Hi everyone!
I am back again to ask community about the CALIBRATION of an IRB2400L M98 S4C robot. My used robot came to me with a replaced 2nd axis's motor, so I assume the calibration is wrong.
What I did:
I made a straight pin tool and attached it to a robot.
I made an another pin and fixed it on a table.
I defined a tool using a 4 points method, by moving a tool around a fixed pin.
After that i placed a 2 meter ruler on a table and moved a tool along it with a MOVEL command.
The path deviation on a travel of about 1.5m was around 10 mm. Hmm. I would like it to be 1 mm max or even smaller.
So, I tried to define a calibration offset for a 2nd axis by taking to account a TCP definition error: I changed calibration offset value a bit, defined a tool, checked the definition error, repeated these steps until the error became minimal.
After these steps I got a line travel deviation of about 5 mm. I think this method worked somehow...
But I would like the robot to be much more accurate.
In addition, I think there was not only 2nd axis calibration error, because when I reorient a tool around a fixed point, the tip of a tool deviates 10 mm or more.
Q1: In theory, may this method be used to get calibration values or am I wrong?
Q2: Can I find anywhere information about IRB2400L M98 robot links dimensions to make reverse kinematics calculations?
Q3: Can I get the precise joint rotation values from the controller? With the precision comparable to calibration offset values.
What else I want to do:
The calibration procedure is described in a manual. It is performed by use of a calibration pendulum (or earlier by wyler's levelmeter 2000). I looked at Wyler's sensors characteristics, and they seem not to be of super high accuracy. I think I could make my own pendulum using Murata inclinometers... The main problem is I do not know the calibration positions to put inclinometers on a robot. I found two calibration surfaces on a base and on a 2ndf axis arm covered by steel plates. Think they should be perpendicular each other. I have no idea how to mount a sensor to the 3, 4 and 5 axises. There are some sensor holders described in a manual, but I cannot find any flat and accurate surface on a robot to mount them and have no idea how the holders should exactly look like.
Q4: Can anyone help me about sensor holders and mounting positions for 345 axises?
Q5: Maybe a simple 0.02mm/m level should be enougth to calibrate a robot? Of course, I will first align the base of a robot to a level, after that will put a level onto a calibration plate.