why the rotor in a synchronous motor is excited by a dc source ?

A synchronous motor requires a continuous, constant magnetic field in the rotor. This field locks onto the rotating magnetic field produced in the stator by the AC input. In small synchronous motors the field is produced by permanent magnets in the rotor. In larger motors, the rotor field is produced by electromagnets, DC excited.

The power to the motor comes from the AC input which must be multi-phased to produce rotation. The DC excitation just produces the rotor field, but not power output. The DC power is usually a small percentage of the AC power.

If the rotor field were AC excited, the poles of the rotor would flip back and forth north to south to north at the AC frequency and could not lock onto the similarly flipping or rotating stator poles. North would lock onto south and south would lock onto north and since both were filliping at the same rate they would remain locked to each other and the motor would not rotate.

In normal induction motor operation, the stator's rotating field induces a current in the rotor by transformer action. Because of the rotation, the rotor frequency is the slip frequency.

In a way, the motor can be considered to be a parametric convertor device. The pump is driving the stator. The signal is mechanical and at the shaft rotation speed. The idler is at the difference frequency (electrically speaking) current in the rotor. If you drive the rotor at dc, you force the slip frequency to be zero.

This gets me to think about a possible way to control motor speed. I would like to know if such a thing has been tried.

Consider an electronic drive that can vary in frequency. After starting a wound rotor machine, set the drive frequency to zero., You now have a synchronous motor. To change motor speed, change the rotor drive frequency. With a positive phase sequence (the same as the stator), the motor speed should be reduced. With a negative phase sequence, the speed should be increased.

It may be necessary to somehow block the slip current so as not to get normal induction torque. For synchronous operation, there is no steady state slip current.

Bill

multi-phased to

I don't know if anyone has tried to vary motor speed that way or not. It sound like it could work. But, in synchronous machines with DC rotors, the rotors are usually not laminated but sold iron. The winding inductance is usually high because nobody cares since they are usually DC excited. Because of this inductance and the lack of laminations, putting AC on these windings may be problematic, but it could work in principle.

Modern variable speed drives for induction motors do vary the frequency to control the speed. They rectify the line to DC then switch the DC with thyristors or transistors to three phase variable frequency AC to drive the windings. These are induction motors so there is no rotor excitation required. They also adjust the firing angle to control the motor voltage keeping the volt-time product constant.

| In normal induction motor operation, the stator's rotating field induces | a current in the rotor by transformer action. Because of the rotation, | the rotor frequency is the slip frequency. | | In a way, the motor can be considered to be a parametric convertor | device. The pump is driving the stator. The signal is mechanical and at | the shaft rotation speed. The idler is at the difference frequency | (electrically speaking) current in the rotor. If you drive the rotor at | dc, you force the slip frequency to be zero. | | This gets me to think about a possible way to control motor speed. I | would like to know if such a thing has been tried. | | Consider an electronic drive that can vary in frequency. After starting | a wound rotor machine, set the drive frequency to zero., You now have a | synchronous motor. To change motor speed, change the rotor drive | frequency. With a positive phase sequence (the same as the stator), the | motor speed should be reduced. With a negative phase sequence, the speed | should be increased. | | It may be necessary to somehow block the slip current so as not to get | normal induction torque. For synchronous operation, there is no steady | state slip current.

I had suggested the idea a few months back to have a polyphase rotor field rotating in opposition to the field rotation of a polyphase stator as a way to get the rotor to run as double the syncronous speed, e.g. 6000 RPM for 50 Hz or 7200 RPM for 60 Hz. It got some suggestions that it would not work. At least one suggestion said that starting from 0 RPM would be a problem in syncronization (but I felt that if it had enough poles to make a nice uniform field, this might be avoided). Coupling the power into the rotor would be a question. I don't like the idea of slip rings. I was wondering if some way to couple power with a single phase 360 degrees around a magnetic "slip ring" might be doable (with 3 of them, 1 for each power phase). Maybe it might not work from a dead stop, but could work if initial rotation is started.

Why not use a brushless DC motor driven at any speed you wish?

1. No slip rings.
2. Can be run at any reasonable speed.
3. Can use almost any available power input.
4. Can be put in servo loop to control speed to external standard.

Brushless DC motors consist of two coaxial magnetic armatures separated by an air gap. In certain types of motor,

? The external armature, the stator, is fixed. ? The internal armature, the rotor, is mobile (the rotor can also be external in certain cases). The stator is the induced part of the machine. The rotor is the inductor of the machine.

In brushless DC motors, the internal armature, the rotor, is a permanent magnet. This arma-ture is supplied by a constant current (DC). The external armature (stator) is polyphased (3 phases in our case) and is covered by poly-phased currents. The pulsation of these currents is ?. We say that the machine is a synchronous machine because, if ? is the angular speed of the rotor, we have the relation: ?=?/p

In a Brushless DC motor, the rotor is a permanent magnet, this type of motor has almost the same properties and physical laws as a DC current machine.

An electric motor transforms electrical energy into mechanical energy. Two main characteris-tics of a brushless DC motor are: ? It has an electromotive force proportional to its speed ? The stator flux is synchronized with the permanent magnet rotor flux. The back electromotive force (as we will see in this document) is the basis of one the ways of driving brushless DC motors with the ST72141 microcontroller in sensorless mode.

On Fri, 04 Apr 2008 19:47:30 -0700 VWWall wrote: | snipped-for-privacy@ipal.net wrote: | |> I had suggested the idea a few months back to have a polyphase rotor field |> rotating in opposition to the field rotation of a polyphase stator as a |> way to get the rotor to run as double the syncronous speed, e.g. 6000 RPM |> for 50 Hz or 7200 RPM for 60 Hz. It got some suggestions that it would |> not work. At least one suggestion said that starting from 0 RPM would be |> a problem in syncronization (but I felt that if it had enough poles to |> make a nice uniform field, this might be avoided). Coupling the power |> into the rotor would be a question. I don't like the idea of slip rings. |> I was wondering if some way to couple power with a single phase 360 degrees |> around a magnetic "slip ring" might be doable (with 3 of them, 1 for each |> power phase). Maybe it might not work from a dead stop, but could work if |> initial rotation is started. | | Why not use a brushless DC motor driven at any speed you wish? | | 1. No slip rings. | 2. Can be run at any reasonable speed. | 3. Can use almost any available power input. | 4. Can be put in servo loop to control speed to external standard.

That's certainly possible to do. But I was interested in exploring the effects of different configurations. Even if it wouldn't be practical, I was curious if it would work.

| In brushless DC motors, the internal armature, the rotor, is a permanent | magnet. This arma-ture is supplied by a constant current (DC). | The external armature (stator) is polyphased (3 phases in our case) and | is covered by poly-phased currents. The pulsation of these currents is ?. | We say that the machine is a synchronous machine because, if ? is the | angular speed of the rotor, we have the relation:

That's three phase DC, right?

| An electric motor transforms electrical energy into mechanical energy. | Two main characteris-tics of a brushless DC motor are: | ? It has an electromotive force proportional to its speed | ? The stator flux is synchronized with the permanent magnet rotor flux. | The back electromotive force (as we will see in this document) is the | basis of one the ways of driving brushless DC motors with the ST72141 | microcontroller in sensorless mode.

Well, at least it is syncronized to the DC.

|

I would not call that a DC motor. The motor is not getting DC; the controller is. The motor gets modulated pulsed AC. It may well be the standard practice to call this a DC motor, but if so, it is yet another case of terminology misapplied. I'd call it a "pulse control motor". A more clever controller could drive it from an incoming three phase AC supply directly, by switching in the supply phase that has the right voltage and polarity at that instant (in three phase AC power, there's always at least one phase in each polarity at any time). It sure would not be anything DC in that case.

AC is alternating current because the current "allternates" the direction in which it flows. The current to the motor does not do so. The term "phase" does not only apply to one of the rotating vectors of a multiphase AC supply.

Everybody else calls them brushless DC motors, even the little ones in your computer case that drive cooling fans. These use Hall effect switches rather than back EMF for switching and do actually run on DC input. The current in their windings goes through "phases" as it is switched, but it's not AC. Three phase DC motors are used as the platter motors in hard drives, driven by the 12V DC supply. If this terminology is misapplied, it has been so for many years!

