DC Motors Vs AC induction (single or three phase)

[John HaineParticipant @johnhaine32865]

Posted by Martin Kyte on 03/02/2016 13:52:59:

" the flux in a transformer core is, ideally, zero" ? WHAT

No field = No coupling = No transformer

Martin

Martin, it's true! An ideal transformer has infinite inductance and 100% coupling. If you think of it transmitting power from primary to secondary, look at the currents and their relative sense, the fields they generate cancel out in the core. Practical power transformers get quite close to this. When there is no secondary load the core flux is determined by the primary inductance which is usually pretty big (in a mains transformer), which leads to a rather low quiescent "magnetising current". At high load currents the flux in the core is determined by the leakage inductance of the windings, which measures the extent to which the coupling is non-ideal. The resulting flux makes the core buzz and get warm, and if the current is high enough the core starts to saturate. But you don't have to size the core for the flux generated by the load current but the "leakage" flux.

I think an induction motor is pretty non-ideal as a transformer, but still this means that the core flux probably isn't as high as one might think for the power being generated.

Again going back to the OP's question. My Novamill has a Baldor DC motor with a maximum speed of ~6000 rpm and rated at 375 watts, driven from a KBE speed controller. The motor is really quite small as it is entirely enclosed in the mill column, and seems to be about 7 inches diameter and the same long. On the other hand I removed a half-horse, about the same power, single phase induction motor from my VMB and replaced it with a 3 phase motor with VFD. The single phase induction motor is at least double the size of the Baldor I's say, probably because (a) its single phase so needs more metal and (b) runs at only about quarter the speed.

[AjohnwParticipant @ajohnw51620]

The only problem with windings that have a infinite inductance is that they will also have an infinite resistance to any ac current which would mean driving them with an infinite voltage across them.

John

–

Edited By Ajohnw on 03/02/2016 19:35:58

[Martin KyteParticipant @martinkyte99762]

Back to the transformer question. If you have no field coupling the primary and secondary windings then you don't have a transformer. With an ideal transformer you have no losses not no field.

[AjohnwParticipant @ajohnw51620]

I suspect what John means Martin is no net flux in the core which is some what different to no flux.

The wiki link is poor really. There is no point in trying to explain an ideal transformer as all sorts of anomalies will crop up.

John

–

[Clive FosterParticipant @clivefoster55965]

Back to motors. I found this :- **LINK** rather nice guide to the general characteristics of electric motors. A little out of date, 2007 publication I think but mostly compiled from earlier information, so its weak on brushless motors but it covers the all the standard traditional motors well with nice clear graphs to show whats going on in the torque / power / rpm department. I have all the information in other places but I'm keeping a download of this because it pulls things together nicely in one set of covers.

The companion VFD one :- **LINK** may be of interest too.

Main site has some other efficiency guide that many be of interest too :- **LINK**.

When it comes to the size / power / rpm relationship the calculation tends towards a constant when the underlying technologies are similar. So for a given power a slow motor is larger than a fast one. Probably the major differential is "how hard", for want of a better term, the technology is driven. Drastically oversimplifying old style induction motors running at lower peak magnetic field levels and lower winding temperatures are bigger than modern ones able to run hotter and closer to magnetic saturation. One gain with the old style design is a flatter torque curve. But you pay for it in the electricity bill due to lower all round efficiency. Big ole motors do hang on better under load than modern ones, especially the high efficiency ones which are highly optimised to work well at close to rated power and drop off relatively quickly when slowed by the load. There may also be a physical size element in the slip required to generate power. On a large rotor more physical movement is needed for any given % of slip. If this is the case it would explain why small induction motors in the 1 HP range are relatively weak. Nice plot showing this in the first link above.

Clive.

Edited By Clive Foster on 04/02/2016 10:13:25

[Russell EberhardtParticipant @russelleberhardt48058]

What is important for machine tools is available torque. For a lathe that is available torque at the chuck. I say "available" because when the chuck is spinning but there is no cutting there is very little torque being generated. The torque is produced by the cutting force at the tool and must slow the motor in order for the motor to produce a balancing torque. If it is slowed down too much it will stall and that is what must be avoided.

