Additional Axis 3HAC021395-001 RevF En

8/9/2019 Additional Axis 3HAC021395-001 RevF En

1/218


2/218


3/218


4/218


5/218


6/218


7/218


8/218


9/218


10/218


11/218


12/218


13/218


14/218


15/218


16/218


17/218


18/218


19/218


20/218


21/218


22/218


23/218


24/218


25/218


26/218


27/218


28/218


29/218


30/218


31/218


32/218


33/218


34/218


35/218


36/218


37/218


38/218


39/218


40/218


41/218


42/218


43/218


44/218


45/218


46/218


47/218


48/218


49/218


50/218


51/218


52/218


53/218


54/218


55/218


56/218


57/218


58/218


59/218


60/218


61/218


62/218


63/218


64/218


65/218


66/218


67/218


68/218


69/218


70/218


71/218


72/218


73/218


74/218


75/218


76/218


77/218


78/218


79/218


80/218


81/218


82/218


83/218


84/218


85/218


86/218


87/218


88/218


89/218


90/218


91/218


92/218


93/218


94/218


95/218


96/218


97/218


98/218


99/218


100/218


101/218


102/218


103/218

4 Configur ation

4.3.3.2. Defining the user frame for a rotational single axis

1013HAC 021395-001 Revision: F

C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g

h t s r e s e r v e d .

Result

The result of the calculation is displayed (expressed in the world coordinate system). Thefollowing values are shown:

If the estimated error is acceptable, press OK to confirm the new user frame.

If the estimated error is unacceptable, press Cancel to redefine the calibration.

L is ted v al ues Des cr ip ti on

Method Displays the selected calibration method.

Max error The maximum error for one positioning.

Min error The minimum error for one positioning.

Mean error The accuracy of the robot positioning against the tip.

Cartesian X - Z The x, y, z coordinates for the user frame.

Quarternion 1-4 Orientation components for the user frame.

Continued


104/218

4 Configuration

4.3.3.3. Defining the user frame for a two axes positioner

3HAC 021395-001 Revision: F102

C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


4.3.3.3. Defining the user fr ame for a two axes posi tioner

Parameter file requiredIt is possible to define positioners with more than one axis. To achieve the best possible

performance from such a positioner, a set of data, describing its kinematic and dynamic properties (among other things), must be defined. This data cannot be defined in the system parameters, but must be read from a parameter file. If no file was supplied with themanipulator, the manipulator cannot be coordinated with the robot. It can, however, bedefined as a number of separate external axes.

Differences between one and tw o axes posit ioner The principles for defining a user frame for a two axes positioner are the same as for a oneaxis positioner, see Defining the user frame for a rotational single axis on page 99 . However,

note that the axis must be moved to a negative position (see step 9 below).For a positioner with more than one axis, a 4 point calibration is performed for each axis.

User frame definition procedure

Acti on

1. Tap the ABB menu - Calibration .

2. Select the two axes positioner

3. Select Base Frame .

4. Select 4 Points for Axis 1 .

5. If you have a MultiMove system, select which robot to use for the calibration.For non-MultiMove systems, go directly to the next step.

6. Select Point 1 .

7. Jog the robot as close as possible to the reference point.

8. Modify the position by tapping Modify Position .

9. Move the first axis to a new negative position.

10. Repeat the steps 6-9 for the points Point 2, Point 3 and Point 4.

11. If you want, you can save the entered calibration data to a file. Tap Positions and thenSave . Enter the name of the file and then tap OK .To restore this calibration, the file can be loaded from Positions - Load , instead ofperforming steps 6-10.

12. Select 4 Points for Axis 2 and repeat step 5-11 for the second axis.

13. Press OK to calculate the user frame for the positioner.

Continues on next page


105/218

4 Configur ation

4.3.3.3. Defining the user frame for a two axes positioner


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


Result

The result of the calculation is displayed (expressed in the world coordinate system). Thefollowing values are shown:

If the estimated error is acceptable, press OK to confirm the new user frame.

If the estimated error is unacceptable, press Cancel to redefine the calibration.NOTE!

When defining a work object for a coordinated motion, the user frame part of the work objectis left empty (unit frame). Instead the user part is computed when needed using the kinematicmodel and the joint position for the mechanical unit.

L is ted v al ues Des cr ip ti on

Method Displays the selected calibration method.

Max error The maximum error for one positioning.

Min error The minimum error for one positioning.

Mean error The accuracy of the robot positioning against the tip.

Cartesian X - Z The x, y, z coordinates for the user frame.

Quarternion 1-4 Orientation components for the user frame.

Continued


106/218


107/218

5 Commutation

5.1. Commutate the motor


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


5 Commutation


OverviewThis chapter describes how to use the call service routine commutation, so the additionalmotor runs properly.

The service routine commutation is used to:

Find a commutation value for a synchronous permanent magnet motor.

Check motor phase order

Verify the pole pair parameter value is correctly typed in.

Check resolver connection

How to install a new motor

Set the motor in s afe mode/ normal mo deThe system parameters can be changed in the FlexPendant or in RobotStudio Online. On theFlexPendant tap Control panel/Configuration/Topics/Motion/Drive System. Then change the

parameter current_vector_on to TRUE or FALSE in the DRIVE_SYSTEM. If the motor is to be set in safe mode set the parameter to TRUE or normal mode FALSE.

CAUTION!

If the motor is not properly installed, it can run away and destroy itself or other equipment.To avoid this set the motor in safe mode.

Ac ti on Note

1. Set the motor in safe mode by changing the systemparameter current_vector_on to TRUE.

Set the motor in safe mode/normal mode on page 105 .

2. Start the service routine Commutation. See Operating manual - IRC5 withFlexPendant sectionProgramming and testing -Running a service routine .

3. Check motor phase order connections. Check motor phase connectionsorder on page 106 .

4. Check resolver connection. Check resolver connections onpage 106 .

5. Move the motor to commutation position. For the pre-Commutated motor: Check the

motor phase connections. For the none commutated motor:

Commutate the motor by updating thecommutation offset.

Check the motor phaseconnections on page 106 .Update commutation offset onpage 106 .

6. The commutation is now finished and the motor isready to use. When exiting, the program ask if themotor is to be set in normal mode. The motor canalso be set to normal mode by changing the

parameter current_vector_on to FALSE.

See Set the motor in safe mode/normal mode on page 105 .



108/218

5 Commutation



C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


Check motor phase connections order

By stepping the motor in positive direction from the service routine, the motor shaft shall turnin counter clockwise direction. If the shaft is seen from the resolver side and clockwise fromthe drive shaft side.

xx0600003335

If the motor is turning in the wrong direction then the motor phases has been swapped. Trychanging RST to SRT, RTS or TSR.

Check the pol e pair parameter Check if the pole pair parameter is loaded with the correct value by stepping the motor fromthe service routine. The motor shall turn 1/16 of a revolution for every step command.

Check resolver connectionsFrom the service routine step the motor in positive direction. The resolver is connectedcorrectly if the motor angle in the jogging window is increasing. Otherwise check the wiringof the resolver.

