Is there a relationship between the Ethercat task and the gpascii task?

[hjpark]

1. When running the Power PMAC IDE with EtherCat enabled, a communication error occurred. This situation appears occasionally. (FW2.5.0.4 / IDE 4.3.2.19)

 

2. When I shut down the Power PMAC IDE when the Ethercat motor makes noise, the noise is gone (FW2.5.4.0 / IDE 4.3.2.19)

 

I think there is a relationship between the Power PMAC IDE and Ethercat. Please explain what kind of relationship it is.

[Omron Forums Support]

There shouldn't be any relationship between the IDE and EtherCAT. You might want to try closing watch windows to see if any commands they are be sending could be the problem.

 

What Power PMAC CPU is this? (460? 465? Dual Core ARM?)

[hjpark]

The cpu is Dual core ARM.

 

The watch window was on, but the query item was empty.

 

When it comes to master shift, is it correct that EtherCAT and IDE have no relationship at all?

 

I think there is a relationship when you look at this phenomenon.

[Omron Forums Support]

Is either core of your CPU heavily loaded? Can you show your "Tasks" tab in your IDE "Task Manager"?

 

Do you have a Shift Time defined in your "Distributed Clock" tab? I know default settings work most of the time. Ideally "Cycle Time" should be set to your servo/ecat period and "Shift Time" should be set to half of that.

[hjpark]

I set the Distributed Clock settings as you said. I'll check the CPU load.

[sutty]

I also know this kind of phenomenon where a motor sounds noisy. Several reason:

* bad ECAT-setup (steady noise)

* wrong ServoExtension (should be set so ECAT-Update matches servo-freq) - steady noise

* noise during opening/connecting/closing IDE resp. especially IDE-Taskmanager to PPMAC (gpascii) in combination with heavy cpu-load

[hjpark]

Ethercat disconnection when exiting the IDE still occurs. (FW2.6.1.0 / IDE 4.5.1.3)

Commonly, this problem occurs when there are many Ethercat slaves and the CPU load is high.

Is there any solution other than changing to quad core?

[steve.milici]

Please place a G-TKS ticket on this issue for ODT Engineering to review.

[markr]

I have also seen this, with the motor noise problem becoming worse with more ECAT slaves and set to hot connect - this was seen while using an NYB2C.  I have also separately measured that connecting and disconnecting to the IDE temporarily affects ECAT timings. 

Strangely, the motor noise disappears while 'gathering' in the plot tool. As soon as the 'gather' is stopped, the issue re-appears. 

[tecnico]

This is also my experience, can't figure out why but also DCClockDiff appears to be much more stable (and with lower value) when gathering.

Could this be an unwanted effect of having a non-default value of ECAT[i].RTStateCheckCount?    (50)

[steve.milici]

Gathering should have no effect on the DC Clock but “RTStateCheckCount” too fast could be an issue.

[tecnico]

So what is a "proper" value for  RTStateCheckCount?
As for the gather effect on DcClock, it shouldn't but it appears to have some effect.   

 

[steve.milici]

The value of “50” that you show should be acceptable. I have done some testing (FW 2.7.1.2 IDE 4.6.3.9 on CK3E) and have found no correlation to DcClock issues with the “RTStateCheckCount” mechanism or data gathering. You may be experiencing a different issue. I might suggest you contact your local Omron technical support for a review of your EtherCAT usage.