velocity-position control

[nebisman]

Can anyone give information about position-velocity feedback control of induction motors and it digital implementation ?

I would be grateful for links or books where detailed study cases are considered.

My interest is nonlinear control of these systems.

THANKS

NEBISMAN

 

[JabaThaSlut]

Encoders in general come in 2 main types, Incremental or Absolute. An absolute encoder outputs ( usually in gray code or similar ) the exact radial postion as a number. Likewise, an incremental encoder simply outputs a square wave from which you must count the pulses. Most incremental encoders used for motor position control have at least 3 outputs as a minimum. (some encoders also provide the inverse outputs as well, used for error checking.)

The outputs are generally known as A, B and Z. A and B are continuous 50% duty cycle TTL level signals with B being electrically 90 degrees out of phase from A. The reason for this is that you can read whether A or B is first to rise, thus if turning for example clockwise, A will rise first or anti clockwise B will rise first. Counting the pulses will give you ratational position information and if needed, speed.

The Z pulse is generally a single pulse once per revolution. It is most often used within a homing or start position setting procedure. A common form of its use is for example a motor driving a machined lead screw. During the "homing" procedure, the motor will drive in a high speed, moving whatever is attached to it (eg an X,Y table) A cam striker fitted to this table activates a limit switch. When this switch is operated, the motor reduces speed to a creeping pace whilst the controller looks for the next instance of the Z pulse. When this pulse is received, this position then is recorded by the controller as it home datum from which all position of the motor is measured.

Motors used for positioning are commonly called AD or DC servo motors. They are specifically designed with a very high number of poles (compared to a standard induction motor). The reason being that the greater the number of poles, the far more acurate speed and position can be controlled. The electronics involved in controlling the motor are generally very complicated. This is due to designs requiring hi speed, hi torque yet lo currents. Just about all aspects of the motor ( acceleration etc) can be controlled to fine detail. Vector control is very popular at the moment.

Search the web for info on motion control and you will find what you seek. How do I know all this?...Coz I work as a Motion control and Industrial Robotics Engineer for more years than I can remember


Jaba

P.S. Sorry for the logwinded reply, I tried to keep it short and sweet.

 

[dkace]

Hallo nebishman,
I can only give you bibliography on generally sensors which have special topics on velocity - positioning sensors. The books have general information on physics of the sensors.
So check them out :
Ramon Pallas-Areny , John G. Webster:Sensors and Signal Conditioning, John Wiley & Sons,Inc. 1991

Jacob Fraden: AIP Handbook of Modern Sensors.Physics, Designs and Applications, American Institute of Physics NY 1993

Julian W. Gardner: Microsensors. Principles and Applications, John Willey & Sons Inc. 1994

Tran Tien Lang : Electronic Measuring Systems

Using any search engine with the words " velocity sensors", or "positioning sensor" you can find allmost every industry that constract such sensors. Of course both last topics are a quite good intro to these sensors.

Good luck on your search! :idea:
:wink: :wink:

 

[nebisman]

Thanks JabaThaSlut

Could you give me more information about vector control of machines

It si suitable to realize this kind of control over microcontrollers of I need DSP for this implementation?



Thanks a million


nebisman

 

[JabaThaSlut]

nebisman,

h**p://w*w.unicous.com/html/vector.html

Is a nice easy to understand starting point. Also jump over to hotbot or lycos and type "flux vector control" or "flux vector drive" (include the inverted commas so the whole phrase is searched). This will bring up countless sites with info etc on this topic. Look at the manufacturers web sites also as many have lots of info on the theory of operation.

A DSP processor is best suited to these types of drives as the theory of operation requires very large numbers of calculations in real time to control the motor current. Using vector control brings many of the advantages of DC servo systems into the AC arena.

Jaba

 

[RegUser_2]

For a linear displacement a LVDT does great job. For a limited rotation, the RVDT is the best. Both the Displacement Transducers rely on the principle of disbalance of the induced signal into a set of complimentary coils in cojunction with phase-sensitive readout. The other way to do the linear displacement is to use the capacitive transducers, but the long-time stability
(or 1/f noise knee) is worst in this case, what is important for a long-duration measurement, except if calibration is performed often enough to pull the error within the needed limits for a siccessfull controll feedback sensor.

 

[JabaThaSlut]

LVDT's and RVDT's are excellent choices, but only for short travel distances. Where continuous rotational position or long lineal travel is required, some form of encoder device is really the best and only way to go. For lineal movement, devices like magnetic strip inductosync or etched glass ( very similar in function to a rotary encoder ) could also be considered. If the means of lineal propulsion is via a rodless pneumatic cylinder, one of the latter devices are a great choice.

Another device that could be considered is the resolver / DC tacho combination. a Resolver generates a series of sinusidal output waveforms very similar in content to an incremental encoder but the device is based around inductive principles.

Jaba