KRC4 + Rotary + Motion Track + EtherCAT --> (HALF) DISASTEEEERRR!!!

  • Hi to all!


    I've been figthing for almost one month to set up a brand new KR210 R3300 K ULTRA, KRC V8.2.19 (Kernel KS V8.2.123) and the help from KUKA has been quite useless so let's see if here I'm more lucky! :winking_face:


    I've 2 completely different errors, non related one to the other (I think), one is the impossibility to make the OMROM ETHERCAT modules to communicate properly with the KUKA EtherCAT (this robot came directly from germany to my company with Ethercat), the other one, being able to set up properly the motion track that after many test I made it move for 4 cm, but then I've got an error.


    So let's start with the details:


    1.- I'm using a GRT1-ECT device from OMRON that is a EtherCAT SmartSlice, just an I/O device with some DIGITAL I/O modules. I've got the XML ESI file from OMRON, but or WorkVisual (2.4.3 build 78) is not recognizing it properly or the ESI file is not defined following the standard (that I doubt b'cos the device is etherCAT certified...) the thing is that on WorkVisual I cannot deactivate the analog inputs/outpus of the device, neither specify how many input/output bytes it needs (I'm using 2x8 output + 2x8 input modules). So the Device fails to initialize, the exact error is:


    KSS13018
    <SYS-X44> Error during ECat stack initialization. Device: GRT1-ECT
    EtherCAT Communication Unit for Sm [Init: init command failed -
    PREOP to SAFEO]


    I'm having the SAME error when I try to communicate it with the 3G3AX_MX2_ECT, that is an EtherCAT communication device for an MX2 Spindle from OMRON. I try not to connect both of them at the same time to try to isolate the problem... but nothing... In the last installation I had to put a DEVICE NET gateway to be able to control both devices, BUT I don't want to do it b'cos then if devNet fails at any time you must restart manually the I/O driver of the robot, so the user must become Admin, etc etc etc...


    Any Idea/Suggestion? I'm quite pissed with both KUKA and OMRON, they just blame each other...




    2nd PROBLEM, the motion track.


    I'm using a home made motion track based on a Mx_260_KSP40_00-136-892/00-119-767_V01 engine from KUKA, using a 190/7 reduction ratio. The only way to set properly the Geometry was using the WorkVisual 2.4.3 build 78, with the previous build 55 there was no flange inside the Geometry for this engine...


    The robot orientation is the same than the motion track orientation, both +X are aligned, so I think that no transformation is needed, and as I just want it to move!!!


    The machine.dat inside R1/MADA looks like this:


    [...]
    DECL EX_KIN $EX_KIN={ET1 #EASYS,ET2 #ERSYS,ET3 #NONE,ET4 #NONE,ET5 #NONE,ET6 #NONE} ;EXTERNE KINEMATIKEN #NONE,#EASYS,#EBSYS,#ECSYS,#EDSYS,#EESYS,#EFSYS,#ERSYS
    [...]
    DECL ET_AX $ET2_AX={TR_A1 #E2,TR_A2 #NONE,TR_A3 #NONE} ;EXTERNE ACHSEN #NONE, #E1, #E2, #E3, #E4, #E5, #E6
    CHAR $ET2_NAME[20] ;NAME DER TRANSFORMATION ET2 MAX. 20 ZEICHEN
    $ET2_NAME[]="Trac lineal"
    FRAME $ET2_TA1KR={X 0.0,Y 0.0,Z 0.0,A 0.0,B 0.0,C 0.0} ;FRAME ZWISCHEN A1 UND FUSSPUNKT DER KIN IN TRAFO ET2
    FRAME $ET2_TA2A1={X 0.0,Y 0.0,Z 0.0,A 0.0,B 0.0,C 0.0} ;ZWISCHEN A2 UND A1
    FRAME $ET2_TA3A2={X 0.0,Y 0.0,Z 0.0,A 0.0,B 0.0,C 0.0} ;ZWISCHEN A3 UND A2
    FRAME $ET2_TFLA3={X 0.0,Y 0.0,Z 0.0,A 0.0,B 0.0,C 0.0} ;ZWISCHEN FL UND A3
    FRAME $ET2_TPINFL={X 0.0,Y 0.0,Z 0.0,A 0.0,B 0.0,C 0.0} ;ZWISCHEN MESSPUNKT UND FL
    [...]


    $ET1 is a rotary that is working fine, so I'm not posting more info on that one.


    After many test, I make the E2 to move, BUT after 1 or 2 seconds (movement smaller that 4cm) the robot stops and I got this error :


    KSS12032
    Robot Controller <-> safety controller, Cart. position deviation


    To be able to clean this error I have to unmaster and master again the E2 axis...


    Again, from KUKA the response is being VERY slow and useless... And as they don't give a manual of detailed list of errors, I'm feeling quite alone in this moment.


    Any idea of what can be happening?


    I'll appreciate any idea!! Thanks in advance! And sorry for my long post, maybe I should put it in 2 different post...

