April 25, 2019, 04:07:17 PM
Robotforum | Industrial Robots Community

 World Coordinates Error cannot move linear.

Author Topic:  World Coordinates Error cannot move linear.  (Read 1825 times)

0 Members and 1 Guest are viewing this topic.

July 29, 2018, 10:50:18 AM
Read 1825 times


my before post last massage net replay anyone.

can't find witness marks.
how to zero position master this robot and fix this problem.
mastered this robot j1 face for x direction,j2 and j5 parallel the z direction, j3 parallel the x direction.
wrist lock and jog robot mod now world coordinate is OK before the problem.but not full correct.
wrist unlock and move the robot x direction.tool is not in y and z planes same point.offset these point move x direction.example is x direction.move any other direction.same problem
move the robot in joint  mod j2 axis.why move j2 motor and j3 motor.i think cannot move j3 motor.
please help me.

Today at 04:07:17 PM
Reply #1



July 31, 2018, 02:49:31 PM
Reply #1


What robot model are you working with? Post a picture
If the marks are erased is going to be difficult to calibrate

In the calibration posture the
main arm (link 2) is vertical
upper arm (link 3) is horizontal :
at 90 degrees with link 2
link 5 is in line with link 3

J1,J6 are unimportant

What is your application???
You might no need to calibrate accurately the arm

If you use it for pick and place joint interpolation is enough

Welding sealing requires good calibration

Sent from my iPhone using Tapatalk

July 31, 2018, 04:33:08 PM
Reply #2


my robot is """FANUC ARC MATE 100IB""".

"robotero"  i try this method.but not clear linear  movement.
x and y have small curve.but z have more curve.
and can move this linear only wrist jog lock method.
unlock the wrist jog tool is moving other direction.

this robot use for welding.
i want high accuracy calibrate this robot.

please help me

thank you.
« Last Edit: July 31, 2018, 04:34:40 PM by AML1994 »

August 02, 2018, 12:14:10 AM
Reply #3


August 02, 2018, 01:40:02 AM
Reply #4


thank you your pdf.
i checked witness mark.but haven't witness mark in this robot.
please give me another solution.
master zero position or fixture position?

August 02, 2018, 05:10:30 AM
Reply #5


Your robot's zeroing signs may have disappeared. In the "JOIN" position, the zero points of the engines do not change much. However, due to not working too much over time, some motors may deviate by 1-1.2 degrees.
In my opinion, do not reset the reference point of your robot, but reset it with reference to the zero points you see in the "JOIN" position. We've worked on this before, we have tried many methods, but it's very hard to catch factory values ​​at home. We are fine-tuning our source points according to the zeroing we have done.

August 03, 2018, 12:53:53 PM
Reply #6


I've heard but never seen FANUC zero fixture tool.

As far as I know you could use it to get 0 position again.

Today at 04:07:17 PM
Reply #7



August 14, 2018, 01:38:28 PM
Reply #7


"""   robotero""" thank you.

but not high accuracy liner movement.
teach 2 point using wrist lock world coordinate mode.
running robot have small curve.
wrist unlock robot cannot move.because wrist rotate more turns.
cannot find witness mark.(add picture below.)
moved robot x axis.but robot move x and z axis.
same are y,z axis.

please help me.

August 14, 2018, 01:46:10 PM
Reply #8


teach 2 point using wrist lock world coordinate mode.
running robot have small curve.
wrist unlock robot cannot move.because wrist rotate more turns.
cannot find witness mark.(add picture below.)
for the mastered robot.how to master this robot.
moved robot x axis.but robot move x and z axis.
same are y,z axis.
how to design tool frame for this robot

please help me.

August 14, 2018, 03:38:04 PM
Reply #9


None of the bolt heads on any of my robots have paint on them.
It looks like your robot has been painted.
The witness marks probably need to be uncovered.

August 14, 2018, 04:01:50 PM
Reply #10


how to fix non linear problem.
robot always move small curve

August 15, 2018, 12:48:30 PM
Reply #11


If you have another robot of the same model, you might be able to cheat out a zero position mastering.

Take a mastered and functional robot that is identical, and use a miracle point or digital protractor to manually align all of the axis on the bad robot to the exact same angles as the good robot at 0, then zero master the bad robot. The only parts i am concerned about are J1 and J6, because i am not sure how you would be able to positively ID them. For J1, maybe drive to a hardstop and single axis master it to whatever the hard stop location is supposed to be? Not very accurate. For J6, if it is continuous rotation, maybe just pick an orientation, and it should be fine.

Alternatively, Fanuc should have a method with a special fixture which they could use to re-master a robot without witness marks.

August 15, 2018, 04:34:45 PM
Reply #12


""Robo_Eng_13"" hot to do it?
please explain.
thanks for your anser

August 15, 2018, 04:58:22 PM
Reply #13


In these instructions, Robot 1 will be your current problematic robot, and Robot 2 will be a known good robot of the exact same model as Robot 1.