Check the motor phase connections

Step through the commutation angles to make sure that the cables are connected to the right phase. For best result commutate with a free mounted motor. There are a numbers of correctcommutation angles (same as pole_pair parameter). If the difference is a number of 6.283185/

pole_pair values - the commutation is ok. Otherwise all motor phases shall be moved one stepforward or backward (same order! RST -> STR or TRS). A commutation value set by themotor manufacturer is normally more accurate than a value updated with this method.

Update commutation offsetTo get a good commutation position the motor must not be affected by gravity or large frictionfrom equipment connected to the motor. For best result commutate with a free mountedmotor.

When the motor is aligned, the resolver commutation parameter can be set. When the parameter is set the database is also updated.

Continued


109/218

6 Tuning

6.1. Tuning the commutation offset


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


6 Tuning

6.1. Tuning the commutation o ffset

GeneralBefore using an additional axis, you must tune the motors commutation offset. This requiresthat you connect a DC power source between two nodes and then measure the position of themotor.

NOTE!

ABB motors are precommutated with the commutation value 1.5708. Therefore, an ABBmotor does not require tuning of the commutation offset.

PrerequisitesThe motor must comply with the specifications in Motors on page 201 . The resolver mustcomply with the specifications in Resolvers on page 206 .

Required materialThis is a list of what you need to perform the tuning:

Measuring procedureThis procedure describes how to measure the commutation position of a motor.

Material Description

PC with Test Signal Viewer The software Test Signal Viewer is delivered on theRobotWare CD.

Power supply 24 V (DC).The power supply should be equipped with a relay that trips

at short circuit. Otherwise a fuse will burn every time thepower is applied.Check the motor data to see the current required from thepower supply.

2 cable sets Cables to brake release and motor phase.Each cable set includes one plus and one minus cable.

Motor documentation Motor data sheet and electrical connection drawing.

Ac ti on1. Deactivate the axis of the motor you want to tune.

2. Switch off the controller.

3. Disconnect the power cable to the motor.

4. Disconnect the motor from the gear (or in some other way make sure the motor is notaffected by external torque and friction).

5. If the motor is using a brake, release it by connecting the power supply to the contactpins for the brake release.See motor specifications for max brake current, which contact are for the brake releaseand the polarity of the contacts (if any).

6. Ensure that the brake is released by manually turning the motor.



110/218

6 Tuning

6.1. Tuning the commutation offset


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


7. Connect the power supply with the plus cable to the phase S (V) and the minus cable(0 V) to the to the phase T (W).

A short pulse is enough to move the motor to its commutation position. Disconnect thepower after the voltage pulse.

8. Connect the power to give another voltage pulse to the motor. If the motor is already inits commutation position it should not move this time.

9. Disconnect the power supply from the brake release, so that the motor brake is on.

10. Reconnect the power cable from the drive module to the motor.

11. Activate the axis.Do not move any mechanical unit.

12. Start the controller.

13. Configure Test Signal Viewer, selecting mechanical unit and the signal Resolver_angle.Zoom in on the signal so you can read at least 2 decimals.Note that the number of commutation positions are equal to the number of pole pairs.E.g. a motor with 2 pole pairs have 2 possible values for this measurement. It does notmatter which of the commutation points you are measuring.

14. Set the measured value to the parameter Commutator Offset in the type Motor Calibra-tion .

Ac ti on

Continued


111/218

6 Tuning

6.2.1. Introduction


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


6.2 Tuning

6.2.1. Introduc tion

OverviewThe servo control parameters can be adjusted (tuned) to achieve the best possible motion

performance.

This section contains a complete description of how to tune the axes and is divided into thefollowing sections:

Tuning of axes, complete procedure on page 112 details the complete procedure oftuning the axes, including necessary preparations and references to more detailedinstructions of setting tuned values for the different parameters.

Separate procedures that detail how to set the respective tuning parameters.


112/218

6 Tuning

6.2.2. Defining signals in Test Signal Viewer


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g



Definin g test signalsTest Signal Viewer is used during tuning of the parameters. This section shows the signalsthat may be used, and how these are defined from the menu Commands - Define Test Signal.

Signal identifier

Some of the signals may be defined from a drop down list as shown in the figure below:

en0400000763

- speed

- torque_ref



113/218

6 Tuning



C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


Signal id entifier, manual

Some of the signals may be defined manually by entering numbers as shown in the figure below:

en0400000764

6 speed

9 torque_ref

55 positive torque_limit

56 negative torque_limit

57 torque feed forward

Continued


114/218

6 Tuning

6.2.3. Tuning of axes, complete procedure


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g



GeneralThis section details the complete procedure for tuning of the axes, which is done in order toachieve the best possible motion performance of the robot system.

WARNING!

The system is unstable and therefore dangerous during the tuning process, since bad parameters or parameter combinations may be used! The safety procedures of the robotsystem must be carefully followed throughout the tuning process.

How to set the parameter valuesSystem parameters for tuning are preferably edited by using RobotStudio Online. They canhowever also be edited on the FlexPendant, and even directly in the configuration files by

using a text editor. System parameters are listed in the section System Parameters on page149 . More information, including parameter names in the configuration files, can be found inTechnical reference manual - System parameters .

PreparationsThe following preparations must be made before tuning a servomotor axis:

Set up the Test Signal Viewer software according to the instructions in the pdf-fileTest Signal Viewer , enclosed with the program.

Make sure that the additional axis is commutated and calibrated. Any position may bedefined as the calibration position.

Select or use default tuning parameters so the axis may be jogged without stoppingdue to speed or torque supervision.

Tunin g calib rated axes, overview

Tuning a calibrated axis means setting values for several parameters in the motionconfiguration type Lag Control Master 0 . The procedure below is an overview of thecomplete tuning procedure and includes references to more detailed instructions of how toactually set each parameter.

Ac ti on

1. Make the Preparations on page 112 .

2. Use RobotStudio Online or the FlexPendant. Select the configuration topic Motion andthe type Lag Control Master 0 .

3. Set initially tuned values for the parameters Kv, Kp and Ti. This is further detailed in thesection Initial tuning of Kv, Kp and Ti on page 115 .



115/218

6 Tuning



C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


4. Set a value for the parameter FFW Mode (feedforward mode) by choosing one offollowing modes:

No : This is the easiest configuration with no adjustments of the position lag. (thismode is called 0 in the configuration file)

Spd : This is the recommended configuration. In this configuration the controllerreceives information about the desired speed of the axis. As a result, the positionlag is considerably reduced compared to the No configuration.

Trq : In this configuration the controller uses the desired speed and accelerationof the axis to calculate the desired motor torque. This requires knowledge of themass moment of inertia of the axis, which must be supplied by the user. For thisreason this configuration is more difficult to tune, specially for the axes affectedby gravity. It is only recommended for experienced users.

The controller is driven by the position lag. This is the offset in time from the point ofgiven command to the point of actual performance of the command. The lag may beadjusted with the different modes in the parameter FFW Mode .Depending on which mode is chosen for the FFW Mode , further tuning parameters canbe set later on.

5. Set a value for the parameter Inertia , depending on the chosen FFW Mode : FFW Mode = No : set Inertia to 0 FFW Mode = Spq : set Inertia to 0 FFW Mode = Trq : calculate and set Inertia according to the section Specifying

the inertia on page 121 .