  • Hi Vergeithor,


    probably another useless remark on behalf of KUKA on your second topic (configuration of the external kinematic):


    Did you update update your safety configuration after creating the external kinametics config? Note you do not need to configure the external kinematic ($ETx == #NONE) if you just want to drive a standalone axis.


    Did you try to switch external kinematics ET1 and ET2 ? As far as i remember there was/is a restriction for the KL (=ERSYS) always being the first external kinematic.


    Fubini

  • Ciao Fubini,


    Thanks for the try! :winking_face:


    Yes, I did update the safety configuration, in fact every time that I activate a new configuration from WorkVisual I have to manually do it, and there's not a single message / error on the pendant before moving the robot, so I think everything is ok.


    The ET1 is a rotary that must be interpolated with the robot, so I'm not sure if this is considered an standalone axis, or I just do not understand what do you mean that I not have to configure the external kinematic...


    I'll give a try to exchanging ET1 and ET2, it worth the effort, sounds even plausible or logical :winking_face: :winking_face:


    Thanks!!! I'll keep you updated!

    &quot;¿Funciona? NO LO TOQUES!&quot;

  • Hi again!


    Unfortunately swapping the E1&E2 axis didn't work, I still have the same error... I can move the rotary without any problem (now E2) but once I move the linear (E1) it gives the error...


    Thanks anyway

    &quot;¿Funciona? NO LO TOQUES!&quot;

  • I could be mistaken, but I think if you have your robot sitting on the track, even if they are aligned there is a transformation, typically the "B". This is because the flange of the robot is not pointing up. On mine I believe it is B-90, as my robot is on the track backwards. However, if this is wrong on a krc2 it will not give an error, it will just behave badly;)

  • Hello wes_mcgee!


    You are absolutely right! I did not post yesterday but I was checking that, I'm attaching a document that explains better the transformations, I hope that it is useful for somebody! (the document is Aus Deutch but I think that it's very easy to understand).


    BUT, as the robot is not moving properly yet, the transformation is not important at all at this moment, the error is so weird.


    As an update I can say that I UPDATED the robot KSS to 8.2.20 and nothing, same problem with both, the linear track and the etherCAT ISSUE.


    ----------------------------------------
    ----------------------------------------


    I did find something even more weird, so I hope that this will illuminate anybody; I'm explaining STEP BY STEP:


    1st- -Imagine that we start with the error on the screen:


    KSS12032
    Robot Controller <-> safety controller, Cart. position deviation


    2.- To be able to "confirm" this message I've to "unmaster" the axis E2 and "master" it again.


    3.- Now I move the robot, and as we just master, the E2 is in 0,0, and so the Cartesian Value of X ($tool = nullframe, $Base = nullframe) b'cos I put the robot just in X=0.
    So after some test I realize that I CAN move the E2 from -19.99 to 19.99, almost 4 cm. If I move the E2 further than this "window" the had the error.


    4.- AFTER confirming the error mastering E2, I decide to put a value in Z of the ET2_TA1KR, so for testing I put the real high, 1640


    5.- SURPRISE, the ERROR APPEARS WHITHOUT MOVING THE E2, and is NOT possible to confirm it, not EVEN mastering again.... ¿¿¿¿¿????


    6.- I reduce the Z value of ET2_TA1KR to 10 and start again AND SURPRISE, I can move the E2 BUT my window has reduced from -19.99 <-> 19.99 TO -17.5 <-> 17.5 !!!! If I ,move the robot outside this windows I get the error.

    I change the Z value now to 19, and my window is reduced to -6.2 <-> 6.2 !!!


    I changed the Z Value to 20, and the error and its acknoledge are in the message window appearing and dissapering continuosly... obviously, is in the limit of the inital 19.99 and is getting crazy...


    I changed the Z to -10, and the result is the same than with Z = 10; -17.5 <-> 17.5


    -------------------------------------------
    ------------------------------------------


    That's all folks!


    Thanks for any idea mates!

  • Hi,


    I'll try again:
    For me your problem still sounds that the safety controller is not considerung your E2 axis which is contained in the external kinematic ET2. Furthermore your external kinematic is of robroot-type (ERSYS), that is in your case the robot is mounted on E2. Is this what you intended?
    Message KSS12032 "Robot Controller <-> safety controller, Cart. position deviation" is set if the standard controller has a different cartesian position (only x,y,z no orientation considered) than the safety controller. This check is doen because the standard controller and safety controller use different forward kinematics functions. Hence a consistency check is done. The maximal allowed deviation between standard and safety controller is as far as I know 20 mm (which fits perfectly to your +-19,99 mm deviation). By remastering your external axis you reset its axis value to zero and the forward kinematics equation of ET1 only gives $NULLFRAME as result, so basically you get the same result as if there is no external kinematic.


    To check this theory please check under Configuration->Safety Configuration->Machine Data (View) -> check machine date. If you get a message "the machine data of the safety configuration are up to date" please take a look at the listed machine data. If the safety controller integrated your external kinematic correctly the value "Axis Number for ROBROOT kinematic 1" should read "8", meaning the 8-th system axis (6 robot axes + 2 external axis) is used inside the safe controllers forward external kinemtics equation.