Power is just the speed x torque product. Torque at the chuck depends on the gearing thus the better performance from machines with a geared head or belt speed change.

Is anyone else annoyed my the comments on Top Gear such as, "This engine is better because it produces lots of torques (sic)"? Again it is the available torque at the wheels that is important and it depends on the gearing as much as the torque produced by the engine.

Russell.

[colin hawesParticipant @colinhawes85982]

A brushed DC motor tends to develop a higher starting torque against a load compared to a synchronous AC motor but has more components likely to eventually cause failure and it is also likely to be noisier. I once owned an old motor that seemed to combine the best characteristics of both types, starting as a brushed motor and changing to synchronous by means of a centrifugal device. I'm not quite sure how it did this but I think the centrifugal device lifted the brushes and shorted the armature windings. An induction motor is usually continuous rated, runs cooler and will go for decades without problems. That makes it very suitable for industrial machines. In my experience speed control of the motor is rarely essential. Colin

[MWParticipant @mw27036]

"Torque at the chuck depends on the gearing thus the better performance from machines with a geared head or belt speed change."

Exactly, but as we know, fully geared machines with a vast range of speeds are very expensive.

Belt drives are inconvenient to change, especially when idyler wheels are involved, you could just stick it on 500 rpm like i do with my pillar drill mostly but a VFD is actractive for its simple design and flexibility combining a mechanical control with a modest selection of either belts or gears with an electronic adjustment, i would argue give you the best of all worlds without being a pain to switch between them.

P.s Colin, i'd fully agree, these AC squirrel cage motors are as tough as old brown boots, i've had 2 DC motors fail me and none of my AC ones have ever died.

Michael W

Edited By Michael Walters on 04/02/2016 14:05:07

[Colin WhittakerParticipant @colinwhittaker20544]

Why are AC motors more powerful than DC motors?

Well a HP is a HP is a HP. But if you want to consider the size needed for a given HP then things are not equal.

The star wars project wanted huge amounts of power for minimal weight and was using airborne DC motors/generators.

Induction motor power is limited by the rate of flux change per second. Once the steel of the stator/rotor is saturated then no more power is possible unless the frequency is increased. That's why airborne electrics work at 400Hz giving a factor of 7 to 8 reduction in weight. Why not increase the frequency still further? Well the choice of 50Hz was to keep the hysterisis losses tolerable with the steels available in the early 20th century. A few years later steels had improved and a 60Hz was possible. In the 1950s steels were available that could support 400Hz and that became the aircraft standard. I have heard of 1000Hz motors but that must around the upper limit.

DC motors use the same copper and steel but rotate independent of the supply frequency. There will still be steel limits to the speed of a DC motor but the commutator will normally have exploded before the steel overheats.

While it might be possible to build a 400Hz induction motor powered hand held drill for those of us stuck on the ground a DC universal motor (or compressed air motor) is the only way to go.

[Russell EberhardtParticipant @russelleberhardt48058]

Posted by Michael Walters on 04/02/2016 14:00:22:

Belt drives are inconvenient to change

That depends entirely on the design of the machine. My 1950s lathe has eight speeds selected by belts as well as a back gear to give a total of 16. Changing speeds with the belts just takes seconds with just one lever to remove the tension and the belts just slip across from one position to another.

However I also have a VFD to give the best of both worlds – particularly useful when facing a large diameter when it helps to be able to wind the speed up as the cut progresses towards the centre.

Russell.

[Neil WyattModerator @neilwyatt]

Slight twist to the tale.

BLDC motors like most electric motors have constant torque up to their rated RPM where it starts to drop off.

EXCEPT that as they are, in effect, a sort of servo-motor a good controller can supply extra power at lower speeds allowing temporary torque boost of 100% or more – at the cost of increased heating.

Neil

[MuzzerParticipant @muzzer]

When you say "DC motors", I think you are meaning to say "commutated" ie the field in the rotor is rotating relative to the rotor itself – and the field on the stator is stationary, so frequently the stator uses permanent magnets.

