Commutation methods

The important characteristics of a servomotor, such as its very smooth running, high efficiency and optimum thermal utilization, are strongly influenced by the commutation. Commutation refers to the transfer of current from one winding to the next. The moment at which commutation takes place must be harmonized with the magnetic field of the rotor if the servomotor is to operate most effectively.

Rotary servomotors

Mechanical commutation

These motors, which use brushes, generate the alternating fields necessary for operation of the motor through sliding contacts, whose geometrical arrangement switches the current paths. Brush losses and wear are disadvantages of this simple, mechanical commutation method.

Electronic commutation

These modern motors generate the alternating field needed for operation of the motor by means of an electronic circuit which is not subject to either wear or friction. The type of motor and the encoder system in use determine the commutation method.

Absolute encoder system (motor feedback) within one rotation

Samples of this type of encoder system includes: Resolver, EnDat, BiSS and HIPERFACE

Two different commutation methods are involved here:

Mechanical adjustment of the encoder

The motor's encoder system is mechanically adjusted at the factory (the encoder and rotor are matched to one another), but the rotor position is unknown.
The commutation angle is determined once by the P160 command, using the IDN "P0-0-165_Command mode_Static current vector" and the IDN "P-0-057 "Electrical commutation offset". This means that the corresponding mechanical angle coming from the encoder system is displayed and read out in P-0-0058, and is saved in the IDN "P-0-0150_Parameter chanel_Adjustable commutation offset" (motor database). In order for the parameter to be used, the IDN "P-0-0150_Parameter chanel_Commutation mode" (motor database) must be set to 3: "Adjustable offset". The associated value of the IDN "P-0-057 "Electrical commutation offset" is also saved in the motor database.

Electronic adjustment of the encoder system

	Synchronous motors! Electronic adjustment is only required for synchronous motors. In the case of a synchronous motor, the magnetic field of the rotor is generated electronically, and therefore can be set appropriately for the electromagnetic field of the winding.

Depending on the encoder system there are, again, two different commutation methods:

The encoder is always attached to the rotor by the manufacturer in the same rotary position, but the rotor position is not known.
The commutation angle is determined once by the P160 command, using the IDN "P0-0-165_Command mode_Static current vector" and the IDN "P-0-057 "Electrical commutation offset". This means that the corresponding mechanical angle coming from the encoder system is displayed and read out in P-0-0058 and is saved in the encoder system's data store (exceptionally) and in the IDN "P-0-0150_Parameter chanel_Adjustable commutation offset" (motor database). For the parameter to be used, the IDN "P-0-0150_Parameter chanel_Commutation mode" (motor database) must be set to 3: "Adjustable offset". The associated value of the IDN "P-0-057 "Electrical commutation offset" is also saved in the motor database. This method requires a encoder system having a data store and a data line.
The angle between the encoder system and the rotor is determined by the motor manufacturer using a command that is specific to the encoder and is communicated to the encoder system. The encoder system stores this angle, using it for internal calculation, but the rotor position is unknown.
The commutation angle is determined once by the P160 command, using the IDN "P0-0-165_Command mode_Static current vector" and the IDN "P-0-057 "Electrical commutation offset". This means that the corresponding mechanical angle coming from the encoder system is displayed and read out in P-0-0058 and is saved in the encoder system's data store (exceptionally) and in the IDN "P-0-0150_Parameter chanel_Adjustable commutation offset" (motor database). For the parameter to be used, the IDN "P-0-0150_Parameter chanel_Commutation mode" (motor database) must be set to 3: "Adjustable offset". This angle is always included in internal calculation processes. This method requires an intelligent encoder system.

Non-absolute encoder system (feedback) within one rotation

Samples of this type of encoder system includes SIN / COS 1Vss, TTL

In this case, a special commutation procedure (wake&shake) must be run in order to determine the commutation angle. This angle is stored internally, and is taken into account during operation. If the AX5000 is switched off, or if the "EtherCAT-State machine" is switched into "Pre-op" or a lower state, the commutation angle will be lost because the encoder system is not absolute. "Wake&shake" can only operate without error when the drive system is running steadily; in other words, there must not be any vibrations affecting the motor from outside. In addition, a stability investigation using the default values of the "IDN P-0-0165" is necessary the first time the system is operated.

	Oscillatory system! It is important for this stability investigation to examine the application in advance and to determine the oscillation that is potentially most problematic. This case can occur under load conditions, or may be found when unloaded.

WARNING

Risk of injury from uncontrolled movements!

In the method described below, the motor shaft is brought directly to a certain position. Make sure that your application permits this movement, secure the surroundings to prevent unintentional entry, and make sure that nobody is in the hazardous area.

