Position control of AC induction motor

Need more info:

Is it geared and at what ratio?

Separate motor on each axis or ONE as stated?

What is the size of the loads? Dead band requirements, if any (
stability )?

Position to a fixed points or is it tracking a moving object?

What is the input and feedback(s) signals?

Yes there are a lot of off the shelf items available to do this and methods.

Lanarcam wrote:

Dennis Mchenney wrote:

The load is a camera, I don't know the inertia moment as of now.
Stability is required for vision at one mile (hope this helps!)
I don't know about deadband


Position to fixed points.


Input : RS422 message to be defined: position to reach, sweeping
area,...
feedback : 64000 counts encoder.

Lanarcam, a properly-gimballed camera is not much load at all and if you are
not gearing the motors in some way, you basically won't get the precision
you are asking for.

You could buy three small VVVF drives and fit encoders to the motor shafts
as per the instructions, but the shudder when stopped will totally stuff
your picture quality.  This is one reason why people use DC servomotors for
your application.

Suggestion:  Go and buy a commercial servo-based PTZ head for your camera.
Even a professional, expensive one will be cheaper and work better than
using AC drives.

I hope this helps.

Cameron:-)

Cameron Dorrough wrote:

This certainly helps to get down to earth .)

The camera is specified and is somewhat larger than the PTZ you
suggested.

In fact contary to what I thought we are using a 3 phase DC
Brushless motor. The driving current is 3 phase AC, with speed
prop to frequency hence the confusion. Sorry about that, I should
change the title of the topic, but I don't know if I can.

The motor is from Artus with 0.29 mrad guaranteed precision without
gear for our application.

Artus will not provide the drives, so we need to find one by ourselves
and make the whole system work.

Our concern for now is to find a drive with position control
capability. We also have thermal and mechanical dimensions
constaints.

Cameron Dorrough wrote:


This certainly helps to get down to earth!

Contrary to what I have said the motor is a 3 phase DC brushless.
The driving current is 3 phase AC hence the confusion.

The camera is specified and is  somewhat larger then the PTZ you
suggested.

The motor is an Artus BM8618Z02. Artus will not provide the drives.
We must find one and make the whole system work.

The precision guaranteed by Artus is 0.29 mrad.

Lanarcam wrote:


You have provided insufficient specifications to make any meaningful
suggestions. Artus manufactures custom motors, generally for military
applications. Given this, simply providing a motor number is meaningless
as is the 0.29 mrad spec. You will need to provide the motor drive
parameters in order to select a drive ( Kt, Kv, Max RPM, Rt-t Lt-t
Number of poles, Halls switches?, Ipeak, Irms, Pmax). I also question
the "25000 Count Resolver". Are you sure it is a resolvers? Resolvers
will typically use a resolver to digital converter which is binary in
nature and the resulting counts per rev will be a power of 2. In
addition, for the accuracy and response you are looking at resolvers
would not be the feedback of choice. Good ones have an accuracy
typically in the 5 arc-min range, and have considerable lag. I suggest
you contact Artus to get complete motor specifications, and their parent
company, Pacific Scientific which manufactures a full line of servo
drives and controllers.

--
jeff

jeff wrote:


Thanks for the tip, I will do that.


25000 is rounded up. Artus suggested the feedback.


Artus declined to design the drive, overworked!
We will ask if Pacific Scientific can supply it.

The problem is we are not experts in this field and the design
was done by mechanical engineers who did not care about
electronics. I am convinced we did that the wrong way, but
we have to make it work hopefully.

3 phase DC. Now I'm lost.

Bob Watkinson wrote:

Artus datasheet:

Brushless motor
Supply voltage : 24 - 32 Vdc
Phases number: 3
Poles pairs numbers : 10
Nominal voltage : 28 Vdc
Resistance ph - ph, R : 2.8 ohms
Self ph - ph, L : 0.6 mH

I saw DC brushless motor in one of their documents, but that
could be a tongue slip.

Lanarcam wrote:

Although there are significant differences in the details of their
designs, the basic magnetic circuits and coil excitation sequences of
brushless DC motors and stepping motors are nearly identical. The
primary difference between their operations is that coil currents are
determined externally in steppers, but commutated internally in the
brushless motors. Both types run from DC power supplies and both have AC
(perhaps with offset) in their coils. The spec above makes sense to me.

The speed and torque of a brushless DC motor as the supply voltage is
varied is roughly that of a series-wound DC motor, with difficulties
arising at low voltages because of the internal electronics. Do your
motors have a separate supply terminal for the electronics? That would
allow operation at lower speeds and allow finer positioning.

I think you have a tough design task with that servo loop.

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

Lanarcam wrote:


Bob: Yes, Brushless DC. It can also be characterized as synchronous AC.
Three phase windings on the stator, permanent magnets on the rotor. it
relies on some sort of external commutation, typically hall switches to
six-step it through one pole-pair.

Lanarcam: Any information on the resolver/encoder/feedback device yet?
Is the motor equipped with hall switches?

The inductance looks high enough for most modern PWM drives. If the
motor is looked at as resistance limited then it has a peak current of
10 Amps. Small PWM amplifiers by Coply or Elmo would be appropriate
depending on your feedback/commutation device.

--
jeff

jeff wrote:

yet?

No hall switches.
Resolver:
- 3 Windings
- X4 : +/- 2 arc min / Nx360 deg
     : +/- 1.5 arc min / +/- 45 deg
14 bit tracking resolver to digital converter
- Accuracy : 8 arc min

Lanarcam wrote:

   ...


How are the coil currents decided on? Algorithmically from the measured
angle? (Before wondering how to servo the motor, one needs to know how
to make it turn.)

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

Jerry Avins wrote:


The resolver provides absolute position within one rev which is used by
the drive to do a sine, sine120, and sine240 lookup for the motor. The
number of poles of the motor is a configuration parameter.  Current is
then apportioned to the windings appropriatly. With some drives it is
necessary to mechanically phase the resolver to the motor poles. In
other cases such as the AMC drive this is accomplished automatically
during commisioning. In this case it simply creates a reference rotating
vector and observes the resolver position for a couple of revolutions.
fron that it knows the motor pole angle as a function of the resolver
position.

--
jeff

jeff wrote:

Thanks. I think there has to be more than that if the motor is to turn,
rather than be locked in place. Torque requires some displacement
between the magnetic axes of rotor and stator.

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

Jerry Avins wrote:

The field rotational speed is proportionnal to the frequency.
By regulating the frequency with respect to an error between the
setpoint and the measured speed, you force the rotation of the
stator field and the rotor catches up.
Please, correct me if I am wrong.

Lanarcam wrote:



Thanks It seems reasonable. There are other ways I can imagine to do it,
too. I don't know what is actually done, that's why I asked.

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

Jerry Avins wrote:

You have described the control scheme for a typical open loop vector
drive. While this will work for spindle drives and other variable speed
work, it is wrong for a closed loop brushless DC motor. It has no zero
speed stiffness. At the inner loop the drive is operating in torque mode
with the current apportioned to each phase based on the rotor position.
See: http://www.embedded.com/2000/0010/0010spectra.htm  about half way
down for an explanation of sinusoidal commutation.

--
jeff

jeff wrote:

Thanks to your hint, I found this link:
http://www.worldservo.com/html/arc.htm
Certainly not trivial, would you recommend this drive, are
there similar products?