Last week I went to a customer plant to solve a problem they had. They use 1 IRB340 with IRC5 controller and PickMaster 3.10.
Their problem was the Next:
When PickMaster sent to the controller the Pick Position the angle was in the wrong direction. For example if the item had an angle of 45? the robot went to -45?.
When I saw the problem, I verified all things: Vision calibration, conveyor calibration, etc… and the only thing I think it was different is the robot x direction is in the negative x conveyor direction.
I solved the problem by RAPID receiving the PickTarget angle and then making the movements with a Reltool where Rz:=2?PickTargetangle;
The problem is I don’t know why the pickmasters works wrong. Is a usual problem or is a unusual?? somebody can help me?? I must explain the good solution to the customer .
Did you verify that you calibrated the camera in the same direction as the conveyor, that is, not in the same way as the robot since it is rotated in the wrong direction.
Sorry, Where I can find the tool coordinate system? I verified the toolData rotation and it was like the PickMaster standard program, I suppose that is ok.
Last week I went to another PickMaster line and I found the same problem. And for my surprise the robot was placed like the other case, the robot x direction is in the x negative conveyor direction. (conveyor with vision).
I’m going always to solve the customer problem without time to find the solution and I only can modify the RAPID code to change the 4axis rotation, can somebody explain me what is happening? Sorry I can’t send any backup file for the moment because I don’t have it.
I am having what sounds to be a similar problem. I have two IRC5 controllers w/340’s tried both 5.6 and 5.7 baseware just to see if that would help. Not sure what the reason for the problem is but with some coding like Toni I could temporarily overcome it. From the cameras point of view 0 deg would be in the conv xpos direction with positive rotation counterclockwise but the pick rotation 0 deg point seems to be in what would be the neg Y dir of the conv and positive rotation is clockwise. As a part is seen say at 10degs the pick is significantly off, but repeatable. I drew a diagram of zero from the cameras perspective and then how it is used at the pick, with rotation being in opposite directions and 0 off by 90 degrees. I hope that makes sense to someone.
Toni, maybe you could post your code used, I have an equation figured just not sure how to put that in quaterions.
We have been testing the similar fenomena in our lab att ABB Robotics. We have not been able to reproduce this behaviour. Could you send the line, project and pmw files. I’m also interested in what calibration paper you have been using. Could you also make a simple sketch of how the configuration looks like with cameras and robots.
I was asked to send this per Mathew Garcia ABB US.
Attached is the complete line, project, and rapid code edited to fix the rotation issue. Robot still has rotation problem with standard rapid. Please test line and project without included rapid in your facility and post the results.
Thank You, that works like you said. Although,like you, I don’t like having to use this work around. Hopefully someone finds a permanent solution soon.
The recent errors with IRB 340’s turning in the wrong direction is now a known issue. There is a new gearbox introduced so for the new robots there has to be a adjustment in the eio.cfg file. There is or will be an RSI issued shortly. Pls contact Serop Product support to have the IRB 340 error resolved.
Chris and I are wondering, do they have a phone number and are they in the US or Sweden? Or better yet could you please just post the fix on this forum? We are almost a month behind schedule now due to this problem and would like the real fix as soon as possible. Thanks again for all your help during these times of trouble. We will miss Matt Garcia greatly.
If you are not able to download this you need to contact your local ABB office in your country for assistance.
You also need a new drive module key that ABB Robotics Aftersales can generate. The key that needs to be inserted is called IRB340 Type C. If you are a ABB external contact your local ABB office for assistance with this issue.
Note.
This is not an official notice for this error. It’s posted to speed up the process of solving the problem for those affected.
P.S Robert H Krukowski in ABB US has gotten this information and might be good person to contact for integrators in the US.
