direction thus in the center of the first pole the other two windings actualy add rather than subtract to the field, maybe you need to study some winding schemes, its been a while since I messed about, I remember re winding the stator of a large dc motor and it never seemd as powerfull again, I was convinced i rewound it exactly the same, its probably easy to miss something and be 1 slot away from a much more powerfull motor.

Maybe it would be helpfull to look at a book on the subject, I found a book called 'electric machines' very usefull, cant remember the author.

let me see if I can draw it in ascii ...

____________________ \\ / Ia-> \\ / X X / \\____________________/ \\ / \\ | pole A | \\ ____________________ / \\ / Ib-> \\ / X pole a+b X / \\____________________/ \\ / \\ / X pole b+c X / \\____________________/ \\ / \\ / X pole c+a X / \\____________________/ \\ / \\ / X pole a+b X / \\____________________/ \\ / \\ /

... not sure if this is corect as its a bit hard to draw/folow ...

Ia+Ib+Ic always = 0, so when Ib is at its positive peak :- Ia and Ic will be

0.5 negative ... therefore if you look at pole B it is imediatly surounded by a clockwise direction curent from winding b plus clockwise current from half of winding a and c b + a/2 + c/2 = 1.5

If you look at pole c+a then it is surounded by an anticlockwise curent of a/2 + c/2 + b = 1.5 here is where you wil find the oposite pole to b.

when there are many more slots the windings are just spread out, the overlap ocurs gradualy, but at the pole centers the same rules should apply, any 1 pole should never be completly overlaped by adjacent pole windings at its center.

hope this helps

Colin =^.^=