What do you do for a living that gives you so much time to think about this sort of thing? :-)

Consider a synchronous generator that is open circuited and the rotor excited with DC. Then there will be a voltage (or voltages for a polyphase machine) induced in the stator at a frequency linearly dependent on speed. If the machine is a 60 Hz machine, the frequency will be 60 Hz at synchronous speed. Now if the machine is run as a motor, it can only run at synchronous speed as at any other speed the generated internal voltage (or back emf) will be at , say 59 Hz, and the supply at 60 Hz, there will be a

1Hz beat frequency with peak torques and currents considerably higher than rated torque and current - considerable damage can result and if the machine is large enough, the effects can cause serious problems in the power system. There will also be 0 average torque so the motor will stall and try to pull up its mounting bolts and go walkabout-.

In theory one could change the speed of a motor by gradually changing the frequency of the rotor- say increasing from 0 to 1 Hz and if done slowly and carefully enough, the motor speed will drop so that the magnetic field of the rotor -as seen from the stator- will still be at synchronous speed as far as the stator is concerned, without actually pulling out of step -i.e. staying away from the maximum torque point.

This is complicated and not worth the effort.

If you want to change the speed- then use a variable frequency drive and, in that case it is easier to use the simpler and more forgiving induction motor. Note that in an induction motor the induced rotor voltages are at slip frequency and the currents will produce a magnetic field that is rotating at slip speed with respect to the rotor-and this field always appears at synchronous speed as seen by the stator.

------------ There have been induction machines used which are doubly excited. Some of these used a form of slip frequency feedback. I believe the Schrage machine was one of these. All suffered from problems of size and relatively poor performance (I recall one machine which was about 3-4 times the size of a normal induction motor of the same rating). Certainly forcing a current at a desired "slip" frequency would do as you wish but blocking the normal slip currents would be a problem and may have been the cause of the poorer performance. Electronic drives supplying the rotor could be an improvement over these machines but if you can do this, you can go with a normal induction motor and use a VFD supply for the stator and avoid a lot of complications.

On Fri, 04 Apr 2008 22:01:32 -0700 VWWall wrote: | snipped-for-privacy@ipal.net wrote: |> On Fri, 04 Apr 2008 19:47:30 -0700 VWWall wrote: |> |> |

|> |> I would not call that a DC motor. The motor is not getting DC; the |> controller is. The motor gets modulated pulsed AC. It may well be |> the standard practice to call this a DC motor, but if so, it is yet |> another case of terminology misapplied. I'd call it a "pulse control |> motor". A more clever controller could drive it from an incoming |> three phase AC supply directly, by switching in the supply phase that |> has the right voltage and polarity at that instant (in three phase AC |> power, there's always at least one phase in each polarity at any time). |> It sure would not be anything DC in that case. | | AC is alternating current because the current "allternates" the | direction in which it flows. The current to the motor does not do so. | The term "phase" does not only apply to one of the rotating vectors of a | multiphase AC supply.

It sure looked like it was alternating from some of the diagrams. Sure, not alternating in the way you get from the wall outlet. But the pulse go one way sometimes, and the other way other times. See figure 3 on page 7. See figure 4 on page 8.

The document showed an example where a simple north-sorth bar magnetic was figuratively used as the rotor. I suppose a pulse sequence only in one polarity could still pull it around in rotation and in sync. But it would operate better if there was a corresponding reverse pulse to pull the other end of the magnet. Additionally, reversing the field of each winding just as the magnet pole passes, so it pushes away afterwards, could help even more.

Also, even though the document describes "PWM" (pulse width modulation), some of the diagrams seem to suggest it is instead using "PDM" (pulse density modulation). PDM is where pulses at a very high rate are used, each at a fixed time interval, but varying in how many are on or off in a window of time, as opposed to a single pulse that varies in width. PDM can be used to more readily vary the average current in the winding over time, where desired. If the PDM rate is high enough, the winding inductance smooths it out. You could make a decent sine-like waveform that way. The advantage of PDM over PWM is that high switching rate allows less reactive component to smooth it out.