Oscillatory system

It is necessary to analyze the vibration pattern of an oscillating system, and to take appropriate damping measures. Oscillations always have their effect in Phase 2 of "wake&shake"; oscillations are not particularly critical in Phase 1.

Commutation methods 3:

Decaying oscillation

The amplitude (k) and the decay time (l) of this kind of oscillation must be found. The parameters IDN-P-0-0165 "Commutation pos control: Kp" can affect both the amplitude (k) and the decay time (l). The parameter IDN-P-0-0165 " Second phase duration" should be greater than the decay time (l).

Commutation methods 4:

Constant oscillation

This kind of oscillation is unacceptable, as a stable regulation process is not established. The parameters IDN-P-0-0165 "Commutation pos control: Kp" must be checked and modified if necessary. If this does not achieve the desired result, you must damp the vibration mechanically.

Commutation methods 5:

Rising oscillation

The motor shaft is brought to freely definable electrical positions by impressing an appropriate current during this investigation. When this injected current is switched off, the motor should remain in the position that it has reached. BECKHOFF recommends positions of 0°, 90°, 180° and 270°. In critical applications, eight positions (0°, 45°, 90°, 135° ...315°) should be selected instead of four. The current injection is parameterized in the IDN P-0-0165 under "Static current vector", while the freely selectable electrical position is set in the IDN P-0-0057. "Wake&shake" should be carried out in each position; stability of the system is only ensured when this has been done successfully.

	Oscillating system! A mechanical remedy must be provided if the application oscillates. You can carry out the commutation up to a degree using wake&shake but should carefully select the parameters for the IDN "P-0-0165" to make the effect of the oscillation as small as possible, since too much post-pulse oscillation will cause a commutation error. This is because the angle measured after completing the command will be entered as the commutation angle.

WARNING

Risk of injury from uncontrolled movements!

The motor shaft will be moved in steps during the process described below. In Phase 1 the maximum electrical movement is 8 x (the value from "P-0-0-0165_Fist phase position monitoring limit"). In Phase 2 it is 0.5 x (the value from "P-0-0-0165_Fist phase step width"). This formula can only be applied if the previous investigation of stability has been concluded satisfactorily. Make sure that your application permits this movement, secure the surroundings to prevent unintentional entry, and make sure that nobody is in the hazardous area.

The wake&shake commutation function consists of two phases. An approximate determination of the rotor position is carried out in Phase 1, while Phase 2 determines the position more precisely. The aim of the commutation function is to determine the precise position of the rotor with a minimum amount of movement.

Due to the pairs of poles, servomotors exhibit a direct relationship between the electrical and mechanical rotation. One electrical rotation always corresponds to one mechanical rotation divided by the number of pole pairs. A motor with a single pair of poles is illustrated in the following example for the sake of simplifying the calculation.

Parameterization is carried out using the IDN P-0-0165 "Commutation offset calibration parameter". The quoted angles always refer to electrical rotations!

IDN P-0-0165 - Commutation offset calibration parameter.

Parameter	Default	Description
Command mode	0: Static current vector	Selection between two commutation methods
Activation	0: manual	Selection of when the commutation process is started
Static current vector		Commutation methods
Current level	Stationary current in %	Current intensity of the current vector (value = 100% x P0-0093 / P0-0092)
Duration	3000 ms	Period for which the parameterized current is maintained so that any oscillations that may be present can settle, allowing an optimum commutation angle to be reached
Wake and shake		Commutation methods
First phase current vector	Stationary current in %	Current intensity of the current vector (value = 100% x P0-0093 / P0-0092)
First phase ramp up time	100 ms	Time for the current vector "a" to reach its parameterized magnitude
Second phase current level	Stationary current in %	Current intensity of the current vector (value = 100% x P0-0093 / P0-0092)
Second phase ramp up time	500 ms	Time for the current vector "g" to reach its parameterized magnitude
Commutation pos control: Kp	0,04	Amplification factor. Warning: If "0" then Variant 2 will be carried out in Phase 2
Wake and shake expert		Warning: Only experienced users should change the following parameters!
First phase pos monitoring limit	0.5 degrees	Minimum angle of rotation of the rotor required to detect movement
First phase step width	22.5 degrees	Current vector offset or segment detection angle
First phase waiting time after step	150 ms	The time from detection of movement and the next step in Phase 1 or between Phase 1 and Phase 2 (any oscillations in the system have time to settle)
Second phase duration	3000 ms	Period for which the parameterized current is maintained so that any oscillations that may be present can settle, allowing an optimum commutation angle to be reached
Error monitoring (range of motion)	90 degrees	The maximum movement of the rotor before it is switched off, since there would otherwise be a risk that the motor would make an uncontrolled movement.

