Movements J and L

Quote from VHS

Is linear motion much faster than joint motion ?

Most time the same movement will be faster with joint movement than linear.

Quote from VHS

Is it correct to check it as in the photo (J 100% = L 4000 mm/sec ?) ?

You can check it in that manner but the 4000mm/s depend on the robot model. Most robots will never reach the speed of 4m/s. The robot axis need to accelerate and decelerate between the positions, so you need a big trajectory or an extrem fast robot.

Quote from VHS

My simulation ...

Which simulation, and how do you simulate?

If you use roboguide and just run the selected program via virtual TP, the timer values are not accurate.

Quote from VHS

.. while the distance to be covered is greater

What do you mean? You should compare movements between same coordinates.

I've noticed this as well. Linear moves do, sometimes, produce faster motions. Something in Fanuc's motion planner will make the arm move faster when a linear move is commanded vs a joint move.

It was years ago, but I verified it in Roboguide.

Turn on TCP traces, and then color them by speed. You will be able to find out.

The TCP makes a longer movement in joint, but the servos will most likely move the exact same distance in degrees. It's just a difference of synchronization of the servo motors.

Joint should theoretically be faster, but sometimes it isn't because Fanuc.

One cannot categorically state that linear moves are faster than joint. Like HawkME said, sometimes are faster. For example, if you are approaching singularity with a linear move, you will really be working axes 4 and 6, and their limitations will cause it to slow, so as to not damage the robot. Big reorientations can also slow a robot down. In either case, all axes have to arrive at the programmed position at the same time and the torque limits of any one of the axes will limit the speed achieved.

According to the handlingtool manual on joint speed "the actual speed of the movement is dependent on the speed of the slowest axis." I have always interpreted this to mean that the slowest axis involved in the motion will move at the percentage programmed and all other axes will move at a rate that gets them to the destination at the same time.

Another option that I have found to improve Joint movement speed is the Path Priority mode against Cycle time Priority. In theory it should work the opposite (setting robot to cycle time priority should theoritically produce faster moves) but Path Priority seemed to produce a lot faster JOINT moves.

Check SCR_GROUP use_tbcart nad use_tbjnt system variables.

Just curious since I have not touched a Fanuc in quite some time, but when did they finally bump the linear speeds up from 2000mm/s?

I believe the range of linear speed command depends on robot model.

There is also an option to use MaxSpeed instead of a number of mm/s which will determine which joint will hit a speed limit for that move and use that to calculate the actual maximum linear speed possible for that motion.