Posts by stefanristevski

hey guys,

i hava a huge problem, and it is very very urgent.

I have been trying to configure KUKA KRC4 extended with 7 external motors in simulation with 5 axes in KRC4, 4 motors, 2 by two are in master slave configuration.
Catalog element for master-slave configuration was sent to us by KUKA.

My problem, after testing 5 or 6 configurations, suddenly the controller started displaying the message "no more user memory available".
Now i cannot select the sub, nor can i select any program, also i cannot activate any new configuration.

any suggestions for solution??

i succeeded, it worked nicely.

thank you panic mode. i will try and post results on the forum.

Hey guys,

i have a question.
is it possible to program kuka kr360 with krc4 kss8.3 with soft coordinates.
What i mean is, is it possible to teach an end point, but to give some tolerances i.e. if kuka senses some external force or additional load to move in opposite direction within some given limits.

We need this to teach kuka precise pick and place coordinates where attached tool is very rigid and high precision is needed.

Thank you.

robot studio does not recognize P parameters like:
N10 P1=10
N20 G01 X 100+P1 F1000
Also it does not recognize M functions, which are user defined.
Further it does not recognize external variables, which are crucial for our application.

does anybody know any alternative to solve the problems i am having.

from the web search that i did, best that i could understand is that ABB has a translator from G-code to RAPID.
are you talking about that? if so, does this work with external variables and miscellaneous functions in G-code like M-functions.

hey guys,

does any one know if there exists an option like KUKA.cnc but with ABB robot.

thanks

can i have some more information

Hey guys

i have some questions about new KUKA.simPRO software.
We recently acquired a trial of the new version of KUKA.simPRO, version 3.

Install went well, trial activation was ok, however whenever we insert a robot in a new project,
visualization does not show anything, on the side bar though robot properties are visible.

Another problem that exists is whenever we right click on the visualization area and left click on "center in 3D view" the program crashes.
Same thing happens when Drawing tab is clicked.

i shot an email to KUKA, still no answer.

any advice, or does anybody know why the program crashes?

in the attached pdf file there are pictures of the described problem above.

thanks

hey guys,

i am conducting research for our company, we are looking to buy an industrial robot solution with CNC.
I tried the web, but I could not find any information about COMAU robots and CNC.
So is it possible to run G-code with COMAU robots.

Thanks

yeah I tried, I did not find any info there.

Hey guys,

I need technical info about air hose connector on KUKA KR360 located at the flange.
Picture in the attachment.

Thanks for the explanations.

I did perform mastering on A2 with a dial gauge, and orthogonality has improved.

I have some additional questions, though.

While performing mastering with a dial gauge, [bold]start-up->master->dial gauge[/bold], assigned value to axis_mot differs from the value assigned at the exact same position with [bold]start-up->master->reference[/bold].
Is this difference in the motor position the one stored in mames?

And one more thing, mastering A2 was obvious, since was unmastered when the robot got here, however, do I have to perform mastering on the remaining axes?
Can A6 be mastered with a dial gauge?

thanks guys

Quote from Fubini

Calculate $MAMES_ACT[] - $MAMES[] for an axis and you will see the offset.

Sorry bit confused, offset from what? the actual mastering position?
So as i understang this offset should be zero or very close to zero for AA robot?

Quote

Look into the RDC-Folder you will see a *.cal File

I was not able to find the *.cal file.
I did found it though after extracting RDC data from robot info.
However there was not any information in it.

Quote

Or look into C:\KRC\Roboter\Log. You should find a file called Mastery.log.

I could not find this file as well.

majorRob

Thank you SkyeFire,

I have done some experiments with mastering A2 axis with by offsetting its mastering from the reference position, and clearly it requires precision in mastering.
We are buying a mastering device.

One question though, is it possible to read factory mastering positions of the motors?
And what is actually stored in .mam file.

regards

Hey skyeFire
thanks for the lengthy reply.

Quote from SkyeFire

Has this robot been through First Mastering (no load) followed by the Set Offset Mastering (with the measuring-tool load)? This is one of the first things you need to do with an AA KUKAbot before you ever start using it.

Except for mastering of A2 axis, done visually, rest of the robot axes are not mastered.

Quote from SkyeFire

Is the AA model active? There is a system variable to turn it off -- if the model is deactivated, you should see a constant warning message on the Smartpad that never goes away. Also, you can check under HELP>INFO and see if the phrase "Positionally Accurate Robot" appears.

