Clarification of the motor rotor position

[Physicslearner500039]

I was referring to the following statement in the document which confused me and also attaching the document for your reference. "First, we need to know the rotor position. The position is typically related to phase A." How the above statement is true. Please help me to understand. I am assuming that Phase A is current here. The statement is on page 1.

 

[jim hardy]

How the above statement is true.

It's a definition, not a calculation.
It's just the choice of a reference point. You measure rotor's angular position from some agreed on point .
Where do you choose to place your zero degrees mark?
Author chose centerline of phase A for his zero.
Next couple sentences describe it.

We can use
an absolute position sensor (e.g.,
resolver) or a relative position sensor
(e.g., encoder) and process called
alignment. During the alignment, the
rotor is aligned with phase A...

http://www.ti.com/lit/an/bpra073/bpra073.pdf might be of help.

 

[Physicslearner500039]

Thank you very much. But this question will clear lot of doubts in my mind. Suppose i take any point as refernce or starting point. Then when i consider park transform the electrical angle will be our assumed refernce point or the actual electrical angle which we get through resolver. Please advise. I am very very confused.

 

[jim hardy]

Any measurement requires two points - the one being measured and the one from where you measure it - your reference point.

Ever been on a merry-go-round? It has something to teach us.

Imagine yourself very small and sitting on your motor's rotor as if it were a merry-go-round, holding a magnetic compass.
The compass will align with the rotating magnetic field made by the three phase currents in the stator. It'll spin at synchronous speed, just as the rotor itself is doing.
So to you it will appear stationary since you're both spinning.
If you brought your protractor along you can measure the angular difference between rotor centerline and that rotating field.
The compass needle shows you where is the rotating field
and you can see which direction is the rotor pointing because you wisely marked its centerline before climbing aboard..
Just as on a merry go round you can estimate the angle to another passenger on it even though the world is whirling by behind both of you.

The Park transform(as i understand it) gives you that angle to the pretty girl who's sitting on the merry-go-round seat perfectly aligned with the stator's rotating magnetic field.
The sensor is 'aligned' so that it reads zero when both the rotor and the stator's rotating field are aligned with centerline of phase A. That's your reference.

So,
the angle you get from Park will be that angle you got from the sensor compared to a calculated(or measured) angle of the phase A sinewave.
Comparison is probably by subtraction, I do not pretend to follow all the algebra in the Park transform .
If you choose a different zero point all your angles will be offset by that much difference which only complicates your algebra.
The equations look fearsome but the concept is not so difficult - not at all unlike phasor notation..

That's the mental trick to grasping the concept. Use your imagination (and all your senses) and build from what is already familiar. I assume they still teach phasor notation - perhaps by another name? Been fifty years ago now for me..

I hope this helps. Draw a picture in Paint and post it ?

Sorry to be so wordy. But thought for me progresses in tiny steps. I try to maintain one step per line of text.

old jim

 

[Physicslearner500039]

So much thank you for your time. I am completely following the attached document and implementing it. Your reply has helped me to clear few doubts. Could you please review my understanding. My system is a Permanent magnet NMEA motor and i get the feedback from an encoder attached to it. After discussing with you and after seeing the following picture in the attached document.

Step1: Apply a constant voltage to phase A by switching on switch "a". This will make the motor rotate and align with phase "A"

Step2: When received the motor command and based on the requested Torque (I). Apply the transformations and derive the voltage to be applied to the stator. Using the space vector modulation apply the the pwms to the switches.
Step3: once the motor starts rotating get the encoder pulses and from this identify the rotor position. This will be the electrical angle.
Very briefly am i correct in the understanding? Please help.

I have doubts regarding the I nominal. what does it represent? How much value should I take?

 

[jim hardy]

I think there's quite a bit more to it than that .
I've never done FOC, just am aware of it so i can't tell you 'how to' do it.

in your other thread

https://www.physicsforums.com/threads/rotor-magnetic-field-and-stator-magnetic-field.918667/

is a link to a one hour video by a TI engineer on how to implement FOC
it's not something you will learn to do in one day
here's the link again

https://training.ti.com/field-oriented-control-permanent-magnet-motors

Please watch it. In the first few minutes he summarizes the steps, then goes into detail about each of them.

 

[Asymptotic]

I have some experience in ancient analog servo drives (programmed using trim pots and "personality" circuit boards, not by digital communications), and take this with a grain of salt, but ...

1. Two 'switches' in addition to 'a' must be closed to lock the rotor to a "zero", home position.
In this example, the closed switches are a, not b, and not c.

The drives I'm familiar with employed current limiting when the output bridge was so configured to prevent motor windings from overheating and burning up.

2. I didn't read deeply on how this TI chip uses an incremental encoder for position feedback, but my impression was it re-learned where home is on every power-up. My preference is a sin/cos resolver or other form of absolute encoder. That said, now that the motor rotor has been magnetically driven to a home position, the general idea is to align a position feedback device to the rotor so it is also 'home'.

Traditionally this was done by loosening the position encoder shaft coupling, physically rotating it until measured feedback was also at zero position, then retightening the coupling. More modern, higher precision drives often have active electronics on the motor encoder circuit board, and an ability to save an offset position in non-volatile memory. Preliminary zeroing can still be (and often is) done using the traditional method, but final zeroing is by adding or subtracting an offset value to raw position feedback until perfect motor matching is obtained, saving this offset value, and from then on using this offsetted position feedback to commute the motor.

 

[Physicslearner500039]

Thank you both for the replies. Yes what you said is correct i need to close the not b and not c. Yeah i have gone through these videos a number of times and have been reading about FOC for a long time. I thought i understood the concepts. Do you feel i still need to learn some more concepts? Any particular area where i can concentrate? Please suggest.