Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
EtherCAT Setup - How much is automated?
#1
Hello,

I'm interested in understanding the procedure and workload involved in setting up a PowerPMAC as an EtherCAT master motion controller controlling an EtherCAT slave motor which is supported by Delta Tau.

How much of the interface is automated? Looking at forum posts (such as this one: http://forums.deltatau.com/showthread.php?tid=2676), I see that things like I/O require a mapping between Delta Tau variables and the EtherCAT hardware registers. But for a supported motor is it as simple as getting the PowerPMAC to search the EtherCAT chain for a new device and then selecting this device and assigning it an axis number?

For example, if I have software which knows how to talk to Delta Tau i-variables (ie to set speed and acceleration) would these i-variables also need explicit mapping to EtherCAT registers or would this be handled by automatically by the controller when the connected slave device is selected?

Thanks for any info,

James
Reply
#2
If you have the EtherCAT slaves's ESI file, or the device has its own file contained on itself, the setup software automatically maps/sets up basically everything for you.
Reply
#3
To give a little bit more detail (and it does change a little bit depending on what type of PMAC you have, as we currently support two different EtherCAT implementations):

1) Slave devices should be auto-detected when scanning the network based on ESI files
2) You will need to select which PDO mapping (typically premade in the ESI file) you intend to use depending on what features of the drive you intend to use
3) Some configuration will be required--2-3 parameters, mainly setting up the EtherCAT Clock to make sure all slaves are set properly
4) In System Setup*, you will select which EtherCAT drive is connected to the motor and you will set which mode of operation is desired (Cyclic Synchronous Position, Velocity, or Torque modes are supported at this point)
5) Verify that the correct PDOs are selected for output register (TargetPos/Vel/Torque--0x607A, 0x60FF, or 0x6071), feedback (Position Actual Value--0x6064), status (Status Word--0x6041), and amplifier enable (Control Word--0x6040).

Again, that's a very broad strokes picture of it, and some drives may require or support additional EtherCAT registers which may need to be set for use, but that's the process for a theoretical "generic" drive.

*or the new screens in IDE 4.x if it supports your hardware
Reply
#4
Thank you both for your replies and the information.

AAnikstein - what are the two different EtherCAT implementations that you support?
Reply
#5
Currently, all PowerPC based PMACs (Clipper, Brick, UMAC and Etherlite) support only the Etherlab stack. All ARM based PMACs (CK3E, CK3M, and future products) support Acontis by default with Etherlab available for backwards compatibility in the newest firmware. The NY51[]-A at this point only supports Acontis, however we intend to add Etherlab support for backwards compatibility later this year.
Reply
#6
(05-04-2018, 08:42 AM)AAnikstein Wrote: Currently, all PowerPC based PMACs (Clipper, Brick, UMAC and Etherlite) support only the Etherlab stack. All ARM based PMACs (CK3E, CK3M, and future products) support Acontis by default with Etherlab available for backwards compatibility in the newest firmware. The NY51[]-A at this point only supports Acontis, however we intend to add Etherlab support for backwards compatibility later this year.

Can you comment on the current thruput comparison of Acontis vs Etherlab? I understand that you have been working to improve speed while retaining ease of use with Acontis.
Reply
#7
(05-03-2018, 12:30 PM)AAnikstein Wrote: To give a little bit more detail (and it does change a little bit depending on what type of PMAC you have, as we currently support two different EtherCAT implementations):

1) Slave devices should be auto-detected when scanning the network based on ESI files
2) You will need to select which PDO mapping (typically premade in the ESI file) you intend to use depending on what features of the drive you intend to use
3) Some configuration will be required--2-3 parameters, mainly setting up the EtherCAT Clock to make sure all slaves are set properly
4) In System Setup*, you will select which EtherCAT drive is connected to the motor and you will set which mode of operation is desired (Cyclic Synchronous Position, Velocity, or Torque modes are supported at this point)
5) Verify that the correct PDOs are selected for output register (TargetPos/Vel/Torque--0x607A, 0x60FF, or 0x6071), feedback (Position Actual Value--0x6064), status (Status Word--0x6041), and amplifier enable (Control Word--0x6040).

Again, that's a very broad strokes picture of it, and some drives may require or support additional EtherCAT registers which may need to be set for use, but that's the process for a theoretical "generic" drive.

*or the new screens in IDE 4.x if it supports your hardware

The above is what may magically happen if you have the perfect Acontis stack supporting PMAC unit with the right firmware and proper IDE version combination, extreme luck and the proper interplanetary alignment of galaxies.

