3 PHASE INDUCTION MOTOR DESIGNING FOR LOW FREQUENCY OPERATION

[Spehro Pefhany]

Thanks to all of you who have replied.

My application is:
--airgap larger than 6mm
--torque 1 Nm
--speed range 20 to 80rpm
--low current
--220V supply
--also have dimension constraints

I have been successful in designing this motor for 50Hz operation with
four pole stator( synchronous speed = 1500 rpm). It was satisfying all
the parameters. Now when I tried to control the motor speed by
applying V/f control the results were horrible. Current rose to 16
amperes( four times of orignal value), maximum torque achieved was
very lower than the required one and also there was saturation
problem.

Why did the current *rise* over the original value? Normally the
current will be LESS because the winding resistance. In fact, AC motor
speed control designers put in a boost factor that increases the
voltage at low RPM to compensate for that (at least that's what I have
done).

Best regards,

Best regards,
Spehro Pefhany

 

[Tim Wescott]

I think that you are just rediscovering why motors ar not built to operate
at such low speeds.

Bill

There are DC motors that operate at those speeds -- I work with a system
that has one. In general, though, with either a DC motor or an
induction motor with a reasonable number of poles you must get very good
shaft feedback and control the motor torque, and you need to view the
motor as a torque source and not a power source. You'll end up
dissipating all your "torque generation energy" as the same amount of
power as if the motor were operating full speed (which is why other
responders have mentioned cooling).

Was your V/F drive perhaps changing the frequency but not the drive
voltage, or did it have a minimum drive voltage below which it could not
go? Certainly trying to control your motor torque by giving it a
too-large voltage and controlling torque by lowering the slip would heat
things up in a hurry.

 

[The other John Smith]

Thanks to all of you who have replied.

My application is:
--airgap larger than 6mm
--torque 1 Nm
--speed range 20 to 80rpm
--low current
--220V supply
--also have dimension constraints

I have been successful in designing this motor for 50Hz operation with
four pole stator( synchronous speed = 1500 rpm). It was satisfying all
the parameters. Now when I tried to control the motor speed by
applying V/f control the results were horrible. Current rose to 16
amperes( four times of orignal value), maximum torque achieved was
very lower than the required one and also there was saturation
problem.

I have tried every thing in my knowledge but to no avail.
I have not tried increasing the number of poles and vector control
yet.

What are your kind auggestions.

regards,
Nasir.

Nasir -

A motor wants a _constant_ Volts-per-Hertz. If your four-pole motor operated
properly at 50 Hertz, it was rotating at an unloaded (approximately
synchronous) speed of 120*50/4 or 1500 RPM. To get down to 20 RPM, you must
reduce the frequency to 50*20/1500 or 0.66667 Hertz. The voltage must also
be reduced by the same ratio, approximately). The reason I say approximately
is that a voltage boost could be applied to overcome Voltage drop in the
windings at low speed and high load.

If the voltage is not reduced along with the frequency, you will saturate
the iron. You know what happens when you saturate a transformer. A motor is
really quite similar to a transformer except you are extracting mechanical
energy rather than electrical energy. If a transformer is designed for 100 V
at 50 Hz, you can apply 200 V at 100 Hz with no problem assuming there is
sufficient insulation to support the higher voltage. This is true for motors
as well.

I have operated a 4-pole, 40 hp, 230 V motor as an 80 hp, 460 V motor. Of
course the 80 hp is available only at about 3600 RPM rather than the
name-plated 1800 RPM.

You can go the other way, too, and that is what VFDs do for you. They reduce
the frequency and voltage while maintaining a constant Volts/Hertz ratio.

If you were controlling your motor with a VFD, try turning off the voltage
boost. Be sure to set the VFD for your design V/F. Pay very close attention
to the VFD setup instructions such as base speed and number of poles. There
are a number of setup parameters which will affect your V/F.

Good luck.

John

 

[Don Kelly]

Thanks to all of you who have replied.

My application is:
--airgap larger than 6mm
--torque 1 Nm
--speed range 20 to 80rpm
--low current
--220V supply
--also have dimension constraints

I have been successful in designing this motor for 50Hz operation with
four pole stator( synchronous speed = 1500 rpm). It was satisfying all
the parameters. Now when I tried to control the motor speed by
applying V/f control the results were horrible. Current rose to 16
amperes( four times of orignal value), maximum torque achieved was
very lower than the required one and also there was saturation
problem.