6. If parameter FFW Mode is set to Spd , the following parameters are also available: Bandwidth , should be left at its default value Delay , should be left at its default value.

7. If parameter FFW Mode is set to Trq , the following parameters are also available:

Bandwidth , should be left at its default value. Adjustment is detailed in thesection Tuning of Bandwidth on page 122 .

Delay , should be left at its default value (0.004). In rare cases, increasing thevalue may reduce the speed overshot.

Resonance frequency (flexibility compensation filter), should initially be left at itsdefault value. May be adjusted once the other parameters are set, if thetorque_ref signal is oscillatory due to mechanical resonance. How to tune thisparameter is detailed in the section Tuning of resonance frequency on page 124 .

Resonance damping (flexibility compensation filter), should be left at its defaultvalue (0.01).

8. In the type Acceleration Data , set tuned values for Nominal Acceleration and NominalDeceleration as detailed in the section Tuning of Nominal Acceleration and NominalDeceleration on page 125 .

9. In Lag Control Master 0 , set finally tuned values for the parameters: Kv, Gain Speed Loop Kp, Gain Position Loop Ti Integration Time Speed Loop

This is further detailed in the section Final tuning of Kp, Kv and Ti on page 128 .

Ac ti on

Continued



116/218

6 Tuning



C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


Tuning uncalibr ated axes, overview

Tuning an uncalibrated axis means setting values for several parameters in the motionconfiguration type Uncalibrated Control Master 0 . The procedure below is an overview ofthe complete tuning procedure and includes references to more detailed instructions of howto actually set each parameter.

Ac ti on

1. Make the Preparations on page 112 .

2. Use RobotStudio Online or the FlexPendant. Select the configuration topic Motion andthe type Uncalibrated Control Master 0 .

3. Set the initial tuning values of Kv, Gain Speed Loop , Kp, Gain Position Loop , and TiIntegration Time Speed Loop . Use tuned values, set during tuning of a calibrated axis.Use the values so that the axis is movable regardless of the position.

4. Set the tuning value of Speed Max Uncalibrated . Maximum speed for uncalibrated axis(rad/s on motor side).

5. Set the tuning value of Deceleration Max Uncalibrated . Maximum deceleration foruncalibrated axis (rad/s 2 on motor side). Recommended value: Nominal Deceleration *Transmission Gear Ratio . How to set the value for the nominal deceleration is detailedin the section Tuning of Nominal Acceleration and Nominal Deceleration on page 125 .

6. Set the tuning value of Acceleration Max Uncalibrated . Maximum acceleration for uncal-ibrated axis (rad/s 2 on motor side). Recommended value: Nominal Acceleration *Transm Gear Ratio . How to set the value for the nominal acceleration is detailed in thesection Tuning of Nominal Acceleration and Nominal Deceleration on page 125 .

7. Set the final tuning values of Kv, Gain Speed Loop , Kp, Gain Position Loop , and TiIntegration Time Speed Loop , as detailed in the section Final tuning of Kp, Kv and Ti onpage 128 .

Continued


117/218

6 Tuning

6.2.4. Initial tuning of Kv, Kp and Ti


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


6.2.4. Init ial tuning of Kv, Kp and Ti

GeneralThis section details how to make the initial tuning of the parameters Kv, Gain Speed Loop ,Kp, Gain Position Loop and Ti Integration Time Speed Loop .

The general strategy is to tune Kv first while keeping Kp constant and without integral effect(Ti is set to a high value, e.g. 10), then tune Kp to its maximum value without vibration/oscillation; finally tune Ti and other parameters.

NOTE!

Note!

Check that the additonal axes motor file contains the correct motor data.

Parameter descripti onThe parameters Kv, Gain Speed Loop , Kp, Gain Position Loop and Ti Integration Time Speed

Loop belongs to the type Lag Control Master 0 and are further described in the section SystemParameters on page 149 .

RAPID programProgram a back-and-forth motion of the axis. For the final tuning of the control parametersof the axis it is convenient to use the TuneServo command. Each procedure below includesan example of a RAPID program that can be used.

Initial tuning o f Kv

The procedure below details how to make the initial tuning of the parameter Kv, Gain Speed Loop .

WARNING!


Ac ti on

1. Set the parameter FFW Mode to No.

2. Make following changes in the type Lag Control Master 0 :

set the value of parameter Kp, Gain Position Loop to 5 set the value of parameter Ti Integration Time Speed Loop to 10 (a big value in

order to eliminate integral portion).Restart the controller for the changes to take effect.



118/218

6 Tuning



C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


3. Increase the Kv value by 10% in each motion loop, until the axis starts to vibrate/oscillate or a clear vibration can be heard from the axis, either during motion or when

stationary. The axis velocity supervision may also indicate speed failure.The following RAPID program can be used to increase Kv by 10% of the default valueset in the loaded configuration file:

MODULE Kv_tunePROC main()

VAR num i;VAR num per_Kv;VAR num Kv;

TuneReset;FOR i FROM 0 TO 40 DO

per_Kv:=100+10*i;Kv:=1*per_Kv/100;TPErase;TPWrite "per_Kv = "\Num:=per_Kv;TPWrite "Kv = "\Num:=Kv;TuneServo STN1,1,100\Type:=TUNE_KP;TuneServo STN1,1,100\Type:=TUNE_TI;TuneServo STN1,1,per_Kv\Type:=TUNE_KV;MoveJ p1,v1000,z50,tool0;MoveJ p2,v500,z50,tool0;MoveJ p1,v1000,z50,tool0;WaitTime 1;

ENDFORENDPROC

ENDMODULENote! The velocity data and test positions may be modified depending on the type ofrobot and axis to be tuned.The RAPID instructions are described in the Technical reference manual - RAPIDInstructions, Functions and Data types .The Torque_ref signal in the Test Signal Viewer may be used to evaluate the degree ofvibration/oscillation. A typical plot of the signal is shown in the figure Illustration,torque_ref plot on page 117 .

4. Once an unstable point is reached, divide the current value of Kv by 2.

5. Enter the new value in the parameter Kv, Gain Speed Loop .

Ac ti on

Continued



119/218


120/218

6 Tuning



C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


Initial tuning of Kp

The procedure below details how to make the initial tuning of the parameter Kp, GainPosition Loop .

WARNING!


Ac ti on

1. Leave the initially tuned value of the parameter Kv, Gain Speed Loop and the earlier setdefault values of Kp, Gain Position Loop and Ti Integration Time Speed Loop unchanged.

2. Increase the Kp value by 10% in each motion loop, until the first signs of overshootingare observed in the velocity plot.The following RAPID program may be used to increase Kp by 10% of the default valueset in the loaded configuration file:

MODULE kp_tunePROC main()

VAR num i;VAR num per_Kp;VAR num Kp;


per_Kp:=100+10*i;

Kp:=5*per_Kp/100;TPErase;TPWrite "per_Kp = "\Num:=per_Kp;TPWrite "Kp = "\Num:=Kp;TuneServo STN1,1,100\Type:=TUNE_KV;TuneServo STN1,1,100\Type:=TUNE_TI;TuneServo STN1,1,per_Kp\Type:=TUNE_KP;MoveJ p1,v1000,z50,tool0;MoveJ p2,v500,z50,tool0;MoveJ p1,v1000,z50,tool0;WaitTime 1;

ENDFORENDPROC

ENDMODULENote! The velocity data and test positions may be modified depending on the type ofrobot and axis to be tuned.The RAPID instructions are described in the Technical reference manual - RAPIDInstructions, Functions and Data types .

