We have a line with 4 IRB360 and we have ca 1 time in the month during production an unsychronisation fenomen without any warnings in the pickmaster log, all the robots are still working but have wrong pick of ca 150mm in X direction. In this case we think that we have an encoder problem but we check all and ask the customer also to try some inspection for us.
These are the different step that we ask to the customer to try:
Stop and start project => NOK
Close and Start pickmaster => NOK
Restart PC => NOK
Look at the encoder and cable=> there was no problem.
The only way for the customer to solve the problem was to do a Cold start. =>OK
Yesterday it happened again and I ask the customer to try some extra check, I ask him to take the power of the 2 ethernet switch for ca 1 min and then ask him to try again, with this instruction was the problem solved. It seems to be an ethernet issue but I can not explain why this only happened ca 1 time in the month and that we do not have some warning in the pickmaster log. The problem now is that I do not know which switch is giving the problem, if it happened again I will ask the customer to try 1 by 1 but maybe has someone got the same problem and can give some tips to solve this issue?
When saving the log files we could see this :
Log entries: 5576, 1F813040. System time: 02/06/12 17:13:09 Sync time: 2012-02-06 16:13:09 719 webwiz/622/PickMaster_Log_step_1.zip
There is a difference between system and Sync time, could this cause the problem?
We have had a similar problem that turned out to be a third party software application “Java Updater” that caused an issue with the Time Sync services that pick master uses.
The problem was also erratic as the Java Updater would run randomly to look for an update.
Is the 24VDC camera supply from the robot ? as cycling the power on the camera resolves the problem.
We removed the software and the problem went away.
We have experienced a similar problem. In our situation, the pick position was always off by the trigger distance for the camera. It behaved like the robots were one camera scene off.
Our installation used Insight cameras communicating to Pickmaster as external sensors. We found that the processing time of the camera had to be about 1/4 of the time between triggers or this situation would occur. We were never able to understand why the processing time was so limited and it has reduced our vision capabilities on the system.
If your camera triggers are frequent or you have long processing times, it could be contributing to the problem.
Mark, I will check if we have some third party software installed on this PC.
Craig, we are using the basler scout camera from ABB, our conveyor speed is about 350mm/s and our trigger distance is about 200mm. I do not think that this could be our problem but thanks anyway for you help.
I noticed you have a lot more “Warning 4318” than “Warning 4317” in
the PickMaster log. These warnings will typically show up in same numbers if there
is a “normal” time sync problem. I counted to 156 “4318” and only 24 “4317”.
This could indicate some problem/instability/weakness of the trigger/strobe signals.
Maybe there is a problem with the shielding ?
Your are right we have a lot of these warnings, but if you take a good look all these warnings happened when the robots are restarted after a stop during production due to collision or door request from the operator and when the conveyors are still running. I think this is a normal warning in my case. The most warning 4317 are concerning conveyor 3, this is a conveyor without vision and triggerring on lugs with a sensor and is also happening after a robot restart when conveyor is still running. By experience when I got a network problem I get the same warnings in continue and the robots stop to pick.
I will first check the idea of Mark to be sure that there are not third party software trying to connect for updates but thanks a lot for you help.
Mark, a question for you, What will you mean by cycling the power on the camera ? Will you mean that we should cut it off some times? In our case we use a separate 24V power supply.
When we had this problem we would cold start the robot and think this was the solution.
However because our camera was powered from the Robot we were removing and then re applying the camera’s 24VDC supply during the cold start process and it was not the Cold start that was rectifying our problem but the “cycling” of the camera power.
Error 4318 is also an indication of something affecting the time synchronization services like a firewall or third party program and this error was also present during our problem.
If your EKI 2725 ethernet switch/router has a built in DHCP server you may want to disable this aswell.
I went last week to the customer and check for our problem but I did not really find third party program. I have check all network settings and change some parameters to be sure. I have change some settings in our Siemens WinCC flex runtime. I have also wired our ethernet switch to the ground. I hope that the problem will now disappear with all these changes.