I have tried every thing in my knowledge but to no avail.
I have not tried increasing the number of poles and vector control
yet.

What are your kind auggestions.

regards,
Nasir.

-----------
At 220V, 50Hz, 4 poles you have a motor which has a synchronous speed of
1500rpm. You are trying to run this at 20 to 50 rpm with V/f control. The
impedance of the motor at the lower frequency will be low.You have
saturation -probably because you can't reduce the voltage in line with the
frequency and there is a resultant flux increase- both this and the high
current seem to imply a higher voltage than you should have at low frequency
and is likely a result of saturation. You have indicated nothing about the
basic R/X ratio of the rotor and that will have an effect. The idea of
designing a machine is to design one for the speed range for which you most
commonly use it. That is -design on the basis of a synchronous speed of the
order of 50-75rpm- with what ever frequency you have (50Hz is a bit extreme
for this-can you get down a bit more?). The resultant machine will be big
and bulky because of poles/iron requirements. It can be done but there are
better options available at the torque and speed range that you want-
seriously look at a brushless DC type of machine-as used in some turntables
for records.

 

[Ed]

Thanks to all of you who have replied.

My application is:
--airgap larger than 6mm
--torque 1 Nm
--speed range 20 to 80rpm
--low current
--220V supply
--also have dimension constraints

I have been successful in designing this motor for 50Hz operation with
four pole stator( synchronous speed = 1500 rpm). It was satisfying all
the parameters. Now when I tried to control the motor speed by
applying V/f control the results were horrible. Current rose to 16
amperes( four times of orignal value), maximum torque achieved was
very lower than the required one and also there was saturation
problem.

I have tried every thing in my knowledge but to no avail.
I have not tried increasing the number of poles and vector control
yet.

What are your kind auggestions.

regards,
Nasir.

For the same torque output, the current should be the same at both
frequencies if the volt/hz ratio is kept constant.

You mention saturation at the lower frequency. Can you tell us what
the RMS voltage was at the motor for the 50 Hz test comparted to the
lower frequency? Can you vary the frequency between these two points?

It seems that your ratio between the voltage and frequency is changing
as the frequency is decreased. Is the frequency generated by a design
of yours or is it a commercial product that you have purchased? The
odd part of this would be the fact that you do not have the torque
requirement at the lower frequency even with the saturation.

I am more familiar with the motor side of this and look at the drive
as a black box that provides the motor with what is required. If the
motor will perform at 50 Hz, it should be capable of what you are
wanting at the lower freqency providing it is provided with the proper
inputs.

Good Luck,

Ed

 

[Rob Paisley]

I am working on the designing of a 3 phase induction motor for
operation at 20 rpmm. But i am not finding the exact
relationships/foumulas which hold true at low frequencies(i.e below 5
Hz). Could anyone guide me about a "booK" which deals with the abouve
mentioned subject or can give me some guidelines which should be
observed while designing at low frequencies.

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

If you want to apply a low frequency sine wave to a typical
induction motor, first try applying it to an iron core transformer and
observe the output versus input waveforms on an oscilloscope.

The results may affect your design.

Rob.

 

[Rich Grise]

[email protected] (NASIR) wrote in message


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

If you want to apply a low frequency sine wave to a typical
induction motor, first try applying it to an iron core transformer and
observe the output versus input waveforms on an oscilloscope.

The results may affect your design.

Rob.

At one place I worked, they had a 5 HP 3-phase motor and a VFD. When
you cranked up the VFD, you could hear its oscillators whining. It
was very weird, because the pitch would increase, and from the timbre
it sounded like a variable-width pulse train. And as you continued to
increase the setting, the whine would reset in freq. to the low end,
and increase again, but with a little different timbre, like the
pulses were wider or something.

Kinda like a dragster going through the gears - same freq. band,
different timbre.

So I don't think you want to drive your motor with anything like
a 2Hz sine wave!

Presumably, they have a 3-phase oscillator that's locked to some
harmonic of the natural rotor frequency, and presumably pulsed
or PWM, but the fundamental freq. sounded like it was in the 400Hz
~ 1KHz range.

Hope This Helps!

 

[Bill Sloman]

I am working on the designing of a 3 phase induction motor for
operation at 20 rpmm. But i am not finding the exact
relationships/foumulas which hold true at low frequencies(i.e below 5
Hz). Could anyone guide me about a "booK" which deals with the abouve
mentioned subject or can give me some guidelines which should be
observed while designing at low frequencies.