1. Take Robot 2 to its zero position.

2. Use a miracle point or digital protractor to measure the base of Robot 2 to determine if it is level or tilted.
2.1 If level, no action needed.
2.2 It tilted, rotate J1 of Robot 2 so that the robot faces directly uphill or downhill (to make math easier)

3. Repeat Step 2 for Robot 1.

4. Determining your offset. Measure the slope of the base for Robot 1 and Robot 2. Whatever the difference is, remember that. You will be adding it (or subtracting it as appropriate) to your measurements.

5. Use the miracle point or digital protractor to measure the angle of the lower arm of Robot 2. Add (or subtract as necessary) your angle offset, then carefully jog J2 or Robot 1 until the miracle point or digital protractor indicates you are at the desired angle.

6. Repeat step 5 for J3.

7. J4 should be measured to level, if your robot is aimed uphill or downhill at J1.

8. Repeat step 5 for J5.

9. For J6, i would assume that you are near the middle of the range. Measure across two bolts on J6 or Robot 2, then jog the faceplate so that it matches between any two bolts.

10. Perform zero master on Robot 1.

11. Last is to try and get J1. Move J1 of robot 1 and robot 2 to the same direction hard stop.

12. Perform a single axis master on Robot 1 to the angle value of Robot 2.

This is the best estimate i could come up with, if i were in a desperate place.

Today at 04:07:17 PM
Reply #14



August 16, 2018, 02:10:36 AM
Reply #14


haven't good position robot.
how to did it?

August 16, 2018, 11:36:05 AM
Reply #15


Sounds like you are out of luck. Call Fanuc. They should have a fixture to master it without scribe marks. Once it is re-mastered, add new scribe marks, so you can recover in the future. Fanuc sells the little glue on pointers that come by default, or you can just put a clear scratch in the paint across both sides. I like the arrows.

August 19, 2018, 03:58:49 AM
Reply #16


I am sorry, I couldn´t reply early. I am busy

If your robot has calibration errors just add move intermediate point in you weld path

You can attempt calibrating the robot WITHOUT master fixture :

Is possible because there are multiple solutions to reach the same position/orientation
When the robot is correctly calibrated, the tool position/orientation will be the same when configuration is changed

But when robot is not calibrated correctly position/orientation will be different

Then you move the robot axis to calculate the correct angle
Finally, use the computed value to single axis master

Only J2, J3, J4, J5 are critical

First, roughly calibrated/master the robot
Tool Frame  is unimportant, make it zero

Then adjust the calibration starting from J5

Fix a pointer reference
Then move robot to touch the tip of the pointer
Record the point twice (using Joint interpolation)
Change the configuration of the second position (Flip to Non flip or viceversa)
Move robot to second position
If robot is not correctly calibrated, Tool tip would not return to the tip reference pointer
Take note of the programmed J5 value
Moving only J5 bring the Tool Tip as close as possible to the pointer
Read the value of J5 axis
Add both values and divide by two
Without moving the robot single axis calibrate using the computed corrected angle
Master robot

That would correct J5 axis

Try this first, then I will send instruction to correct the rest

August 20, 2018, 05:04:43 PM
Reply #17


""robotero"" cannot understand your last post.

mastered and calibrate robot  to move the witness mark.

but my linear error not fix.

teach same z axis and x axis different y axis 2 point.but robot running small curve.(have a small video clip.try to upload.but cannot upload for this reply)

normally world coordinate system move the robot z direction.but robot move the x and z direction.

that time checked xyz position value.only changing z.but WPR  values are changing.

August 21, 2018, 05:03:56 PM
Reply #18



To roughly calibrate just move the robot to the posture shown in the pdf document

Zero master the robot


Then in the same menu Calibrate the robot


To check that you correctly calibrate check the position values

Press POS button

ALL joint values should be ZERO

Check which of your TOOL Frames is NULL (All values ZERO)


Check the Tool frame definition by pressing DETAIL

Make any NULL TOOL Frame the active TOOL

SHIFT+COORD, then press the number of the NULL TOOL

Then display the position in World coordinates

The position with a NULL Tool should be

X= 150+615+100 = 865
Y = 0
Z = 600 + 100 = 700

W = 180
P = -90

In this position the wrist is in singularity (J5=0)
To jog thee robot in world coordinates, move J5 first in joint mode

Then jog in world coordinates

Follow the procedure to calibrate J5

August 22, 2018, 01:58:39 PM
Reply #19


i did this proceed-ear.
x value is 860.
i teach 2 point.
same x and z point different y position.
have same problem.move other direction small curve.add image teach point and 2 point between start and end point.

Share via facebook Share via linkedin Share via pinterest Share via reddit Share via twitter

World Coordinates error

Started by AML1994 on Fanuc Robot Forum

8 Replies
Last post July 25, 2018, 04:16:13 PM
by AML1994
converting base coordinates to world coordinates?

Started by knowledgesharing on KUKA Robot Forum

5 Replies
Last post April 28, 2014, 08:18:30 PM
by Greedy
World Coordinates

Started by Junkycosmo on Fanuc Robot Forum

3 Replies
Last post January 26, 2016, 07:08:46 PM
by pdl
World Coordinates

Started by Junkycosmo on Fanuc Robot Forum

4 Replies
Last post September 09, 2016, 01:04:49 PM
by BigFrank612