A typical plot with speed overshot in Test Signal Viewer is shown in the figure Illustration,speed plot on page 119 .

3. Once an overshot is observed, subtract 1 from the current value of Kp.If an overshooting is observed at a later time, the Kp must be reduced to an even lowervalue.

4. Enter this new value in the parameter Kp, Gain Position Loop .

Continued



121/218

6 Tuning



C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


Illustration, speed plot

How to choose the speed signal in Test Signal Viewer is detailed in section Defining signalsin Test Signal Viewer on page 110 .

xx0400000648

Initial tuning of TiThe procedure below details how to make the initial tuning of the parameter Ti IntegrationTime Speed Loop .

WARNING!


A Normal plot

B Overshoot

Ac ti on

1. Leave the initially tuned values of parameters Kv, Gain Speed Loop and Kp, GainPosition Loop unchanged.

2. Set the value of Ti Integration Time Speed Loop to 1.0 as default.

Continued



122/218

6 Tuning



C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


ResultThe parameters Kv, Gain Speed Loop , Kp, Gain Position Loop and Ti Integration Time Speed

Loop are now initially tuned. Continue with the tuning process as described in next section.The complete tunig procedure is detailed in section Tuning of axes, complete procedure on

page 112 .

3. Reduce the Ti value by 10% in each motion loop, until an overshot is observed on thevelocity profile in Test Signal Viewer.

Looking at the torque_ref signal can also help to determine the critical value of Ti.The following RAPID program may be ued to reduce the Ti value by 10% of the defaultvalue set in the loaded configuration file:

MODULE ti_tunePROC main()

VAR num i;VAR num per_Ti;VAR num Ti;


per_Ti:=100-10*i;Ti:=1*per_Ti/100;TPErase;TPWrite "per_Ti = "\Num:=per_Ti;TPWrite "Ti = "\Num:=Ti;TuneServo STN1,1,200\Type:=TUNE_KV;TuneServo STN1,1,250\Type:=TUNE_KP;TuneServo STN1,1,per_Ti\Type:=TUNE_TI;MoveJ p1,v1000,z50,tool0;MoveJ p2,v500,z50,tool0;MoveJ p1,v1000,z50,tool0;WaitTime 1;

ENDFORENDPROC

ENDMODULENote! The velocity data and test positions may be modified depending on the type ofrobot and axis to be tuned.The RAPID instructions are described in the Technical reference manual - RAPIDInstructions, Functions and Data types .

4. Once an overshot is observed, stop reducing and instead increase the tuned value by5-10%, in order to remove the effect.

5. When the effect is removed, enter the new value in Ti Integration Time Speed Loop .

Ac ti on

Continued


123/218


124/218

6 Tuning

6.2.6. Tuning of Bandwidth


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g



GeneralThis section details how to set the tuned value for the parameter Bandwidth .

Parameter descriptio nThe parameter Bandwidth is found in the data group Lag Control Master 0 and is furtherdetailed in the section Lag Control Master 0 on page 157 .

Illustration, speed overshot, bandwidth tuningThe figure below shows a typical plot in Test Signal Viewer when tuning the parameter

Bandwidth .

xx0400000652

Tuning of bandwidth

The procedure below details how to tune the parameter Bandwidth .

Ac ti on

1. Use properly tuned values for the parameters Kv, Gain Speed Loop , Kp, Gain PositionLoop and Ti Integration Time Speed Loop .

2. Use the default value (25) for the parameter Bandwidth .



125/218

6 Tuning



C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


ResultThe parameter Bandwidth is now initially tuned. Continue with the tuning process describedin the next section, and in section Tuning of axes, complete procedure on page 112 .

3. In order to verify and tune the default Bandwidth value, run a motion program to see ifthe speed overshot occurs. If the overshoot is out of the tolerance, reduce the value of

Bandwidth under the system resonance frequency.The system resonance frequency may not always be identified. Therefore the valuemay be varying several times and it may be necessary to perform a check of the speedovershoot and of the following error between command position and actual position.Normally, the bigger the Bandwidth value, the smaller the following error.

A typical speed overshoot plot is shown in the figure Illustration, speed overshot,bandwidth tuning on page 122 .

Ac ti on

Continued


126/218

6 Tuning

6.2.7. Tuning of resonance frequency


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


6.2.7. Tuning of resonance fr equency

GeneralThis section details how to set the value of the system parameter Df (resonance frequency).

Parameter descriptio nThe parameter Df is found in the type Lag Control Master 0 and is further detailed in thesection Lag Control Master 0 on page 157 .

Tunin g of reson ance frequencyThe procedure below details how to tune the resonance frequency.

WARNING!

The system is unstable and therefore dangerous during the tuning process, since bad

parameters or parameter combinations may be used! The safety procedures of the robotsystem must be carefully followed throughout the tuning process.

ResultThe parameter Df is now tuned. Continue with the tuning procedure as described in the nextsection, and in section Tuning of axes, complete procedure on page 112 .

Ac ti on

1. Measure the distance between the resonance peaks (in points) on the plot of thetorque_ref signal. Divide the sampling frequency with this value:resonance frequency = sampling frequency / distance between resonance peaksThis calculation gives a rough estimate (in 1/s).Note! The sampling frequency of the logged data depends on the sampling interval thathas been selected in the configuration of Test Signal Viewer.

2. The value of the resonance frequency should be in the range 3 to 25 (default 100). To

tune the resonance frequency, program a short back-and-forth motion of the axis atmaximum speed. The axis should not be allowed to reach full speed before decelera-tion. Use the TUNE_DF argument of the TuneServo command to adjust the resonancefrequency and examine the torque_ref signal.

Adjust the resonance frequency until the oscillations in the torque_ref signal aredamped out.

3. Enter the tuned value of the resonance frequency to the parameter Df .


127/218

6 Tuning

6.2.8. Tuning of Nominal Acceleration and Nominal Deceleration


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g



GeneralThis section details how to tune the parameters Nominal Acceleration and Nominal

Deceleration .

Parameter descripti on

The parameters Nominal Acceleration and Nominal Deceleration belongs to the type Acceleration Data and are further described in the section Acceleration Data on page 149 .

PreparationsThe following preparations must be made before performing the tuning:

If an axis has a varying moment of inertia, Nominal Acceleration and Nominal

Deceleration should be tuned with the maximum moment of inertia. For a description ofhow to specify the inertia see Specifying the inertia on page 121 .

If gravity has an influence on the axis, then Nominal Acceleration should be tuned with amotion accelerating upwards gravity. Nominal Deceleration should be tuned with astopping motion (deceleration) while moving downwards in gravity direction.

Program two test points for acceleration and two test points for deceleration with thefollowing requirements:

Velocity: choose a velocity that is approximately 50% of the maximum speed of theadditional axis.

Distance: the distance should be chosen to ensure that the axis stabilizes at the programmed velocity before deceleration starts.



