"Safe device communication error KUKA Servo Pack 40A (KSP) 1" error message

  • Problem solved !

    Before replacing my faulty KSP, I had swapped it's position with another one, to assess the issue.

    When I received a replacement, I put it where I had previously put the faulty one, and that was causing the error.

    By putting the replacement KSP where the faulty one was originally, the error went away.

    All I need to do now is a mastering with my "dial gauge" mastering tool.


    Could anyone point me to a video showing how to perform this on a KRC4 ?

    So far, I have only found videos with the digital EMD...

  • all components on KSB have assigned FSoE address which is done in strict order.


    when one device is swapped out, KSS can identify which device was replaced and automatically assign the FSoE address of now removed device.


    but when more than one device is replaced at once, or when two devices from same cabinet exchange places, system is unable to automatically adjust for changes.


    one can swap two devices in same cabinet or replace two or more devices but then user must assign correct FSoE addresses using WorkVisual. this is well documented and normally is a straight forward process, but as always there can be some challenges in specific cases.

    1) read pinned topic: READ FIRST...

    2) if you have an issue with robot, post question in the correct forum section... do NOT contact me directly

    3) read 1 and 2

  • OK, thanks, I understand. I hope this thread can help someome with a similar issue.


    When I Google FSoE, I find links to an Egyptian Rock band called "Future Sound of Egypt"...

    That would make a nice addition to my acronyms list.


    Newb question again : does Workvisual have to be installed on a separate computer, or can it be installed directly on the controller's PC ?

  • FSoE = fail safe over ethercat


    That is ethercat equivalent of safety network


    FSoE / Ethercat

    ProfiSafe /ProfiNet

    CIPsafety / EthernetIP

    etc


    Googling may help find crumbs but does very little for deeper understanding.


    If you read Work Visual manual you would know. WorkVisual is application that need to be installed on separate computer. it connects to KRC via network (KLI or KSI).

    1) read pinned topic: READ FIRST...

    2) if you have an issue with robot, post question in the correct forum section... do NOT contact me directly

    3) read 1 and 2

  • All I need to do now is a mastering with my "dial gauge" mastering tool.


    Could anyone point me to a video showing how to perform this on a KRC4 ?

    So far, I have only found videos with the digital EMD...

    No video, but it's not difficult, just finicky.

    1. Align the paint marks on the axis
      1. This will put you slightly on the + side of the actual 0 position
    2. Attach the dial gauge
    3. VERY SLOWLY (usually 1% jog speed), move the axis in the - direction
    4. The 0 position will be where the dial gauge makes an instant transition from positive motion to negative motion. Physically, this is where the spring-loaded pin bottoms out in the V-Notch in the mastering artefact.
    5. Hit "Master"


    It may take a few times to get right. The IMPORTANT thing to remember is to NEVER jog the axis faster than 1% (maybe 5%) when the dial gauge is attached (this goes for the EMD too). And it works best if you only move the axis in the - direction during this process, so as to keep the backlash consistent. If you miss it, just jog + until you're on the + side of the notch again, then try again.

  • No video, but it's not difficult, just finicky.

    Hey thanks SkyeFire !

    If you read Work Visual manual you would know. WorkVisual is application that need to be installed on separate computer. it connects to KRC via network (KLI or KSI).

    I naïvely thought that all the useful software to setup the robot would be on the robot controller's PC.

    OK, so I really need to power a switch in my cabinet, as mentioned in another thread I started about connecting a network to the controller PC.

    Edited 2 times, last by Wall-E ().

  • that should not stop you from using WorkVisual, you can temporarily unplug any cable that connects to KLI, connect your computer with WorkVisual there. when finished, put the original cable back.


    and if you run KSS8.3 or later, you can skip KLI and simply use KSI. make sure your laptop is set to obtain IP address automatically and connect it to KSI port on the KRC.

    1) read pinned topic: READ FIRST...

    2) if you have an issue with robot, post question in the correct forum section... do NOT contact me directly

    3) read 1 and 2

  • No video, but it's not difficult, just finicky.

    I calibrated Axis 1, and the controller informed me that there were no other axii to calibrate, so I guess it was the only one out of whack.