Commutation methods 14: Commutation methods 15: Commutation methods 16: Commutation methods 17: Commutation methods 18: Commutation methods 19: Commutation methods 20: = identifying characters for the description below

Motor with 3 pole pairs	Motor with one pair of poles

Phase 1 - approximate determination of the rotor position (motor shaft)

Step 1

Commutation methods 23:

Commutation methods 24:

Commutation methods 25:

Commutation methods 26: = see the parameter description for IDN P-0-0165 above

Commutation methods 27: = flux vector of the rotor with permanent magnet.

Step 2

Commutation methods 28:

Commutation methods 29:

Commutation methods 30:

Commutation methods 31:

Commutation methods 32: = see the parameter description for IDN P-0-0165 above

Commutation methods 33: = flux vector of the rotor with permanent magnet.

Sequence:
A current vector "a" is developed during the time "b". Due to the rising magnetic force, the rotor "c" is turned in the direction of the current vector "a". The direction of rotation "d" is transmitted to the feedback system and the AX5000, where it is stored and analysed. If the analysis shows that the direction of rotation "d" of the rotor "c" has not changed when compared with that of the previous impressed current, the process continues.

Step 3

Commutation methods 34:

Commutation methods 35:

Commutation methods 36:

Commutation methods 37:

Commutation methods 38: = see the parameter description for IDN P-0-0165 above

Commutation methods 39: = flux vector of the rotor with permanent magnet.

Sequence:
The current vector "a" is again set to the magnitude "e" in the direction of the rotor "c".
The current vector "a" is now again developed during the time "b". Due to the rising magnetic force, the rotor "c" is turned in the direction of the current vector "a". The direction of rotation "d" is transmitted to the feedback system and the AX5000, where it is stored and analysed. In this case, the analysis shows that the direction of rotation "d" of the rotor "c" has changed when compared with that of the previous impressed current. As a result, the sector in which the rotor "c" is located has been found, and Phase 1 is therefore completed.

Example of an oscilloscope display of Phase 1

Commutation methods 40:

Phase 2 - precise determination of the rotor position (motor shaft)

There are two variants of the precise localization that may be used in Phase 2. In the first variant, the rotor only makes minimal movement, but this does require a very stable system with only a slight tendency to oscillate. In the second variant, the rotor can move by up to a maximum of half the sector Commutation methods 41: , but this method is much more tolerant against oscillation.

The value set in the parameter IDN-P-0-0165 "Commutation pos control: Kp" controls which variant is used:
IDN-P-0-0165 "Commutation pos control: Kp" > 0 --> Variant 1
IDN-P-0-0165 "Commutation pos control: Kp" = 0 --> Variant 2

Variant 1 (IDN-P-0-0165 "Commutation pos control: Kp" > 0 )

Commutation methods 42:

Commutation methods 43:

Commutation methods 44:

Commutation methods 45:

Commutation methods 46: = see the parameter description for IDN P-0-0165 above

Commutation methods 47: = flux vector of the rotor with permanent magnet.

Commutation methods 48: = movement of the rotor

Sequence:
The current vector "g" is developed starting from the final position of the current vector "a" in Phase 1. Due to the rising magnetic force, the rotor "c" is turned in the direction of the current vector "g". The movement is passed through the feedback system to the AX5000, and supplied to a control loop. This control loop immediately corrects the direction of the current vector. This algorithm is executed until the parameterized current intensity is achieved, and the current vector approximately coincides with the flux vector. The current is now maintained over the period "h" which ensures that optimum commutation takes place. In this control algorithm, the rotor only moves minimally through "i".

Variant 2 (IDN-P-0-0165 "Commutation pos control: Kp" = 0 )

Commutation methods 49:

Commutation methods 50:

Commutation methods 51:

Commutation methods 52:

Commutation methods 53: = see the parameter description for IDN P-0-0165 above

Commutation methods 54: = flux vector of the rotor with permanent magnet.

Commutation methods 55: = movement of the rotor

Sequence:
After determining the sector "e" in Phase 1, the current vector "g" is placed exactly in the centre of the sector "e", and this current is developed. Due to the rising magnetic force, the rotor "c" is turned in the direction of the current vector "g" until they coincide. In this static alignment, the rotor cannot move more than half the width of the sector "e".