128/218

6 Tuning



C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


Tunin g of Nomi nal Acceleration and Nomin al Deceleration

The procedure below details how to tune the parameters Nominal Acceleration and Nominal Deceleration .

Before beginning the tuning, observe the following warnings:

WARNING!


WARNING!

Kp can affect the torque level (torque_ref). Therefore, acceleration/deceleration tuningshould be verified again if Kp is changed more than 10%.

WARNING!

For low values of Kp (e.g. 0 (motorsused at emergency stop). Do not continue to increase acceleration/deceleration if the effecton torque_ref is small. Always verify torque_ref level at emergency stop.

Ac ti on Note

1. Use Test Signal Viewer software to record the values ofpositive/negative torque_limit and torque_ref for theaxis.

How to define these signals isdetailed in section Definingsignals in Test Signal Vieweron page 110 .

2. Use test positions for Nominal Acceleration and do thefollowing:

1. Run the motion and check the value oftorque_ref for the positive/negative torque_limit .

2. Adjust the value of Nominal Accelerationupwards or downwards in increments of 0.5 untilthe torque_ref signal shows that the axisapproaches, but does not reach, the torque limit.

3. Reduce the final value by 10% to allow forvariations in the mechanical system over aperiod of time.

3. Use test positions for Nominal Deceleration and do thefollowing:

1. Run the motion and check the value oftorque_ref for the positive/negative torque_limit .

2. Adjust the value of Nominal Decelerationupwards or downwards in increments of 0.5 untilthe torque_ref signal shows that the axisapproaches, but does not reach, the torque limit.

3. Reduce the final value by 10% to allow forvariations in the mechanical system over aperiod of time.

Continued



129/218

6 Tuning



C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


Result

The parameters Nominal Acceleration and Nominal Deceleration are now tuned. Continuewith the tuning process as described in the next section. The complete tuning procedure isdetailed in section Tuning of axes, complete procedure on page 112 .

Continued


130/218

6 Tuning

6.2.9. Final tuning of Kp, Kv and Ti


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g



GeneralThis section details how to tune the parameters Kv, Gain Speed Loop , Kp, Gain Position Loop and Ti Integration Time Speed Loop .

Parameter descriptio n

The parameters Kv, Gain Speed Loop , Kp, Gain Position Loop and Ti Integration Time Speed Loop belongs to the type Lag Control Master 0 and are further described in the section SystemParameters on page 149 .

PreparationsThe following preparations must be made before performing the final tuning:

The parameters must be initially tuned as detailed in the section Initial tuning of Kv, Kpand Ti on page 115 .

If the axis has a varying moment of inertia, Kv, Kp and Ti should be tuned with themaximum value for the moment of inertia.

Program two test points with the following requirements:

Velocity: choose a velocity that is approximately 25% of the maximum speed of theadditional axis. The speed must be low enough to guarantee that the axis does notencounter the torque limit but high enough to prevent friction from affecting the result.

Distance: choose a distance that ensures that the axis stabilizes at the programmedvelocity before deceleration starts.

RAPID programThe following RAPID program can be used during the final tuning of the parameters Kv, GainSpeed Loop , Kp, Gain Position Loop , and Ti Integration Time Speed Loop :

PROC main()

ActUnit STN1;

TuneServo STN1,1,TuneValue\Type:=TUNE_KP;

TuneServo STN1,1,TuneValue\Type:=TUNE_KV;

TuneServo STN1,1,TuneValue\Type:=TUNE_TI;

FOR i FROM 1 TO 10 DO

MoveJ p1,v_tune,fine,tool0;MoveJ p2,v_tune,fine,tool0;

MoveJ p1,v_tune,fine,tool0;

ENDFOR

DeactUnit STN1;

ENDPROC

The argument TuneValue is a percentage (1-500) of the set value in the loaded configurationfile. 100% is the normal value.

Further descriptions of the RAPID instructions and arguments are found in Technicalreference manual - RAPID Instructions, Functions and Data types .



131/218

6 Tuning



C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


Illustration, final tuning of Kp

The figure below shows the desired profiles of speed and torque_ref when making the finaltuning of the parameter Kp, Gain Position Loop .

xx0400000649

Final tuningThe procedure below details how to make the final tuning of the parameters Kv, Gain Speed

Loop , Kp, Gain Position Loop , and Ti Integration Time Speed Loop .

WARNING!The system is unstable and therefore dangerous during the tuning process, since bad

parameters or parameter combinations may be used! The safety procedures of the robotsystem must be carefully followed throughout the tuning process.

Ac ti on

1. Make the preparations detailed in the section Preparations on page 128 beforeperforming the final tuning.

2. Use the initially tuned values of the parameters Kv, Gain Speed Loop , Kp, Gain PositionLoop , and Ti Integration Time Speed Loop as default values in the type Lag ControlMaster 0 .

3. Make a final tuning of Kv by using the RAPID program on page 128 with the followingvalues:

Increase the TuneValue for Kv in steps of 5% and observe the torque_ref inTest Signal Viewer. Stop when the axis starts to vibrate/oscillate.

Divide the TuneValue by 2 and run the axis again, while observing thetorque_ref . There should be at most one or two damped oscillations after theacceleration stage. If torque_ref oscillates more than this, decrease its valuesomewhat.

Kv is a critical parameter. A large value will result in a stiff axis and a fast response. IfKv is too small, Kp will also be limited, resulting in an under-utilized axis.

Continued



132/218

6 Tuning



C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


ResultThe parameters Kv, Gain Speed Loop , Kp, Gain Position Loop , and Ti Integration Time Speed

Loop are now finally tuned.

4. Make a final tuning of Kp by using the RAPID program on page 128 and the followinginformation:

Increase the TuneValue for Kp slowly until the desired speed and torque_ref profiles are achieved as shown in the figure Illustration, final tuning of Kp onpage 129 . The position error (lag) is inversely proportional to Kp thus a largevalue for Kp is desirable.

5. Make a final tuning of Ti by using the RAPID program on page 128 with the followingvalues:

Reduce the TuneValue for Ti in steps of 5% until the effect can be seen on theplot of speed as an increased overshoot.

Increase the TuneValue for Ti by 5-10% until the effect is removed.

6. Calculate the final values of Kv, Kp and Ti by multiplying the initially tuned value in theconfiguration file (that was used as the default value) by the given TuneValue dividedby 100.

Default value * (TuneValue / 100)7. Enter these new values to the parameters Kv, Gain Speed Loop , Kp, Gain Position Loop

and Ti Integration Time Speed Loop in the type Lag Control Master 0 .

Ac ti on

Continued


133/218

6 Tuning

6.2.10. Tuning of the soft servo parameters


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


6.2.10. Tuning of the soft servo parameters

GeneralThis section details how to tune the following parameters in the type Lag Control Master 0 :

K Soft Min Factor

K Soft Max Factor

Kp/Kv Ratio Factor

Ramp Time

TIP!

In most applications these parameters do not have to be trimmed and can be left at theirdefault values.

Tuning of K Soft Min Factor The procedure below details how to make the initial tuning of the parameter K Soft MinFactor .

TIP!

The movements in this trim procedure should be done close to the point where the soft servois activated, to minimize the risk of an axis collapsing.

Tuning of K Soft Max Factor In most cases, K Soft Max Factor can be left at its default value (1.0).

If the axis is too stiff at 0% softness, reduce K Soft Max Factor . If the axis is too soft at 0%softness, increase K Soft Max Factor . The tuning can be made in a similar way as for K Soft

Min Factor , but with smaller movements.

Kp/Kv Ratio Factor

Kp/Kv Ratio Factor determines the stability margin for the axis. A value less than 1.0increases the stability. It is not possible to set this parameter to a value larger than 1.0 sincethe stability of the axis would be jeopardized.

Ramp TimeIf Ramp Time is changed, the duration of the activation and deactivation phase will change.A short ramp time can result in a twitch of the axis at activation.

Ac ti on

1. Determine a maximum axis movement for which the axis should not move, when thesoftness is 100%. Such a movement can be 0.1 rad for a rotating axis.

2. Determine a minimum axis movement for which the axis should move, when thesoftness is 100%. Such a movement can be 0.2 rad for a rotating axis.

3. Activate the soft servo with softness 100% and perform the two movements.

4. If the axis moves for both movements, the axis is too stiff and K Soft Min Factor shouldbe reduced. If the axis does not move for any movement, the axis is too soft and K SoftMin Factor should be increased.

5. Repeat step 3 and 4 until the axis does not move for the smaller movement but doesmove for the bigger movement.


134/218

6 Tuning

6.3.1. Introduction


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


6.3 Additional tuning fo r servo guns

6.3.1. Introduc tion

Ab out t un in g of s erv o gun sFor a servo gun there are some additional parameters that need tuning. These parameters

belong to types Force Master and Force Master Control .

NOTE!

This tuning must be done before setting up the torque/ force table described in Tip force on page 75 .

TIP!

To simplify the tuning the STTune instruction should be used.

Ab out t he STTune in st ru ct io n

The STTune instruction is a RAPID instruction used to modify parameter values withouthaving to do a warm start. The argument of the instruction specifies which parameter should

be tuned and the new parameter value. For more information about STTune , see Technicalreference manual - RAPID Instructions, Functions and Data types .

ExampleThis instruction will tune the parameter Ramp when Increasing Force to the value 100 Nm/s.

STTune M7C1B1_S, 100, RampTorqRefClose;

Parameter argumentBelow is a list of the values of the parameter argument used in this chapter, and which system

parameters they correlate to:

Ar gu ment v alues in STTune System parameter name

RampTorqRefClose Ramp when Increasing Force or Ramp Time

KV Kv1 to Kv6

SpeedLimit Speed Limit 1 to Speed Limit 6

CollAlarmTorq Collision Alarm Torque

CollContactPos Collision Delta Position (m)

CollisionSpeed Collision Speed (m/s)

SyncCheckOff Sync check off


135/218

6 Tuning

6.3.2. Protecting the gun


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g



GeneralTo protect the gun from too high forces two system parameters should be set. They are:

Torque Absolute Max , belonging to the type Stress Duty Cycle

Max Force Control Motor Torque , belonging to the type SG Process

PrerequisitesTo set these values one needs to know maximum allowed motor torque for the motor andmaximum allowed tip force, specified by the gun manufacturer. Some sort of force sensor isalso needed.

Preparations

Before starting, make sure that the gun is calibrated. Jog the tips to contact and calibrate theaxis.

Perform a service calibration (from the program editor on the FlexPendant, tap on Debug ,then Call Service Rou. and select ManServiceCalib ).

If you are using the option Servo Tool Control, you need to set the parameter Sync check off (this is done automatically by the ManServiceCalib if you have the option Spot Servo).Setting Sync check off will allow you to close the gun without having performed a tipcalibration. To temporarily set Sync check off , use the STTune instruction:

STTune M7C1B1_S, 1, SyncCheckOff;

If error38104: Overspeed during teach mode, appears during the calibration increase the

value of the configuration parameter teach_mode_speed_max_dsp. Change the value in theservo gun configuration file under SUPERVISION_TYPE. The parameter is not visible fromFlexPendant or RobotStudio Online.

Initial value is 0.28. Increase to about 0.45.

RAPID programCreate a RAPID program containing the STClose instruction. The argument TipForce should be set to 1N and Thickness set to the thickness of the force sensor.

tip_force:=1;

STClose M7C1B1_S, tip_force, sensor_thickness;

WaitTime 1;STOpen M7C1B1_S;

TIP!

If you have any of the RobotWare Spot options, you can replace the three instructionsSTClose , WaitTime , and STOpen with the instruction SetForce .



136/218

6 Tuning



C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


Procedure

This procedure describes how to tune the parameters Torque Absolute Max and Max ForceControl Motor Torque , used to protect the gun from too high forces.

Ac ti on

1. Start by setting the force-torque relationship to unity (See Tip force on page 75 ).Ordered force 1N should give a motor torque of 1 Nm. Note that if the gear ratio, in thetype Transmission , is positive the torque values should be negative.It is enough to define 2 force-torque relationships. In the type SG Process , set Numberof Stored Forces to 2, Tip Force 1 to 1, Tip Force 2 to 2, Motor Torque 1 to 1 and MotorTorque 2 to 2 if gear ratio is negative (-1 and -2 if gear ratio is positive).

2. In the type Stress Duty Cycle , set Torque Absolute Max to the value found in the motorspecifications.

3. Make a warm start of the controller for the new parameter values to take effect.

4. Execute the program and record the tip force. Increase the ordered tip force whilemeasuring the actual tip force until the actual tip force is close to the specified maximumtip force. For future use also find a torque value corresponding to a typical tip force usedduring welding.

5. Because the force-torque relation was set to unity the value in forcedata givingmaximum allowed tip force corresponds to the maximum allowed motor torque. Set thevalues of Torque Absolute Max and Max Force Control Motor Torque to this value.The new value of Torque Absolute Max may be a reduction from the value found earlier(see previous chapters). If that is the case, the values of Nominal Acceleration andNominal Deceleration (in the type Acceleration Data ) may have to be lowered becausethe torque available for acceleration has decreased.

Continued


137/218

6 Tuning

6.3.3. Position Interpolation


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


6.3.3. Position Interpolation

Ab ou t Referen ces BandwidthThe parameter References Bandwidth belongs to the type Force Master and decides the

bandwidth of the position references generated to reach the contact position before switchingto force control.

Procedure

Leave References Bandwidth at its default value (25Hz). If there are some problems withoscillations in torque or speed the bandwidth may be reduced.


138/218

6 Tuning

6.3.4. Tuning ramp times for force ramping


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


6.3.4. Tuning ramp t imes for force ramping

Two ways to con figu re the rampThere are two different ways to configure how fast the force is increased to its ordered value.The first method is by deciding how fast the motor torque should increase per second. Thecorresponding parameter that needs tuning is Ramp when Increasing Force . The secondmethod is by specifying how long time the ramp should take. This approach means that theramping will be steeper for high forces, something that increases the linearity between forceand torque and thereby reduces the calibration effort. The corresponding parameter in thiscase is Ramp Time . To choose between these two approaches a parameter, Use Ramp Time , isused. By setting this parameter to TRUE the second method is used.

Prerequisites

Depending on the type of gun, different parameter values might be optimal. For tuning of the parameters, a force sensor that can read the force during the whole closing procedure would be optimal. If that is not available, use Test Signal Viewer and log position (Signal no. 18),motor speed ( speed ) and motor torque ( torque_ref ).