Considering the number of contributions posted to this thread, it is
surprising that nobody had explicitly made the point that motor torque
depends on the current through the coils, and the voltage applied
across the coils determines the torque only as far as it determines
the current.

IIRR, induction motors are not suited for very low speed operation,
because they rely on the current induced in the rotor by the driving
currents in the static coils, and with very low drive frequencies, you
can't put enough current through the static coils to generate enough
dI/dt at the rotor to generate enough current in the rotor.

Synchronous motors don't have this problem. which is why stepper
motors - which are the devices used when you want really low rotation
rates - are actually synchronous motors.

Most stepper motors are two-phase synchronous motors, but there are
some three-phase parts out there - usually described as 5-step
steppers - which could offer smoother rotation, though a
micro-stepping stepper drive should be adequate for most applications.

 

[NASIR]

Sincere thanks to all of you for your kind support and help.

I have already mentioned my application but here is some additonal
information.

--Application is such that use of brushes (hence dc motor) is not
possible.

--If i keep the V/f ratio constant then torque achieved is very very
low, so that's why i increased the voltage and this four times
increase current is at maximum voltage that is 220 V. Even at this
voltage i was unable to achieve the required torque.

--50 Hz motor is actually built and i simulating it on software for
low speed operation so i can implement any suggestion.

--Alse please keep in mind the large airgap effect and dimension
problem while suggesting.

regaeds,
Nasir.

 

[John Popelish]

Sincere thanks to all of you for your kind support and help.

I have already mentioned my application but here is some additonal
information.

--Application is such that use of brushes (hence dc motor) is not
possible.

--If i keep the V/f ratio constant then torque achieved is very very
low, so that's why i increased the voltage and this four times
increase current is at maximum voltage that is 220 V. Even at this
voltage i was unable to achieve the required torque.

--50 Hz motor is actually built and i simulating it on software for
low speed operation so i can implement any suggestion.

--Alse please keep in mind the large airgap effect and dimension
problem while suggesting.

regaeds,
Nasir.

The combination of large air gap and low speed both imply a large
diameter rotor will be needed. And, of course, the more poles the
better. Have you considered a permanent magnet, synchronous motor? I
would suggest a large size stepper motor, but their air gaps are
typically less than a millimeter.

 

[Yzordderex]

Yes, surprising indeed. You really are going to have to get the
currents under control If you insist on a 3 phase induction motor I
will tell you that it can be done. You will run into some difficult,
but not insurmountable problems. A different motor would probably be
my solution as well. Not sure of the point Bill was trying to make
about low frequency currents, but the frequency of interest may be the
slip frequency. The difference between We and Wm -electrical freq and
mechanical frequency. That is what the rotor bars will see. If you
got 220v to play with then you should be able to design in plenty of
control headroom - Big current slewing no problem, but doesn't sound
like you will need much dynamic control.

Dimension constraints sound like they are pushing you into custom
motor rather than off shelf. (Why the airgap specification? Your
professor throw that in for good measure?)

Have you defined the speed regulation? VFD is basically useless where
you intend to operate anyway. Simple flux control is an option where
magnetizing flux vector is controlled to some degree. Abbondanti
wrote paper many years ago on an analog implementaton he had done.
Worked well. If speed control isn't a major issue then this may work
well for you. I think it maybe simple field orientation control.
Open loop vector (sensorless vector) control is a method where speed
is estimated and appropriate manipulation of pwm is used in order to
control direct and quadrature axis. DQ is just stator current broken
down into the discreet components with a transformation from 3 phase
coordinate system into a rectangular system. Components are the flux
producing current and the torque producing current. You try to
control them independantly and maintain an orthogonal system to
produce maximum torque per amp. Brush DC motors do this by way of
commutator. Closed loop implementations are a step ahead where flux,
speed, position information is feedback to controller in order to
precisley control the process. You can design some smokin control
loops and make the motor dance if that gets you off.

So there ya go!

Good luck with project.

regards,
Bob


PS. The fella who mentioned the weird sounding drive - Sounds like
synchronous carrier. The carrier is kept in synchronism with the
fundamental output frequency. You then have a discreet number of
modulation pulses per output cycle. As the motor accelerates the
carrier slides up as well. Shift points are used to keep the carrier
within a reasonable range. Yes sounds cool.