    Then, I tried to run axis 1 through it's complete angular range and I hit the physical stop on the extreme "+" side : emergency stop caused by over-current in motor 1.

    I suppose my mastering is bad, but man ! If there is so little margin between software and physical stops, it is going to be really difficult to tell when the mastering is "acceptable"...

  • Are soft limits set correctly?

    Are you sure they are not tampered with?

    Are you sure you mastered at correct point? Did you use instructions correctly?

    The pre-mastering position is 1-3 deg in the positive direction from actual mastering position. So you need to look for first dip as slowly moving in the negative direction. Then try finding dead center of the dip. If done incorrectly, it is possible that you are not at the center of the cartridge but on one of the sides, which is some 10mm away. That difference may be all that is needed to get past soft limit and to hard stop .

    1) read pinned topic: READ FIRST...

    2) if you have an issue with robot, post question in the correct forum section... do NOT contact me directly

    3) read 1 and 2

  • What are $SOFTP_END[1] and $SOFTN_END[1] set to? On Axis 1, for most KUKA robots, the $SOFT limits are +/- 185deg by default. And the rubber stop in Axis 1 will usually let you get another 3deg or so before starting to cause torque issues.


    If you jog A1 to 0 deg, which way is the arm facing? For most KUKAbots, A1=0 has the arm pointing straight "forward" (where the "back" is defined by the box where the cables attach to the robot base).


    Try doing a Linear jog along a straight line or edge you can reach inside the cell. If A1 is badly mastered, a Linear motion should show a definite curve over distance.

  • The soft limits for A1 were set at +/- 185deg.

    I really did my best to do this calibration, and the pointy thing was really in the center of the "V"...

    I'll try a linear jog as SkyeFire suggested...

  • not all robots can do +/-185deg on A1. check if MADA and mechanical stop are correct... wrong or tampered MADA (wrong $MAMES[1] or gear ratio for example) can be the cause.

    1) read pinned topic: READ FIRST...

    2) if you have an issue with robot, post question in the correct forum section... do NOT contact me directly

    3) read 1 and 2

  • Thanks folks !

    I re-mastered A1, but this time, sticking my tongue out, and I am now able to reach both software limits !

    I guess there was also a clump of old grease in the way of the sliding plastic block...

    I tried to open the MADA (by the way, what does that one stand for ?) file to see the values of the parameters you suggested to check, but it wouldn't open on the SmartPad... I was logged in Expert mode.


    Oh... in KRC4, how can I force the HMI to stay logged in expert mode ?

    I know it's a stretch to call myself an expert, but since I'm alone in my workshop...

  • MADA = MAchine DAta


    two sets exist - one is robot arm specific (KRC:\R1\MADA) and the other is for controller (KRC:\STEU\MADA).


    so robot parameter and axis workspaces are in the first one.

    cartesian workspaces and output mapping for all workspaces is in the second one.


    to remain logged in, change least time and Announcement time from 300 seconds to zero in Authentication.config

    1) read pinned topic: READ FIRST...

    2) if you have an issue with robot, post question in the correct forum section... do NOT contact me directly

    3) read 1 and 2

  • MADA = MAchine DAta


    two sets exist - one is robot arm specific (KRC:\R1\MADA) and the other is for controller (KRC:\STEU\MADA).

    R1 being "Robot 1" -- I think, long ago, there was a plan to have one cabinet drive more than one robot, but that never happened. Instead, KUKA developed RoboTeam.


    And "STEU" is, I think, an abbreviation of a German word that means "Controller" or "Cabinet".


    So, the /R1/MADA contains files for the physical arm -- motor types, arm lengths, joint relationships, and joint limits. /STEU/MADA contains controller-specific settings, like system I/O signals.


    The file Authentication.config is located in C:/KRC/SmartHMI. It's an XML file. As Panic says, you want to find the LeaseTime settings and change them to 0. LeaseTime is the number of seconds of "idle time" (aka, no one touching the user interface) before the controller automatically switches back to the lowest User Group, with 0 being "never". I forget exactly what AnnouncementTime does, but I've always set it to 0 as well, when making an "always expert" robot.

Advertising from our partners