Motor Shaft Zero Pos

Hello there.
I bought a used servo electric motor last week.
The motor contains a sin-cos encoder and is single-turn. There are 1024 periods sin and cos signal outputs per turn. The signals have a 2.5V offset voltage and are the highest 3V, the lowest 2V. that is, at 8V input voltage, 1V peak to peak is the amplitude of the output signal.
I wrote a FOC algorithm(pretty simple) for the servo I bought and I want to test this algorithm. But in order for the algorithm to function properly, I need to know at what degree the rotor position is in its starting position. this is not possible with sin cos encoders as far as I know. Since the encoder of the motor does not provide an index output, I am stuck. at least I have to take the motor to zero rotor angle at the beginning (before I activate the algorithm). But how can I know the zero angle in such motors?

This appears to be a 3ph AC as opposed to the more popular, (easier to integrate) BLDC.
The motors are essentially identical, just the method of commutation is different.
Normally the rotor is exercised and the current monitored to find the optimum commutation position and from there the encoder keeps track through the controller.
Max.

yes this is an ac servo motor not a bldc.
You mentioned a current commutation above. How can I realize this commutation. Currently, the algorithm can read current from 2 phases and perform foc operations based on these values. But could you please explain in a little more detail how I can find the electrical angle of the rotor through current?

I have always a commutation signal on the encoder, this TI app note may hold a clue, BLDC ECM but it may be adaptable,
Max.

According to the data sheet, the single-turn encoder also has a REF+ & REF- output.
Those should be the zero reference.

You can think of the motor as a bit like a single position stepper motor; apply power to the appropriate combination of windings and it will be pulled to a fixed orientation.

Gradually change the relative winding currents - like "microstepping" a stepper - and you can pull it through a fixed movement.
That should allow you to find the zero ref marker after powering up?

Some machine tool servos use encoders that only retain position with power applied and have rechargeable batteries in the drives to run just the encoder while power is off, so they retain the last position.

rjenkinsgb said:

According to the data sheet, the single-turn encoder also has a REF+ & REF- output.
Those should be the zero reference.

You can think of the motor as a bit like a single position stepper motor; apply power to the appropriate combination of windings and it will be pulled to a fixed orientation.

Gradually change the relative winding currents - like "microstepping" a stepper - and you can pull it through a fixed movement.
That should allow you to find the zero ref marker after powering up?

Some machine tool servos use encoders that only retain position with power applied and have rechargeable batteries in the drives to run just the encoder while power is off, so they retain the last position.

Click to expand...

Good Morning
I guess you looked at the wrong point in the datasheet. REF + and REF- are given in table valid for resolver (page 22). For SinCos encoder, it is specified as REF SIN and REF COS in the table (page 21). i don't know what they are Do you think it contains index information?

I have just try to scope to refcos and ref sin. Ref sin connection is bad so i didnt scope it. But i can scope refcos pin of the sincos encoder.
Its a constant 2.88v. Its not depended to turn or agnle of the shaft. Its just a 2.88 v on every position.
So there is no index i guess

What should i do now ?

Isn't zero degrees when sine=0 and cos=max?

Mike.

Pommie said:

Isn't zero degrees when sine=0 and cos=max?

Mike.

Click to expand...

no this is sincos encoder not resolver. In a round, 1024 times sin is max, while cos is zero. or vice versa.

The result of all optical encoders produce two sine waves, in the sin-cos version, the co-tangent result is used to calculate the position, in the quadrature versions, the two sine waves are squared up and in the simplest form, one is pulse is used for count, the other for the direction indication.
I think you will find there is a 90 deg phase shift between both types.
Max.

v1.5 said:

sin is max, while cos is zero

Click to expand...

How is that different to what I said. Except for 90 degrees different.

Mike.

v1.5 said:

I guess you looked at the wrong point in the datasheet. REF + and REF- are given in table valid for resolver (page 22). For SinCos encoder, it is specified as REF SIN and REF COS in the table (page 21). i don't know what they are Do you think it contains index information?

Click to expand...

Sorry, you are correct - I was in a rush and trying to answer before going out..

The encoder you have is a hybrid digital one.

The drive is supposed to use the digital comms channel to read the absolute position of the motor at power-up, and then it can use the sine & cosine signals to track the rotation.

The sine & cosine are two sets of differential signals. They should each connect to a terminating resistor, eg. 150 ohms (between signal and refsignal) then go in to a differential amp to extract the positive or negative difference output.

The difference voltage from each channel is the true analog sine signal.

If you only want to count cycles rather than interpolate within each cycle, you could just use differential line receivers with terminating resistors, to produce logic level quadrature signals.

See the full data on the signals and interface protocols here: