Motion Diagnostics

A section of our plant has 20 “identical” accumulation systems. Each system has 4+ axis [PF527] controlled by a CompactLogix V28 with N_Tron switches.
On one systems, one of the axis is not traveling to its position correctly after long moves.
A program compare showed no* difference in programs from systems that work and the one that is having issues.
I did notice under the AB:Motion_Diagnostics:S:1 tag in the axis that is having issues the LateControllerToDriveTransmissions counts for the PF 527 increasing while the other axis in this system and the other 19 are not.
I've compared programs, switch and axis configurations and reset Ethernet cables.
They've also replaced the encoder for this axis with no change.


What should I be looking for assuming I'm not heading down the wrong path with the above diagnostics tags?


Thanks,
Bob

probably a chase of the wild goose - but every time I hear "high speed" and COMPACT-Logix mentioned together - I can't help but think of this ...

http://www.plctalk.net/qanda/showthread.php?p=851313&postcount=6

going further, is your problem "random and intermittent" - or is it "repeatable and consistent" ??? ...

There is an asterisk after a [no] in the OP; is there a missing footnote?


Was the problem working for a while without this issue and this just started, or has it been this way since commissioning?

I'll say it's a consistent repeatable issue. I will double check* but I believe they have had this issue since day one on this axis.


*I'm relatively new to this company and the maintenance group in this area try to solve problems on their own to start with.


Sorry, the * was meant for the tag LateControllerToDriveTransmissions

Another wild goose chase: what is the firmware revision?


Because this document https://literature.rockwellautomation.com/idc/groups/literature/documents/rn/1756-rn220_-en-p.pdf

(p. 7, table 10, Lgx001222124, Lgx00121801) mentions an issue about AXIS_CIP_DRIVE, MAW instruction, late/lost transmission count, and missed position prior to firmware revision 20.011. It seems to suggest that the logging of the last or lost updates to ...:S:1 were the cause; can that logging be turned off?

I had seen that also but kind of moved on since these are V28. I don't know if the logging could be turned off.

My colleague slowed this axis down yesterday to 50 Hz from 60 Hz to see what it does.
I report back on that one.

Thanks All!

Here are screen grabs of the motion diagnostics tabs for the problematic axis and a function axis in the same system along with the network between devices.
The subroutines are in a periodic task set for 10 ms with a priority of 10.
I pulled the network cables out of the ducting and away from wiring. No change.
I've re-seated all the network connection. No change.
I swapped HSC cards slots to see if problems followed. No change.
All other devices are communicating over the same CAT 5e without issues [So I'm told].


Any hints/ideas on what could be causing the late controller to drive counts in the problem axis?


Thanks

Double checked myself. The problem systems periodic task with the axis is set at 9 ms and a priority of 9 and two other system that are working fine are 10 ms and a priority of 10.

How are you controlling the axis position? Are you using profiled motion instructions (MAM, MAG, etc) or are you rolling your own positioner and just sending speed commands to the 527 (MAJ)?

Axes can be set up to be just velocity axes or position axes. In the Axis Properties window this is under General / Axis Configuration. You can perform a profiled move on a velocity only axis but it will just get the velocity command based on the move profile sent to it. An axis configured like this is VERY dependent on its velocity accuracy.

Are both the actual, physical position and the feedback position both incorrect or just the physical position?

If just the physical position, have you plotted the actual position to the position command to see where things are coming apart?

Keith

Look at the controller to drive time. They are much different and too long when having issues.

I didn’t write it but it looks like it’s set up as Freq Control and they are using MAM command to send the bin to its new calculated position.
They don’t have feedback on the bin.

I'm mentally stuck thinking it's a communication issue internal to the panel but at a loss on how the other devices appear to be working fine with the way they have the network cables routed/plugged in.

Originally posted by Bob O:

They don’t have feedback on the bin.

Click to expand...

OK, so the axis is getting an open loop velocity command and is expected to go where the velocity command leads it. I would have to guess that the drive is not maintaining the speed it is being commanded.

Unless the velocity command is VERY dynamic I don't think the comm timing is the actual issue. It might be a symptom of an unhappy drive control board but I don't think it is the cause of the positioning issue specifically. The drive telegram is maintaining the same velocity command for expended periods of time. Variability or delay in that telegram isn't going to affect what the drive is doing that much since all the numbers are always the same.

You said that this axis has no feedback. However, in your original post you talk about changing an encoder and later you refer to moving around an HSC. What is the encoder used for?

Keith

I was thinking actual feedback on the bin being driven by the motor. My error.
The encoder is on the motor that is driving the bin through a star wheel type gear.

What do you think could cause the controller to drive late transmissions?

I don't know enough about CIPMotion transmissions to be able to tell what might be causing this. Since it is specific to the one drive I would have to guess it is a soft drive fault. But that is just a guess.

But put the late transmission thing into context. Your course update rate is 5msec. That means 200 packet transmissions every second. Presumably the system was powered up for a while before you took the screen shot. You have accumulated 11 late packet events...out of 200 transmissions EVERY SECOND. While this certainly isn't great news, I personally wouldn't be hanging a consistent positioning error of an open loop axis on a transmission timing anomaly.

So the motor has an encoder. Have you checked the axis commanded position against the axis reported actual position to see if there is a difference? You might not get feedback since it is frequency control. I have never operated a motion axis in this mode before.

Keith