Reality is that it took weeks of reading and effort between a CK3E and a PowerClipper unit to connect to Kollmorgan drive and even when it worked, it just released the brake and before I had a chance to send a new value to make motor spin the ethercat bus would time out, IDE would crash and lots of CTRL+ALT+DEL + END-TASKs, service restarts and a few minutes of reloads and reboots later while searching the forum for the 1209421812908 time for anyone else maybe experiencing the same, low and behold a new firmware version is released and the release notes describe issue experienced above as now fixed for acontis stack only when I only had the Etherlab PoweClipper unit. Fast-forward nearly a year and a CK3E now sits on my desk with the proper acontis ethercat stack that is supposed to actually work. Right next to it there is an ethercat Omron NX coupler which I was once able to connect to the other unit that I no longer have access to, but have yet to succeed on CK3E despite following instructions - the NX may not have the right firmware that supports distributed clock = next step to verify on this, meanwhile projects need to move forward and Beckhoff/Siemens/etc just work so that's what gets used since there is only so much time one can invest in the absurd learning curve needed to master the intricacies to get this hardware on ethercat bus up and running.

If you are looking for intuitive, easy, reliable, mature or broadly supported I would look elsewhere. Just about a week ago I was looking for firmware updates / checking how CK3E support has progressed since I last touched it nearly a year ago, only to find out all CK3E firmware files have been pulled from the filedepot and to contact support by phone if help needed - in short - the prior transparency with error fixes in release notes and ongoing support progress visibility is now gone and whatever issues I run into I can no longer check if a different firmware version may have resolved.

Good luck - I will keep an eye on this thread to see if your ultimate results are better than mine, hopefully that will give me motivation to keep trying on the CK3E and stop using it as a paperweight on my desk reminding me daily to never buy another item like this without first finding a real world customer who successfully implemented same setup I envision for it.
Reply
#8
(12-26-2018, 10:47 PM)dougjr17 Wrote:
(05-03-2018, 12:30 PM)AAnikstein Wrote: To give a little bit more detail (and it does change a little bit depending on what type of PMAC you have, as we currently support two different EtherCAT implementations):

1) Slave devices should be auto-detected when scanning the network based on ESI files
2) You will need to select which PDO mapping (typically premade in the ESI file) you intend to use depending on what features of the drive you intend to use
3) Some configuration will be required--2-3 parameters, mainly setting up the EtherCAT Clock to make sure all slaves are set properly
4) In System Setup*, you will select which EtherCAT drive is connected to the motor and you will set which mode of operation is desired (Cyclic Synchronous Position, Velocity, or Torque modes are supported at this point)
5) Verify that the correct PDOs are selected for output register (TargetPos/Vel/Torque--0x607A, 0x60FF, or 0x6071), feedback (Position Actual Value--0x6064), status (Status Word--0x6041), and amplifier enable (Control Word--0x6040).

Again, that's a very broad strokes picture of it, and some drives may require or support additional EtherCAT registers which may need to be set for use, but that's the process for a theoretical "generic" drive.

*or the new screens in IDE 4.x if it supports your hardware

The above is what may magically happen if you have the perfect Acontis stack supporting PMAC unit with the right firmware and proper IDE version combination, extreme luck and the proper interplanetary alignment of galaxies.

Reality is that it took weeks of reading and effort between a CK3E and a PowerClipper unit to connect to Kollmorgan drive and even when it worked, it just released the brake and before I had a chance to send a new value to make motor spin the ethercat bus would time out, IDE would crash and lots of CTRL+ALT+DEL + END-TASKs, service restarts and a few minutes of reloads and reboots later while searching the forum for the 1209421812908 time for anyone else maybe experiencing the same, low and behold a new firmware version is released and the release notes describe issue experienced above as now fixed for acontis stack only when I only had the Etherlab PoweClipper unit. Fast-forward nearly a year and a CK3E now sits on my desk with the proper acontis ethercat stack that is supposed to actually work. Right next to it there is an ethercat Omron NX coupler which I was once able to connect to the other unit that I no longer have access to, but have yet to succeed on CK3E despite following instructions - the NX may not have the right firmware that supports distributed clock = next step to verify on this, meanwhile projects need to move forward and Beckhoff/Siemens/etc just work so that's what gets used since there is only so much time one can invest in the absurd learning curve needed to master the intricacies to get this hardware on ethercat bus up and running.

If you are looking for intuitive, easy, reliable, mature or broadly supported I would look elsewhere. Just about a week ago I was looking for firmware updates / checking how CK3E support has progressed since I last touched it nearly a year ago, only to find out all CK3E firmware files have been pulled from the filedepot and to contact support by phone if help needed - in short - the prior transparency with error fixes in release notes and ongoing support progress visibility is now gone and whatever issues I run into I can no longer check if a different firmware version may have resolved.

Good luck - I will keep an eye on this thread to see if your ultimate results are better than mine, hopefully that will give me motivation to keep trying on the CK3E and stop using it as a paperweight on my desk reminding me daily to never buy another item like this without first finding a real world customer who successfully implemented same setup I envision for it.



Hi dougjr17!

Finally, how was your experience with the Kollmorgen Drive an EtherCAT interface? Now a days i'm studding to use this alternative to our system. I don't know really if actually all the IDE and Stack of EtherLab are improved. Did you get some success goals with PMAC-EtherCAT-Kollmorgen systems?

Thanks and regards
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)