I would envision a much more sophisticated system for certain uses like independent automotive vehicle drives on all 4 wheels. The spacing of permanent magnet poles could be different than the spacing of windings so you don't have every pole aligning to every winding at the same time (to reduce vibration). For example, 18 magnet poles and 15 winding poles would allow all the magnets to alternate (even number) polarity, and the interval variation would be replicated three times to maintain a reasonable force balance around the rotor. In each 120 degree segment there would be 6 magnets and 5 windings. The windings would be energized according to which magnet poles are approaching/departing in rotation. For vehicular use, I would definitely want a feedback to confirm where the rotor is, since it is subject to variable load drag. There would be at least 4 windings per segment, 12 for the whole stator, that could be simultaneously energized.

| Everybody else calls them brushless DC motors, even the little ones in | your computer case that drive cooling fans. These use Hall effect | switches rather than back EMF for switching and do actually run on DC | input. The current in their windings goes through "phases" as it is | switched, but it's not AC. Three phase DC motors are used as the | platter motors in hard drives, driven by the 12V DC supply. If this | terminology is misapplied, it has been so for many years!

I guess the whole assembly can be called "DC" since it works on DC being fed to the whole assembly. There's still AC going into the motor windings according to the referenced PDF document.

FYI, this same design seems to be used in a VCR I once took apart to see where the lightning killed it. It was the video head motor, too. It had

12 separate windings.

| What do you do for a living that gives you so much time to think about | this sort of thing? :-)

It's a hobby thing. It's just like trying to figure out what chemicals are needed to go boom. But in this case it's what electricity is needed to go boom :-) Note that I don't actually build things unless I already know what it will (most likely) do. And even then, I find more of the fun in the thought process to design it. Electricity is not the only area I do thought experiments in.

If one has a stator field rotating at 60 Hz and a rotor field not rotating, that should produce a 60 Hz mechanical rotation. So what if the stator field is rotating at 59 Hz and the rotor field is rotating at -1 Hz (the minus to indicating the opposite rotation direction)? Still 60 Hz mechanical rotation?

So how is having the stator field at 60 Hz and the rotor field at -60 Hz different from having the stator field at 119 Hz and the rotor field at

-1 Hz?

One point about this is that the synchronous rotor is typically salient pole design and you can't get a magnetic field to rotate very easily on such a design. If you went to a round-rotor type design, you may as well go to a three-phase winding on the rotor with a third slip-ring.

Otherwise the pulsations of the magnetic field on the rotor from a single-phase winding would result in torque pulsations. With single-phase and very low frequency on the rotor (say, 1 Hz for running near full speed), the torsional vibrations would be severe.

But with a three-phase rotor and stator, I think your idea would work. But as far as variable speed, you would need a VFVV drive to apply the right frequency to the rotor. But if instead you just shorted the slip-rings and applied the VFVV drive to the stator, you'd get the same result. And wouldn't need a wound-rotor and you have the very common VFVV / induction motor combination.

With your design, IIUC, you would apply 60 Hz to the rotor in the same phase-sequence (direction) as the stator for starting and slowly decrease the rotor frequency to accelerate the machine. Once at / near zero Hz, you could reverse the phase-sequence (direction) and then ramp frequency back up to 60 Hz as the rotor accelerates *above* synchronous speed.

Since the rotor impedance will rise with frequency applied, the torque developed would be pretty hard to calculate, but I suspect it would drop off as rotor frequency rises.

daestrom

| One point about this is that the synchronous rotor is typically salient pole | design and you can't get a magnetic field to rotate very easily on such a | design. If you went to a round-rotor type design, you may as well go to a | three-phase winding on the rotor with a third slip-ring. | | Otherwise the pulsations of the magnetic field on the rotor from a | single-phase winding would result in torque pulsations. With single-phase | and very low frequency on the rotor (say, 1 Hz for running near full speed), | the torsional vibrations would be severe.

One of the things I was thinking about for this was a motor with many poles but arranged so they don't all align at the same time. So while one set of poles is in exact positional alignment at one phase point of rotation, the others would not be. No energy would be applied at the poles which are at the point of alignment. But the other poles would have some slight spacing in their position, and the corresponding stator pole would be energized to pull the coming rotor pole, and push the leaving rotor pole. This way it could be designed so that there was always a constant power applied somewhere all the time. The DC or polyphase AC supply would then also see a constant power draw.

| But with a three-phase rotor and stator, I think your idea would work. But | as far as variable speed, you would need a VFVV drive to apply the right | frequency to the rotor. But if instead you just shorted the slip-rings and | applied the VFVV drive to the stator, you'd get the same result. And | wouldn't need a wound-rotor and you have the very common VFVV / induction | motor combination.