Yes i checked, the AA is active.

Quote from SkyeFire

What is the value of $LOAD while you are performing these tests? Your measuring device is presumably very very light, but the "default" load that KUKAbots assume when no LOAD_DATA values have been entered will be their factory maximum, worst-case payload. This is to de-rate the robot so that it never exceeds its speed/accel limits if someone neglects to put in accurate load data.
In your case, your measurement device is probably so light, compared to the robot's payload rating, that you can simply set the Mass value and ignore the CoG values (although if you can calculate those reliably, go ahead and add them -- it can't hurt and might help).
I should also point out that the LOAD_DATA array is linked to the TOOL_DATA array by their array index numbers, and both are linked the same way to the "Set Offset Mastering" -- any AA KUKAbot needs to have the SOM done for each tool payload it is going to carry, and the LOAD_DATA values for each of those tools must be correct before performing the SOM. Then, when running a program, the $ACT_TOOL value and $LOAD value must be set to match the tool that is currently on the robot.

I did not enter any load data, so is maximum. I will try to enter the mass at least and run the tests again.

Quote from SkyeFire

As to your specific test: you say that the marble block has been set Orthogonal to the robot, but you don't mention how this is done, or to what part of the robot -- are you working in $WORLD, or are you creating a robot Base that is matched to the angles of the marble block? And if the latter, how?

Marble block has a triangular shape, it is a right triangle and sides forming the right angle are aligned with robots X and Y in $World coordinate system.
How: Attaching a dial indicator on the flange, then running a program in KUKA.cnc where only X changes and then only Y changes such as:
(for X)
G01 G90 X862 Y1219 Z593 A0 B0 C180 F1000
G01 G90 X1062 Y1219 Z593 A0 B0 C180 F1000
G01 G90 X862 Y1219 Z593 A0 B0 C180 F1000
then adjusting the marble block until the indicator shows deviation less than 0.1mm. This is done in XY plane moving in both X and Y direction thus making plane of the marble (top plane) to coincide with XY-plane of $World.
This is OK. Deviations are within 0.1mm.
Same procedure follows in YZ-plane and adjusting the marble block until both planes coincide with each other.

And then when same program is ran in and the dial indicator touches the XZ-plane (i.e. the other side of the marble block) the deviation is few millimeters. Even with incorrect mastering, can an error this big occur?

I should mention that the robot always runs the same program only the dial indicator is adjusted to touch on different planes.

Quote from SkyeFire

When you run this test, you are executing a program, rather than jogging the robot, correct? The AA model is not active when the robot is being jogged manually. Also, how many point are you measuring along each axis? AA improves the accuracy of the end point of a motion, but does not improve the motion between points. So AA will improve the spatial accuracy of points P1 and P2, but will not improve the linearity of a LIN motion programmed between P1 and P2. I would suggest breaking your long linear moves into small segments, and take measurements at each point along the plane. This will help determine if you just have a problem with the specific endpoint you chose, or something more serious.

as i mentioned above i am running a program in KUKA.cnc. Although i tried running a program in KRC as well. I taught the first point (staring point) and then through WoV i added one more point by changing only the X coordinate, such that i get motion along X-axis only.

There is no tool attached on the robot, so no load data is entered.
However in order to switch to CNC a tool is defined identical to robot's null tool, i.e. XYZABC={0,0,0,0,0,0}.
The base is also defined to coincide with robot's World coordinate system.

Quote from animisiewaz

Look at the differences between G-Code and KRL. Deviation is few only microns, or less

This is what I expected as well, however I do not get this result.

Quote from animisiewaz

Analyse movement of axes during test. Maybe one of axes is changing direction of rotation druring LIN motion and then deviation of position appears? Try to move marble block in other position of robot workspace and repeat test again - probably results will differ

I did do this. There is no change of direction of rotation in any of the axes.
And you are right if the marble block is placed elsewhere results might change, however I do not expect difference in millimeters, that is not acceptable.

And I do get linear motion with deviation of approximately 0,1mm, but orthogonality is messed up. Moving in Y is OK, moving in X is not OK.
so I am wondering I something has happened to the robot while transport?
Or its calibration is messed up?