 

[Don Kelly]

[email protected] (NASIR) wrote in message

Considering the number of contributions posted to this thread, it is
surprising that nobody had explicitly made the point that motor torque
depends on the current through the coils, and the voltage applied
across the coils determines the torque only as far as it determines
the current. True.

IIRR, induction motors are not suited for very low speed operation,
because they rely on the current induced in the rotor by the driving
currents in the static coils, and with very low drive frequencies, you
can't put enough current through the static coils to generate enough
dI/dt at the rotor to generate enough current in the rotor.

------
The limitation on current at low frequencies is due to the dominance of
resistance if one tries to keep the V/f ratio constant. If you don't then
saturation becomes a problem.
I have to disagree with the dI/dt statement above. Currents don't induce
currents. The rotor current is due to the rotor induced voltage and this in
turn depends on d(flux)/dt not di/dt. For AC the peak flux density is
determined by the applied voltage, turns and frequency-not current. The
exciting current is then determined by the flux and the magnetic path and
material so its di/dt will depend on the frequency and the peak flux . The
rotor doesn't know this- it only sees the flux produced. The induction
motor is simply a transformer with a rotating secondary and the saturation
and exciting conditions are the same for both.
It is certainly possible to build an induction motor for low speed, low
frequency operation. It will be huge and multi-poled and difficult to cool.
Gears become a viable option and steppers, if they have enough muscle are
another alternative.
-----------

 

[daestrom]

------
The limitation on current at low frequencies is due to the dominance of
resistance if one tries to keep the V/f ratio constant. If you don't then
saturation becomes a problem.
I have to disagree with the dI/dt statement above. Currents don't induce
currents. The rotor current is due to the rotor induced voltage and this in
turn depends on d(flux)/dt not di/dt. For AC the peak flux density is
determined by the applied voltage, turns and frequency-not current. The
exciting current is then determined by the flux and the magnetic path and
material so its di/dt will depend on the frequency and the peak flux . The
rotor doesn't know this- it only sees the flux produced. The induction
motor is simply a transformer with a rotating secondary and the saturation
and exciting conditions are the same for both.
It is certainly possible to build an induction motor for low speed, low
frequency operation. It will be huge and multi-poled and difficult to

cool.

Too true. There has been some experimentation with low speed AC motors for
submarine propulsion. The units were very large and peaked at around 180
RPM. IIRC, they had quite a large number of poles and separately driven
'blowers' to circulate cooling air through the units. I don't recall what
frequency they used at top speed, but I'm sure it was 20 hz or more.

The idea was to eliminate reduction gears. DC drives had been used, but
high maintenance was involved with that. Also brush noise was an issue for
submarine use.

daestrom

 

[NASIR]

Once again sincere thanks to all of you.

--Some of you have suggested use of synchronous motors. Could you
eleaborate that whether i should use switched reluctance motor or
synchronous reluctance motor or some other type?

--Regarding three phase induction motor I will try to implement
a) multi pole design
b) vector control(Have any idea on some web resources dealing with
this)
Would post the results a and b in few days.

regards,
Nasir.

 

[John Popelish]

Once again sincere thanks to all of you.

--Some of you have suggested use of synchronous motors. Could you
eleaborate that whether i should use switched reluctance motor or
synchronous reluctance motor or some other type?

Look into permanent magnet synchronous motors. I think that these are
more efficient (more torque per pound of motor) than the two you list,
above.
http://www.industrialnewsroom.com/fullstory/19174

 

[Don Kelly]

cool.

Too true. There has been some experimentation with low speed AC motors for
submarine propulsion. The units were very large and peaked at around 180
RPM. IIRC, they had quite a large number of poles and separately driven
'blowers' to circulate cooling air through the units. I don't recall what
frequency they used at top speed, but I'm sure it was 20 hz or more.

The idea was to eliminate reduction gears. DC drives had been used, but
high maintenance was involved with that. Also brush noise was an issue for
submarine use.

daestrom

-----
The Swiss use(d?) a 16-2/3 Hz system - and wound rotor induction motors
with resistance control would work nicely at low speed -if they had enough
poles. The problem is that at any frequency there is a flux density
constraint-so pole span must be sufficient and as well the speed requirement
means many poles (20 for about 100 rpm synchronous speed) Room for poles
means a large diameter. Reducing voltage would help but then there is an
additional need for space for larger conductors. Where the saw-off point is
is a matter for more detailed analysis but a meter diameter 1 HP machine
does present some physical problems (based on a breakfast table estimate
before second cup of coffee).