RJ3 IMAGES

[hagsy]

WE HAVE A SPARE ROBOT THAT WE PRACTISE ON AND I HAVE TOOK AN IMAGE FROM ANOTHER RJ3 AND PUT IT INTO THE SPARE RJ3 JUST TO SEE WHAT HAPPENS , ALL THE PROGRAMS AND THE TOOL FRAME READINGS ARE THE SAME AND THE MASTER COUNTS ARE THE SAME ,  THE ROBOT AND ALL THE PROGRAMS ARE WORKING BUT THE ROBOT IS NO WHERE NEAR ANY OF THE POINTS IT IS SUPPOSE TO BE WHEN I FORWARD THE PROGRAM, IS THERE SOMETHING ELSE I NEED TO DO OR CAN AN IMAGE NOT BE COPIED FROM ONE ROBOT TO ANOTHER OR MAYBE IS IT BECAUSE ONE ROBOT IS AN M6i AND THE OTHER IS A 120i

[Fabian Munoz]

Hello:
It is    "BECAUSE ONE ROBOT IS AN M6i AND THE OTHER IS A 120i"    to name one thing. The reach of each robot is different. The M6i will never reach as far as the 120i does


DON’T SHOUT
No matter what forum, writing in all capital letters is considered SHOUTING and is considered very rude. A word or two in caps is fine, but shouting is not recommended

[hagsy]

Thanks for that , and sorry for the capital letters

[hagsy]

By the way you say the reach of the robots are different which i can see but that doesn't explain why when i forward a program the robot tries to twist itself into impossible angles which have no resemblance to the program and give limit alarms for the axis

[Fabian Munoz]

When this happened to me i wasn't playing. I was actually program a production robot and I loaded a  2000i/125 on a 2000i/165.
The 125 is longer than the 165. I noticed the problem when I trying to move in world coordinates. The X and Y was completely off but the Z was actually up and down straight. I tried, like you, to run the program and it was not resemblance either

As far as you "imposible angles" comments, I dont know the algorithms that the robot is using to go to point B from A. I think we are expecting the robot to follow the same path from A to B even with the wrong arm size, but who knows.

If you keep testing, keep posting because we need to know how bad we can screw up a robot !!!!

[anonymous]

Each robot has unique mastering data.

You can restore the image, but you must load the mastering data for the specific robot or type in its unique master counts. Then, recalibrate to associate the master counts and joint angles.

[hagsy]

I copied the master counts over recalibrated but it seemed to make no difference, will keep messing with it as it is a practice robot that we have spare, any suggestions welcome.

[Napierian]

Not only are the reaches different between the robots, the gear ratio of the reducers are different.  this means if you move axis 2 or 3 90 degrees in joint, you will not physically see the arm moved 90 degrees.  This will throw out any of the calculated positions that the robot tries to move to (all points saved as cartesian data). 

You would be better off saving all the points as joint representation, and loading your original image into the robot, then loading these programs in.  By saving as joint representation, the angles are exact, not calculated based on TCP and robot type.

Good luck

[anonymous]

Napierian's latest discourse caused me to realize that the mechanical units are not the same in this situation (yikes). That means the robot libraries are not the same. The kinematic representations differ due to arm length/geometry, transformations ratios for mechanical drives, servo parameters, etc. Ja, the arm may move strangely.

Nevertheless, if you want to transfer paths between arms, Cartesian representation should be used to represent positions, not joint angles. Use joint angles for simple positioners.

[hagsy]

I think you are right having had a look at the manuals for both robots i will try your suggestion and let you know what happens.