Thanks animisiewaz

hey guys,

i have an absolute accuracy question. We recently acquired KUKA KR420R3080 with KRC4 extended. The robot has the absolute accuracy sticker on it.
In the commissioning process we measured straightness and orthogonality of the robot, although the straightness is within 0.1mm deviations, orthogonality is terrible.

Test setup:
We used marble block whose orthogonality according to standards, see pics. One edge of the triangular block was aligned with World-X of the robot and the other edge with the World-Y axis of the robot.
Using dial indicator deviations were measured. The indicator was attached on the flange of the robot.
First on the top of the block we moved the robot +-200mm in both X and Y direction (block's top surface is coincident with World-XY plane of the robot). Deviations in this test in World-Z direction were within limits of 0.1mm which is to be expected, I think. Then we moved the robot on one side of the block, the side coinciding with the World-YZ plane along World-Y direction. In this test as well the deviation in were within 0.1mm. Lastly to check orthogonality we moved the robot in World-XZ plane in World-X direction. In this measurement though the deviation is within few millimeters.

see pictures of the marble block.

Since we are planning to use KUKA.cnc, program was created in KUKA.cnc.
cnc code:
NC code

(Along X direction)
G01 G90 X862 Y1219 Z593 A0 B0 C180 F1000
G01 G90 X1062 Y1219 Z593 A0 B0 C180 F1000
G01 G90 X862 Y1219 Z593 A0 B0 C180 F1000

(Along y direction)
G01 G90 X862 Y1219 Z593 A0 B0 C180 F1000
G01 G90 X1062 Y1219 Z593 A0 B0 C180 F1000
G01 G90 X862 Y1219 Z593 A0 B0 C180 F1000

For comparison a KRC program was created to repeat the same motion, dat file of the KRC program given below.
KRC .dat file Cartesian coordinates
;Along X direction
XP1={X 862.008301,Y 1219.00073,Z 592.997742,A 2.44405310E-05,B 0.000963584229,C -179.999847,S 18,T 35,E1 2.10471456E-38,E2 0.0,E3 0.0,E4 0.0,E5 0.0,E6 0.0}
XP2={X 1062.008301,Y 1219.00073,Z 592.997742,A 2.44405310E-05,B 0.000963584229,C -179.999847,S 18,T 35,E1 2.10471456E-38,E2 0.0,E3 0.0,E4 0.0,E5 0.0,E6 0.0}
XP3={X 862.008301,Y 1219.00073,Z 592.997742,A 2.44405310E-05,B 0.000963584229,C -179.999847,S 18,T 35,E1 2.10471456E-38,E2 0.0,E3 0.0,E4 0.0,E5 0.0,E6 0.0}
;Along Y direction
XP1={X 862.008301,Y 1219.00073,Z 592.997742,A 2.44405310E-05,B 0.000963584229,C -179.999847,S 18,T 35,E1 2.10471456E-38,E2 0.0,E3 0.0,E4 0.0,E5 0.0,E6 0.0}
XP2={X 862.008301,Y 1419.00073,Z 592.997742,A 2.44405310E-05,B 0.000963584229,C -179.999847,S 18,T 35,E1 2.10471456E-38,E2 0.0,E3 0.0,E4 0.0,E5 0.0,E6 0.0}
XP3={X 862.008301,Y 1219.00073,Z 592.997742,A 2.44405310E-05,B 0.000963584229,C -179.999847,S 18,T 35,E1 2.10471456E-38,E2 0.0,E3 0.0,E4 0.0,E5 0.0,E6 0.0}

Here also deviations were different when executing motion from KUKA.cnc and KRC.

Axes of ISG were configured with the axconfigurator. Am i missing some configuration steps?

Am i misunderstanding absolute accuracy?
Should the robot be recalibrated?
And is this setting factory setting or something has happed while transporting?

Thanks

Hello all robotics enthusiasts,

I have KR360R2830 with KRC4 extended, and three external axes. External axes motors are from KUKA
1. MG_8_40_45_S0 as "E1"
2. MG_480_180_30_S0 as "E2"
3. MG_480_180_30_S0 as "E3"

I am making an anti-backlash system that requires two motors to drive one exis. Motors 2 and 3 to be configured as Master-Slave kinematics.
So finally my question is how to configure master-slave in WoV.

in the current configuration in WoV mathematical coupling of motors 2 and 3 has been disabled.

By the way i'll be using KUKA.CNC, for my application, thus master-slave mode should be active in KUKA.CNC as well.

Thanks guys