Using IDN P-0-0165 to affect wake&shake
Parameter	Default	Possible causes that might require a change in the default value
First phase current level	Stationary current in %	Sluggish system, high damping --> increase value Easy-running system, low damping --> reduce value
First phase ramp up time	100 ms	Sluggish system, high damping --> increase value Easy-running system, low damping --> reduce value
First phase pos monitoring limit	0.5 degrees	Application only permits very limited uncontrolled changes in the movement --> reduce value The system has a small amount of damping --> reduce value The loading relationships require more overshoot --> increase value
First phase step width	22.5 degrees
First phase waiting time after step	150 ms	Decay behaviour of the system: Long settling time --> increase value Short settling time --> reduce value
Second phase current level	Stationary current in %
Second phase ramp up time	500 ms
Second phase duration	3000 ms
Error monitoring (range of motion)	90 degrees	Application only permits very limited uncontrolled changes in the movement --> reduce value The system has a small amount of damping --> reduce value The loading relationships require more overshoot --> increase value
Commutation pos control: Kp	0,04	High load stiffness --> increase value Low load stiffness --> reduce value. A special case "0": Variant 2 is carried out in Phase 2

Linear motors

The above description of the commutation process applies equally to rotary motors and to linear motors. Depending on the construction, there are merely some differences of nomenclature (e.g. motor shaft (rotor) @ primary assembly; "degree" @ "mm" (recalculation is needed))

WARNING

Risk of injury from uncontrolled movements!

The primary assembly is moved in steps during "wake&shake". In Phase 1 the maximum electrical movement is 8 x (the value from "P-0-0-0165_Fist phase position monitoring limit"). In Phase 2 it is 0.5 x (the value from "P-0-0-0165_Fist phase step width"). This formula can only be applied if the previous investigation of stability has been concluded satisfactorily. Make sure that your application permits this movement, secure the surroundings to prevent unintentional entry, and make sure that nobody is in the hazardous area.

Linear motors consist of a secondary assembly, whose position is fixed, onto which permanent magnets are attached with alternating polarity and regular spacing. A primary assembly can undergo translatory movement above this magnetic field. This movement is created by generating an electromagnetic field in the primary assembly. Linear motors always have only one pair of poles, and the distance between the poles therefore corresponds to one electrical rotation.

The "Electronic Commutation" section above can be applied to linear motors.

Commutation error "F2A0"

During operation of the motor the commutation is permanently monitored. The following conditions must apply for the AX5000 to detect a commutation error:

The current velocity must be higher than the limit speed set in the IDN "P-0-0069 Commutation monitoring"
The power and acceleration vectors must have different signs.
The current power is greater than 90% of the value in the IDN "P-0-0092 Configured channel peak current".

When these three conditions apply it is very likely that there is a commutation error and that the motor is undergoing uncontrolled acceleration; the AX5000 generates a commutation error and switches the motor torque-free i.e. it stops without control.

WARNING

Risk of injury from uncontrolled movements!

A certain distance will have been travelled from the point when the error is detected until the motor stops. Make sure that your application permits this movement, secure the surroundings to prevent unintentional entry, and make sure that nobody is in the hazardous area. This applies to vertical axes.

	Occurrence of commutation error A commutation error almost always occurs when the axis is commissioned. If this error occurs during regular operation of the axis, then special measures need to be adopted. See next chapter.

Commutation error during regular operation (very rare)

Under special operating conditions the regular operation of the axes can fulfil the three conditions cited above and therefore trigger this error message despite correct commutation. Several examples are given below which, however, occur very seldom:

When the servo drive is operating at the limit (conditions 1 and 3 are met) and external forces cause an opposing torque which then fulfils condition 2, the servo drive generates a commutation error.
The servo drive is operating at the limit (conditions 1 and 3 are met) and an oscillating current is produced due to a rapid change of direction or speed. Condition 2 is then also met, and a commutation error arises.

If these examples do not apply to your application, analyse the application, and try to find the cause. If you are unable to remedy the cause but still wish to operate the axis, there is only one option for suppressing the commutation error:

Parameterise the value of the IDN P-0-0069 to the permitted maximum speed of the motor so that point 1 of the above-mentioned factors cannot apply and the commutation error will no longer appear.

WARNING

Risk of injury from uncontrolled movements!

Increasing the value of the IDN "P-0-0069" to the highest speed always means that the commutation monitoring will no longer cause errors, even when other conditions call for this. This is particularly critical when the motor is being replaced. If the value of the IDN "P-0-0069" is NOT reset, then uncontrolled movements of the motor may occur. Beckhoff recommends that you should NOT increase the value of the IDN "P-0-0069".

	Drive design As a rule, the drive should not be designed at the limit i.e. the current power should reach a max. of 90% of the P-0-0092 "Configured channel peak current" value.