By the same definition when you say "AC motors", you are probably referring to machines where the field rotates within the stator and the field within the rotor is static (or almost so in the case of the induction, asynchronous machine). This includes wound field machines such as clawpole alternators.

The universal motor is essentially a "DC motor" in this unconventional definition, just to cause confusion. Possibly one reason why normal convention doesn't follow these definitions.

Essentially these 2 basic motor classes simply refer to the same electromagnetic concept but one is the other turned "inside out". The conjecture about maximum speeds due to lamination sizes applying to one but not the other is simply wrong I'm afraid. High speed brushed motors require the rotor laminations to tolerate rapid flux swings without silly losses just as much as the stators of induction or PM motors, whether driven sinusoidally or trapezoidally (BLDC).

The rotor in the induction motor sees significant currents and flux is driven by current, not voltage. The fact that the rotor winding is shorted out doesn't alter this. However, the rotor resistance is generally relatively low, as is the consequent winding voltage.

The original proposition that AC is somehow better than DC is debatable. Depends on the context. As I may have mentioned above, a rotor with a permanent magnet will generally be more compact than one incorporating a winding which gives them a size advantage over induction machines. However, they are more expensive and complex to design and manufacture and will generate a voltage without external excitation. You have to consider what happens if the speed is enough that the back EMF exceeds the supply (battery?) voltage.

Interestingly, many hand power tools are now offering brushless versions for a slight premium. No brushes to replace, no sparks and no commutator to wear out. And more to the point, no significant change in motor size and power.

[AjohnwParticipant @ajohnw51620]

Posted by Neil Wyatt on 10/02/2016 14:26:09:

Slight twist to the tale.

BLDC motors like most electric motors have constant torque up to their rated RPM where it starts to drop off.

EXCEPT that as they are, in effect, a sort of servo-motor a good controller can supply extra power at lower speeds allowing temporary torque boost of 100% or more – at the cost of increased heating.

Neil

The characteristics of a dc motor are down to a back emf being generated. At some speed this balances out the load on the motor and it's internal resistance. If the motor speed is forced to drop due to load the back emf drops as well causing more current to be drawn, speed to increase and the back emf to balance out again. There are various aspects that prevent this from being perfect but it's why dc motors some times get burnt out. Controllers too. It makes electronic control less reliable unless motor current is sensed or the set up fused well enough – that aspect is not at all easy when electronic switching is being protected. The easiest answer as far as the electronics are concerned is to use well under rated components. That way quickish blow fuses can be used for protection. It's pretty easy as far as the motor is concerned but in both cases will get a bit complicated because the time the conditions exist matter as well. Some period of overload will be ok – or should be in practice.

There are more variations. Permanent magnet being one plus series wound, shunt wound and a mix of the two.

Where there is a field winding the best form of electronic speed control involves switch mode driving the armature and the field separately. Usually called sep x. It can offer a much better usable speed range. The same would apply to a brushless motor of the same basic type.

John

–

[Neil WyattModerator @neilwyatt]

Posted by Ajohnw on 10/02/2016 17:27:17:

Posted by Neil Wyatt on 10/02/2016 14:26:09:

Slight twist to the tale.

BLDC motors like most electric motors have constant torque up to their rated RPM where it starts to drop off.

EXCEPT that as they are, in effect, a sort of servo-motor a good controller can supply extra power at lower speeds allowing temporary torque boost of 100% or more – at the cost of increased heating.

Neil

The characteristics of a dc motor are down to a back emf being generated. At some speed this balances out the load on the motor and it's internal resistance. If the motor speed is forced to drop due to load the back emf drops as well causing more current to be drawn, speed to increase and the back emf to balance out again. There are various aspects that prevent this from being perfect but it's why dc motors some times get burnt out. Controllers too. It makes electronic control less reliable unless motor current is sensed or the set up fused well enough – that aspect is not at all easy when electronic switching is being protected. The easiest answer as far as the electronics are concerned is to use well under rated components. That way quickish blow fuses can be used for protection. It's pretty easy as far as the motor is concerned but in both cases will get a bit complicated because the time the conditions exist matter as well. Some period of overload will be ok – or should be in practice.

There are more variations. Permanent magnet being one plus series wound, shunt wound and a mix of the two.

Where there is a field winding the best form of electronic speed control involves switch mode driving the armature and the field separately. Usually called sep x. It can offer a much better usable speed range. The same would apply to a brushless motor of the same basic type.

John

–

Standard rules don't apply with BLDC, because they are effectively servo motors (in fact they aren't much different from being stepper motors with relatively few poles), so they can run at zero rpm (no back EMF at all) and still apply variable torque. It takes clever electronics to apply the extra power while keeping the rpm constant and in a good controller that will include current monitoring and modelling temperature rise in the motor windings.

[MuzzerParticipant @muzzer]

Neil – torque is determined by the current in the winding = flux in the air gap (and the angular error). The same rules prevail – no idea what standard rules you believe don't apply? There's no black art at play here and engineers have long had a good understanding of how they work and how to design them. Ask me how I know.

There aren't many motors that can't achieve variable torque at stall with the right controller. Having a fixed (eg mechanical) commutation angle might be a limitation but with a position sensor and suitable software that can be overcome and nowadays that's trivial.

Really noddy BLDC controllers don't control current, just commutate the windings with a Hall switch, using a dumb (typically six-step) scheme. Better controllers also use a current control loop to regulate the torque and if you want to get clever you can implement vector control. Again, that's very straightforward these days.

[Neil WyattModerator @neilwyatt]

Posted by Muzzer on 10/02/2016 19:57:48:

Neil – torque is determined by the current in the winding = flux in the air gap (and the angular error). The same rules prevail – no idea what standard rules you believe don't apply? There's no black art at play here and engineers have long had a good understanding of how they work and how to design them. Ask me how I know.

The 'black art' is just that – vector control. Generating a rotating vector at a constant speed whilst varying the overall current isn't trivial and it's why GOOD BLDC controllers are very good.

Variable torque at stall is one thing, but try achieving it at 20 rpm using an induction motor or a brushed DC one.

Neil

[AjohnwParticipant @ajohnw51620]

All BL is Neil is a dc motor that replaces the com with electronics. It's as simple as that.

John

–

[Anonymous]

Posted by Neil Wyatt on 10/02/2016 20:27:06:

Variable torque at stall is one thing, but try achieving it at 20 rpm using an induction motor or a brushed DC one.

I don't know about brushed DC motors but I don't see what the difficulty is for induction motors? At least one of the inverter designs I was involved with could produce torque at zero speed. That's important for electric vehicle applications as it is the equivalent of slipping the clutch on a hill start.

Andrew

[AjohnwParticipant @ajohnw51620]

The EV's I worked on Andrew did the same with a DC motor.3 tonne GVW, van, single decker bus / coach and a taxi. They all did a real 50mph under normal road conditions not just on the flat, no clutch no gearbox and not a lot of gearing down to the drive either.

Some were also working on ac drives but it was a tough / inefficient method using the semiconductors that were available in the mid 70's so never went anywhere. I'd be more inclined to say impossible really at these power levels with any sensible levels of efficiency. Transistors 1/2 the size of a house brick also had problems competing with thyristors in a number of respects.

The bloke that initially developed the motor could actually pick it up and carry it around – No one else I know could and I expect he used it as a form of weight training.

The starting torque problems of AC motors that I am aware of mostly relate to single phase drive. Hence machine tools mostly using 3.

And as to brushless I can't really see them competing with what sep x can do if needed as the field on them is fixed.

Some of the claims for all sorts of things are crazy. I can remember reading about room temperature super conductors in Nature donkeys years ago. I just can't understand why they haven't hit the market.

mmmm I just wondered what would happen if the magnets were moved further out mechanically.

John

–

Edited By Ajohnw on 10/02/2016 23:37:47

Edited By Ajohnw on 10/02/2016 23:39:17

[Mike PooleParticipant @mikepoole82104]

I am sitting in a car manufacturing body shop and nearly every squirrel cage induction motor has an inverter of some description. All the hoists, high speed shuttles, turtables, trunnion units and lift tables are fitted with encoders on the motor for speed and position feedback, they all position to a stop and in the case of the hoist applications hold the full load at zero rpm before the brake is applied, the motor is also in full control before the brake is released. Positioning accuracy for these applications is set at a few 10ths of a millimeter although the drives can do better. The inverters for this work are made by SEW eurodrive and are Movidrive B units. The 1100 robots use BLDC motors with resolver feedback, Not a DC motor to be found in the place any more.