RAPID programMake a RAPID program containing a STClose instruction with ordered force chosen so thetip force will be a force typically used. The program should also include a STTune instructionwith the argument value RampTorqRefClose . If the parameter Use Ramp Time is set toTRUE this tune instruction modifies the value of Ramp Time else it modifies Ramp when

Increasing Force .

In case Use Ramp Time is set to FALSE the program may look like the one below. To be surethat the speed limitation does not affect the result, include a tune instruction setting the speedlimit to a high value.

STTune M7C1B1_S, 200, SpeedLimit;

FOR i FROM 20 TO 220 STEP 40

STTune M7C1B1_S, i, RampTorqRefClose;

STClose M7C1B1_S, tip_force, thickness;

WaitTime 1;

STOpen M7C1B1_S;

ENDFOR

STTuneReset M7C1B1_S;In case Use ramp time is set to TRUE the values tested should be in milliseconds. Areasonable choice is to modify the FOR loop as below:

FOR i FROM 0.020 TO 0.100 STEP 0.01

TIP!

If you have any of the RobotWare Spot options, you can replace the three instructionsSTClose , WaitTime and STOpen with the instruction SetForce .



139/218

6 Tuning

6.3.4. Tuning ramp times for force ramping


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


Procedure

This procedure describes how to tune the force ramping. What we want is to have as fast risetime as possible without getting any other unwanted effects like overshoot in the force. Alarge overshoot in the force signal is usually also seen in the position and the speed signals.

Make sure the calibration of the gun and ordered contact position is correct. Incorrectcalibration will usually show as an overshoot of the force signal.

Ac ti on

1. Run the RAPID program shown above.

2. If a force sensor is available, look at the overshoot of the force signal. Choose a rampvalue that does not give a large overshoot.

3. After deciding a value of Ramp when Increasing Force or Ramp Time , let the RAPIDprogram run in continuous mode for a while and study the test signals to make sure thatno unexpected effects show.

Continued


140/218

6 Tuning

6.3.5. Speed versus force repeatability


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


6.3.5. Speed versus force repeatability

Delay RampQuite often there is a trade-off between short cycle times and good force repeatability. To get

better force repeatability there is a possibility to have a delay before the ramping starts aftercontact is reached. This delay is controlled by a parameter Delay Ramp , belonging to the typeForce Control . The value is given in seconds. To minimize the cycle time set the value of this

parameter to zero. For better force repeatability try setting it to a value between 0 and 0.05seconds.


141/218

6 Tuning

6.3.6. Tuning the speed limitation


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g



GeneralThe speed limitation is a functionality to limit the impact speed if the real thickness is smallerthan expected or the contact position just badly programmed. In those cases the limitation willmake a major improvement of force accuracy. It will have no effect for contact positions thatare accurately programmed. The parameters that need tuning are found under type Force

Master Control . More information about configuring the speed limits can be found in thedescription of Force Master Control in Technical reference manual - System parameters .

The idea is to find the maximum speed when the contact position is accurately programmed.This will then be used in the controller to reduce speed when the user has made an error indefining the contact position.

What needs to be defined in the system parameters are corresponding values of speed

limitation and gain in the speed loop (Kv) for a number of different motor torques. The parameter No. of Speed Limits gives the number of limitations used. This value can be between 1 and 6, depending on how careful the tuning is done.

ParametersThe torque values are Torque 1 -Torque 6 . If No. of Speed Limits are chosen to, for example,3 only Torque 1 - Torque 3 need values. The unit of these values are torque on the motor side(Nm). Note that the torque values may be either positive or negative depending on the gearratio.

The speed values are Speed Limit 1 - Speed Limit 6 . If No. of Speed Limits are chosen to, for

example, 3 only Speed Limit 1 - Speed Limit 3 need values. The unit of these values is speedon the motor side (rad/s). Note that the speed values may be either positive or negativedepending on the gear ratio.

The Kv values are Kv 1 - Kv 6 . If No. of Speed Limits are chosen to, for example, 3 only Kv 1 - Kv 3 need values. These values must always be positive.

RAPID program Note that the gun must be calibrated before running the program.

Create a RAPID program containing the STClose instruction. The argument TipForce should be between 1N and maximum allowed force. No plates are needed, and the argumentThickness should be set to 0.

tip_force:=1;

STClose M7C1B1_S, tip_force, 0;

WaitTime 1;

STOpen M7C1B1_S;

TIP!

If you have any of the RobotWare Spot options, you can replace the three instructionsSTClose , WaitTime and STOpen with the instruction SetForce .



142/218

6 Tuning



C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


Procedure for tuning speed limits

This procedure describes how to tune the speed limitation.

ExampleAt the ordered force 1, corresponding to motor torque equal to 1 Nm, the maximum speedwas recorded to 50 rad/s.

Set Torque 1 to 1 and Speed Limit 1 to 50. If the gear ratio was positive the speed and thetorque values will have negative sign. Kv 1 may be set to 1 while tuning the speed limits.

KvLarge Kv values will mean a larger speed reduction which will make the gun slower. Theeffect of changing the Kv values may be studied after the speed limitation is entered in thesystem parameters. This can be done by making a program containing a STClose (orSetForce ) instruction and a STTune instruction for tuning of Kv. By studying the effect ofchanging the programmed contact position and tuning the value of Kv an optimal value can

be chosen.

Ac ti on1. Set the force-torque relationship to unity. See Protecting the gun on page 133 .

2. Make sure the speed limitation does not affect the gun closing. Use the STTune instruction to set the speed limitation to a high value.

STTune M7C1B1_S, 200, SpeedLimit;

3. Make sure the gun is calibrated before running the program.

4. Run the RAPID program shown above.Check that the position shown on the teach pendant when the tips are in contact is 0.

5. Use the Test Signal Viewer and log motor speed ( speed ) during the closing of the gunat different ordered forces. For each ordered force find the maximum speed during forcecontrol.

The speed seen in Test Signal Viewer will have two maximum values. The first one isusually the largest one and corresponds to the speed when moving to contact position.The speed we need for the tuning is the second top, which corresponds to the speedwhen the gun tips are in contact.

Continued


143/218

6 Tuning

6.3.7. Improving the results for poor programming


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


6.3.7. Improving the results for poor programming

Close Position A djustThanks to the speed limitation, the force accuracy is usually good even for the cases when thetips are not in contact with the object when the force ramping starts. This will be the case ifthe object is thinner than expected or the user has programmed too large a tip distance.

The opposite case, i.e. when the object is thicker than expected is usually not as good.However, this may be improved by using the parameter Close Position Adjust , in the type SGProcess . The parameter value is set in meters and a typical value would be around 0.0005m.

Activating Close Position Adjust usually gives a large improvement for the latter case butmight have a small negative effect on the results for the first case.


144/218


145/218

6 Tuning

6.3.8. Tuning the calibration routine


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


Setting parameter values

When the calibration works satisfactorily (the gap between the tips is zero), enter the newvalues of the parameters Collision Delta Position and Collision Alarm Torque, in the typeForce Master . The new values should be equal to the values in the tune instructions.