    Fubini

  • Hi again Fubini!


    It's very interesting what you say about the two different ways to calculate the EK that the robot has, it also has a lot of sense, BUT, as you can see on the pictures, EVERYTHING looks ok! (the only problem is that is in that beautiful language that is the German..., they do not even translated the Safety Configuration... thanks KUKA)


    I continue investigating...

  • Furbini, I think I got it,


    You were right, the information on the Safety configuration is not the same that in the MAchine.dat, EVEN if the safety controller say that (picture 02)! Inside the Safety controller I discovered that EVEN if he has two external axis (picture n3), the picture n4 shows how there's not ANY external axis assigned to the EK2, Externe Achsen für externe Kinematik 2 is NONE... shouldn't it be E2???


    Now I just need to be able to modify that Safety configuration to make it match the Machine.dat

  • You are lucky I'am native german: Take a look at your third attached picture. Here you see "External kinematic type" has value "EASYS" (meaning external base kinematic). This value should be "ERSYS" for your external robroot kinematic. This basically brings me back to my first post:


    Quote

    Did you try to switch external kinematics ET1 and ET2 ? As far as i remember there was/is a restriction for the KL (=ERSYS) always being the first external kinematic.


    Did you change your E2-axis to E1 and did you change your ET2 kinematic to kinematic ET1?


    Fubini

  • Quote

    Now I just need to be able to modify that Safety configuration to make it match the Machine.dat


    Sadly you can not modify the safe configuration, because it is generated automatically from your $machine.dat. Your only chance is modifing your $machine.dat to a configuration which generates a correct safety config. Hence still my advice: Try making E2 to E1 and external kinematic ET2 to ET1.


    Fubini

  • I did replace E1 with E2 two days ago, and changed the EK1 and EK2 with no results, exactly the same error, so I put it back to the original configuration E1 Rotary, E2 Linear, what is more, I changed it directly from the Wizard "Configuration -> external axis" and I also try from the dissatrous WorkVisual, but nothing, the same error...

    &quot;¿Funciona? NO LO TOQUES!&quot;

  • Hi Vergeithor,


    My Machine.dat is that:



    Is a linear motion track.


    I Hope this help you.

  • Hi again!


    OK, the problem is partially SOLVED! The linear track is already moving YEEEEES, but the EtherCAT issue is still there.


    The problem is a combination of things, mainly that the WoV is not fully operative and it has some bugs still to correct, and secondly that the controller hangs almost every time that you have to make a "reconfiguration" so some information is not properly activated on the Safety Controller. And of course, as Fubini pointed from the first moment, the Linear Track, MUST BE the ET1 transformation, even if the WoV says the opposite! (thanks Furbini, your help has been the KEY to solve it).


    My solution:


    1.- Create the cell project using the WoV 2.4.8 build 78, (any other is not working properly with external axis that are not inside KUKA devices like their own linear tracks or MGU's)
    2.- Once the project is loaded and activated on the robot ET1 & ET2 have to be MANUALLY corrected using the "Configuration -> External axis" wizard (or modifying the whole machine.dat manually, as they are mixed up).
    3.- After Reconfiguring and rebooting the system due to the hangs (and even if there's NOT ANY ERROR about the data in the safety controller) you have to become Safety Manager and enter the "Configuration -> safety configuration" and check and confirm the variations, and then Reconfigure again the controller (and probably restart again due to the gang).


    The most annoying thing off all the history is that the official response from KUKA up to 13:15 of today was to set properly the values of the reduction gear box, and set the B value of the $EKx_TA1KR to 90...


    Thanks to everybody... BUT... I still have the problem of the etherCAT!!! hehehehehe


    Any idea????

    &quot;¿Funciona? NO LO TOQUES!&quot;

  • Hello the Leg!


    I'm sourprised that you ask me for the EK1100 (beckoff EtherCAT Gateway) b'cos I did not mention it directly in my post, hehehe, but UNFORTUNATELY due to the poor capacity of both OMRON and KUKA to make their devices to comunicate properly with each other, I HAD to install again this EK1100 + EL6752 and comunicate all the devices via DevNet with the inconvenience that as I don't have any PCL controling the bus, everytime there's any problem with the devNET I HAVE to "reconfigure" the I/O buses, so the user has to become admin, etc, etc, etc...


    And now, answering your question:


    EK1100 has 4 state leds, the 2 upper are for +24v (left one for power and right one for bus) both MUST BE FIXED ON. The two leds under these, the left one is LINK-ACT led, this mast blink at bus speed, you can see it blinking about 6-10 times per second (6 - 10 Hz) and the right one is RUN led, this must be FIXED ON.


    I've quite experience with this devices, so if you need any info just ask :winking_face:

    &quot;¿Funciona? NO LO TOQUES!&quot;

Advertising from our partners