Mike

[Neil WyattModerator @neilwyatt]

Posted by Ajohnw on 10/02/2016 22:48:17:

All BL is Neil is a dc motor that replaces the com with electronics. It's as simple as that.

John

–

Is it?

How can you oppose the forces from different windings using a brushed motor?

N.

Edited By Neil Wyatt on 11/02/2016 08:26:18

[Neil WyattModerator @neilwyatt]

Posted by Andrew Johnston on 10/02/2016 22:51:38:

Posted by Neil Wyatt on 10/02/2016 20:27:06:

Variable torque at stall is one thing, but try achieving it at 20 rpm using an induction motor or a brushed DC one.

I don't know about brushed DC motors but I don't see what the difficulty is for induction motors? At least one of the inverter designs I was involved with could produce torque at zero speed. That's important for electric vehicle applications as it is the equivalent of slipping the clutch on a hill start.

Andrew

Read my comment carefully. Variable torque at stall is trivial. Variable torque at constant low speed is not.

[AjohnwParticipant @ajohnw51620]

The level of field excitation plays a huge part in how dc motors behave – including allowing hill hold in the case I mentioned. LOL Downhill hold has more limitations but vehicles generally have brakes but in other case the regen can be varied.

AC inverter drive are constant torque from zero speed with a catch that we all know about however there are various type of inverter rated motors.

John

–

[Martin WParticipant @martinw]

Is it me or have some posts been removed from this thread ???

Martin

[AjohnwParticipant @ajohnw51620]

Posted by Neil Wyatt on 10/02/2016 19:20:20:

Posted by Ajohnw on 10/02/2016 17:27:17:

Posted by Neil Wyatt on 10/02/2016 14:26:09:

Slight twist to the tale.

BLDC motors like most electric motors have constant torque up to their rated RPM where it starts to drop off.

EXCEPT that as they are, in effect, a sort of servo-motor a good controller can supply extra power at lower speeds allowing temporary torque boost of 100% or more – at the cost of increased heating.

Neil

The characteristics of a dc motor are down to a back emf being generated. At some speed this balances out the load on the motor and it's internal resistance. If the motor speed is forced to drop due to load the back emf drops as well causing more current to be drawn, speed to increase and the back emf to balance out again. There are various aspects that prevent this from being perfect but it's why dc motors some times get burnt out. Controllers too. It makes electronic control less reliable unless motor current is sensed or the set up fused well enough – that aspect is not at all easy when electronic switching is being protected. The easiest answer as far as the electronics are concerned is to use well under rated components. That way quickish blow fuses can be used for protection. It's pretty easy as far as the motor is concerned but in both cases will get a bit complicated because the time the conditions exist matter as well. Some period of overload will be ok – or should be in practice.

There are more variations. Permanent magnet being one plus series wound, shunt wound and a mix of the two.

Where there is a field winding the best form of electronic speed control involves switch mode driving the armature and the field separately. Usually called sep x. It can offer a much better usable speed range. The same would apply to a brushless motor of the same basic type.

John

–

Standard rules don't apply with BLDC, because they are effectively servo motors (in fact they aren't much different from being stepper motors with relatively few poles), so they can run at zero rpm (no back EMF at all) and still apply variable torque. It takes clever electronics to apply the extra power while keeping the rpm constant and in a good controller that will include current monitoring and modelling temperature rise in the motor windings.

Wrong Neil. They are driven is a loosely similar fashion to stepping motors. In practice they are perm mag motors where the com has been replaced with electronics. In fact in theory some one could take a perm mag motor and convert it to brushless but they would have to take the drive off the case, a so called out runner.

When they could be found servo motors for some application could be driven from rectified mains. Due to them being so over built for reliable long term operation.

John

–

[Author]

Posts