Make a warm start and test that the calibration works. (Note, if you use the same program thenremove the lines containing the tune.)

Speed and b andwidthTuning Collision Delta Position and Collision Alarm Torque should be enough in mostnormal cases. If the calibration is not satisfactory, try to lower the values of Collision Speed(m/s) or Collision LP Bandwidth and repeat the tuning procedure.

Continued


146/218

6 Tuning

6.3.9. Force ready detection


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


6.3.9. Force ready detection

Definin g when or dered force has been reachedThe parameters Force Detection Bandwidth and Force Detection Speed , in the type Force

Master, are involved in deciding when the ordered force has been reached. They should beleft at default values. The effect of increasing the value of any of these two parameters will

be that the system believes ordered force is reached at an earlier time.


147/218

7 Error handling

7.1. Error management


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


7 Error handling


GeneralThis section details how to handle fault localization after having performed systemconfiguration.

Fault localization is done in the following steps:

1. start the Hyper Terminal to get access to the log file containing information about thecommunication between the controller and the PC

2. check the log file and search for faults

3. edit the file INT_*.cfg(the internal configuration for the kinematic model in question)

4. reload the configuration fileFor a description of how to use a Hyper Terminal see the section How to use a Hyper Terminalon page 147 .

Handlin g errors - an iterative process

Fault localization is an iterative process that must go on until all errors are eliminated. Thefollowing steps describe the order of the process:

Illustration of error messages in log fileThe figure below shows an example of the contents of the log file.

en0500002478

Description

1. Search for error messages in the log file. Identify the first error in the file. Status 0 impliesthat the file is found and the operation has passed. A negative status value implies thatan error has occurred.For an example of the status information see illustration in the section below.

2. Open the configuration file (internal or external configuration file) and correct the firsterror that was found in the log file. The illustration below shows a part of an internalconfiguration file.Note! To avoid new error messages only correct one error at a time.

3. Save the edited configuration file and load the file to the controller. Further details arein the section.

4. Make an I-start, i.e a cold start of the system.

5. Go back to step 1 and proceed with the four steps above until all errors are eliminated.



148/218

7 Error handling



C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


Examples of error messages

The table below gives some examples of common error messages that can show up in the logfile after system configuration.

SummaryError management is necessary to secure that the right configuration file for a certainkinematic model is used. It is also important to check that parameter settings have been donewith correct/allowed values.

Every time a new system configuration is done or axes are tuned, fault localization and errorcorrection must also be done to ensure that any remaining errors will be eliminated.

Error message in log file Descri ption - causeof failure

How to correct th e error

04-01-13 16:33:46 MC0:param.c 1286cfg_get_named_instance for'DRIVE_UNIT' M2DM1failed.

With regard to hardwaredrive unit position andnode position; do notexist for axis rob11_2.

Check the hardware and set thecorrect node values for theparameter use_drive_unit , inthe INT_XYZ_11.cfg file. Theparameter is described in thesection.

04-01-13 17:43:59 MC0:cfg_file.c 2142Mandatory attribute 'name'missing in line 180Status -13 from config compr:/hd0a/STANDALONE001/GantryArea/XYZ_rob11/irbcfg/INT_XYZ_11.cfg.enc

In this case, the sign forthe line breaking, "\", ismissing on line 179 in the

INT_XYZ_11.cfg file.The error messageindicates that the errorexists in some of the linesbefore line 180.

Check the lines for missing orincorrect signs.

04-01-13 17:50:57 MC0: type= ERROR id = SYSTEM code= 136arg 0: pole_pairsarg 1: 180Status -12 from config compr:/hd0a/STANDALONE001/

GantryArea/XYZ_rob11/irbcfg/INT_XYZ_11.cfg.enc

The value of pole_pairs(in this case, on line 179)in the INT_XYZ_11.cfgfile is not within thepermitted limits.The error messageindicates that the errorexists in the parameterpole_pairs , in some ofthe lines before line 180.

Change the value of theparameter pole_pairs in theINT_XYZ_11.cfg file. Theparameter and permitted limitsare described in the section

Continued


149/218

7 Error handling

7.2. How to use a Hyper Terminal


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g



GeneralThis section details how to connect the robot system to a Hyper Terminal. A Hyper Terminalis a software used to log the communication between the controller and the PC.

How to u se a Hyper Terminal

Acti on

1. Connect the PC to the controller. Use the MC/CONSOLE port in the controller and theCOM1 port in the PC.The MC/CONSOLE is further described in the section.

2. Open a Hyper Terminal application. A Microsoft Windows application may be used: Start / Programs / Accessories / Com-

munications / HyperTerminal.3. Use UNIX commands to navigate in the software.


150/218

7 Error handling



C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g



151/218

8 System Parameters

8.1. Acceleration Data


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


8 System Parameters

8.1. Acceleration Data

General

These parameters are applicable to each arm of the external robot in question.

Parameter descripti onThe parameters belong to the configuration type Acceleration Data in the Motion topic.

Cfg name Parameter name Description

name Name Name of the Acceleration Data group. Max 32characters.

wc_acc Nominal acceleration Axis acceleration in radians/s2. If the value is toohigh, the motor will reach the torque limit andresult in poor path performance.

wc_dec Nominal deceleration Axis deceleration in radians/s2. If the value is toohigh, the motor will reach the torque limit and theaxis will overshoot in fine points.


152/218

8 System Parameters

8.2. Arm


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


8.2. Arm

GeneralThese parameters are applicable to each arm of the robot in question.

Parameter descriptio nThe parameters belong to the configuration type Arm in the Motion topic.

Common parameters

Parameters for additional axes

Parameters for non ABB robots


upper_joint_bound Upper joint bound Upper bound for the axis work area (inradians or meters). The axis cannot bemoved beyond this limit during joggingor program execution.

lower_joint_bound Lower joint bound Lower bound for the axis work area (inradians or meters). The axis cannot bemoved beyond this limit during joggingor program execution.


- Independent joint Set parameter to value on in order to activate thepossibility to use independent joint instructions.Default value is off.

- Independent upper joint

bound

Upper bound for the axis work area when

operating in independent mode (in radians ormeters.

- Independent lower jointbound

Lower bound for the axis work area whenoperating in independent mode (in radians ormeters).


name Name Name of the ARM data group, e.g. x.

use_arm_type Use arm type ID name for ARM_TYPE data group.

use_acc_data Use acc data ID name for ACC_DATA data group.

use_arm_calib Use arm calib ID name for ARM_CALIB data group.

lower_joint_bound_min - Minimum value for lower_joint_bound.The unit is radian or meters.

cal_position Calibration position Calibration position. The unit is inradians or meters.


153/218

8 System Parameters

8.3. Arm Calib


C

o p y r i g

h t 2 0 0 4 - 2

0 0 7 A B B

. A l l r i g


8.3. Arm Calib

GeneralThis data group provides parameters that are applicable to each arm of the external robot inqustion.

Parameter descripti on

The following table contains the parameters that belong to the data group ARM_CALIB.

Parameters for non ABB robots

Cf g n am e Par am et er n am e Des cr ip ti on

name Name Name of the ARM_CALIB data group.

8/9/2019 Additional Axis

Documents

Additional Axis 3HAC021395-001 RevF En