My interest in that design with a three phase energized stator and rotor was to see if an uncontrolled syncronous motor could operate at 7200 RPM on 60 Hz three phase power. And if it could, could it properly start up when power is instantly applied to an idle motor. And would it have an efficiency similar to a normal syncronuous motor.

| With your design, IIUC, you would apply 60 Hz to the rotor in the same | phase-sequence (direction) as the stator for starting and slowly decrease | the rotor frequency to accelerate the machine. Once at / near zero Hz, you | could reverse the phase-sequence (direction) and then ramp frequency back up | to 60 Hz as the rotor accelerates *above* synchronous speed.

Obviously this being a controlled motor. But if you have a controller, there wouldn't be any benefit I could see to the added complexity of having windings on the rotor and the coupling to energize them.

I think the terminology is part history, and partly depends on the perspective of the person writing. For example a standard permanent-magnet DC motor with a commutator is universally called a "DC" motor, yet the armature windings are fed with approximately square-wave AC produced by the commutator (with a different phase shift for each coil in the armature). If you build the motor inside-out, with a permanent magnet rotor and the field windings fed polyphase square-wave AC switched according to shaft rotation, what is it? If you include the switching electronics as part of the "motor", then it's still powered by DC and has most of the same characteristics, so it gets called a DC motor. On the other hand, if you separate the electronics from the motor, then the motor is actually a polyphase synchronous motor being driven by a variable frequency generated by feedback from a shaft encoder.

I don't think there's any "correct" perspective, though some are more common than others.

Dave

All of the common electrical machines use alternating current. For induction and synchronous motors, the ac is used without conversion. For dc motors and generators the commutator is use to either rectify ac for dc distribution or invert dc to ac for generating torque.

Bill

The little dry-cell motor you might have made as a Boy Scout did not convert anything to AC. The current in its single coil was reversed in the direction it flowed by its terminals contacting alternate brushes as it turned. It had an alternating magnetic field produced by passing DC through it in an alternating direction. It even produced torque!

With solid state power switches becoming cheaper, I think we'll see many "brushless DC motors" replacing induction or "universal" motors.

| I think the terminology is part history, and partly depends on the | perspective of the person writing. For example a standard | permanent-magnet DC motor with a commutator is universally called a | "DC" motor, yet the armature windings are fed with approximately | square-wave AC produced by the commutator (with a different phase shift | for each coil in the armature). If you build the motor inside-out, | with a permanent magnet rotor and the field windings fed polyphase | square-wave AC switched according to shaft rotation, what is it? If you | include the switching electronics as part of the "motor", then it's | still powered by DC and has most of the same characteristics, so it gets | called a DC motor. On the other hand, if you separate the electronics | from the motor, then the motor is actually a polyphase synchronous motor | being driven by a variable frequency generated by feedback from a shaft | encoder. | | I don't think there's any "correct" perspective, though some are more | common than others.

This makes complete sense.

-------------- Don't worry about constant power- worry about unidirectional torque. What you describe can be done with some versions of stepper motors. Also, this is what you essentially have in a single phase shaded pole induction motor-without the complication of pole switching

------------.

-------------- No. - the rotor field would have to be at synchronous speed with respect to the stator in order to produce anything other than torque pulsations with a zero average and high currents. Note that a conventional synchronous motor doesn't have its field energized and starts as an induction motor- then the field is energized and as long as it is close enough to synchronous speed, it will lock in at synchronous speed. If the speed is not close enough- then it is floor shaking time. Similarly -if the machine was already at synchronous speed and you applied the same frequency AC to the rotor winding you would get the rotor field turning at a speed which was different from that of the stator field and the two fields would act much as they would with a startup from 0 with the field energised with DC. Not good. A very small motor might react quickly enough -for example, a 1/2 inch toy compass needle will sometimes start and accelerate to speed within less than a half cycle . Inertia of a larger mottor will not allow that.

-------------------

-------- EXACTLY!