Posts by Nation

    Details man. You are asking us to spec out a cell with out even answering basic questions.


    Knowing the part size only tells me what size of robot you will need.


    To repeat some of my questions:

    What is your required cycle time?

    What are the tolerances of your parts?


    Some new ones:

    What is the shot size of your glue?

    How accurate do you need to be with the volume of glue dispensed?

    What is the Reynolds number of the glue? If the glue is super thick or thin, it will require special provisions.

    What is the cure time of the glue?

    Does the glue need to go through an oven or some other post process?

    Does the glued LED strip need to be held in place while the glue cures?


    These are all questions you should be able to answer. Any integrator will be asking you the same.

    in this step i need to know, can i program both tasks with one robot?

    Yes.


    or do i need two robots?

    How many parts do you need to make per unit time? You could weld a whole car with enough tool changers, but it would take you 3 weeks to do it.


    I need to know clearly what is the difference if I do these tasks with 2 robots?

    The difference is cycle time. You will be able to make about twice as many parts per unit time. The tasks are now parallelized instead of in sequence.


    Does it help to have better accuracy or does it help to save time?

    What are your part tolerances? A micron? An inch?

    you can use your linear move if you use point representation - joint. By default when you create a linear point, it saved as Cartesian, but you can change it to joint by keeping your linear interpolation type.

    I'm not seeing this in my testing. I ran to a point with J6 at 503 degrees (j6 turn number of 1), and the robot arrived at the point with J6 @ 143 deg (j6 turn number of 0).


    Linear moves ignore wrist configuration numbers. Joint moves do not. Use a joint move to get close to your operation, then use a Linear move for the final moves.

    No, that would not work. Guaranteed crash if you LPOS a position and then try to do an INC move with it.


    I recommend dedicating a PR for incremental and using that.



    Honestly HawkMe's solution would probably be easier to use.

    It invalidates the point as a safety measure. If it left the point as is and you ran it, the robot would attempt to move that distance from wherever it was, likely resulting in a crash.


    If you punch in 100mm in X and zero everything else, the robot will move from wherever it is, to +100mm X in the active user frame. So if it started at 500mm X, it will end up at 600mm X.

    Yeah, the connector is there for if you want to keep the network online if the robot is powered off. Otherwise the card gets its power internally.


    Another question would be if you have the profinet option installed on the

    Turn off power and pull the card and make sure Fanuc actually assembled it. I had that problem and they didn't bother screwing the CP1604 daughterboard to the main board. It fell off when I pulled it out.


    Another time I had a dead board, and had to RMA it to Fanuc.

    You will need to purchase the 3DL vision software options. Otherwise the vision process simply won't be there. I'm not sure the price, but the difference between the 2D package and the 3DL package is about $7,000. How much of that is software vs hardware is something you will need to talk to Fanuc about.