Advice regarding configuration of 842E-CM-S / 5069-L30ERM Via Stratix 2500

Accuracy withing a degree or two consistently would be acceptable. That's why I felt with a course update time of 4ms and a 5.56ms per degree actual speed then it should be ok just using the actualposition.

As an example we might see product via a photoeye at 22 degrees and dissapear at 208 degrees so we check either side of that to make sure the sides haven't fallen open. So a windows is programmed a few degrees away and if the sensor is still high during this window it will fault the machine.

Initially I was seeing big variations in where this sensor saw the product at different speed. But today its running fine. It would be a consistent variation and it wasn't slipping. I was just worried that with the time not synced that maybe that was an indication that something wasn't updating correctly or quickly enough.

It's currently got the encoder, plc and HMI on it's own local network via a stratix 2500 lightly managed switch (1783LMS5) I think this has QOS and supports the time protocol. Later today when production stops I'll move the HMI and the switch over to the other port and just have the encoder direct to the CPU on the other port to see if that helps.

Here is a link to a photo of the time sync tab on the encoder.

https://drive.google.com/open?id=1-3PvfgQ_TLWwXuXKlKXwd2k6ybnC7b_C

Is your 0.1ms the task execution time or the task interval?
Is your task running on a schedule or in the continuous task? If it is continuous then there is a lot of overhead that gets done between runs.

Edit: that image shows the encoder as not sync'd with the PLC. The PLC should be the CIP sync master of the encoder.

Yeah its continuous with a scan time of 0.1ms and an interval time of max 0.3ms.

Yes all routines are under the continuous task.

link to photo...

https://drive.google.com/open?id=1-B8UjLD86NYJO9ItIXYmvDIzOX8T2_AT

Can you give me a bit more info about the time sync. I have ticked the box on the PLC processor to enable time sync and its status indicator is on to say it is the system time master. I assume I should see on the enconder that it is synced with the master / plc right ?

Can you post the equivalent of the screenshot I've attached for your project? The next place to look is at QoS, Smartport and Port Security settings on the switch. I don't believe the switch itself has any PTP settings.

Hopefully this is sufficient...

https://drive.google.com/file/d/1-UjA2SwW4iwjuNJ9_anKyLWcIDdIXC9i/view?usp=sharing

The update period was previously on 4ms I extended it out to 10ms to see if it would sync.

antsrealm said:

Later today when production stops I'll move the HMI and the switch over to the other port and just have the encoder direct to the CPU on the other port to see if that helps.

Click to expand...

Ok, configuration wise it should be working. Do the above and see if it syncs. You're correct in saying the PLC should be the master and all the blank fields in the Time Sync tab for the encoder should be populated. If bypassing the switch does fix the issue, go back to the QoS/Smartport/Port Security settings on the 2500.

Thanks Jeev,

I've got the encoder connected direct now and it's synchronized with a 10ns offset so much much better then before. It can stay like this, so no need to worry about the switch. I've brought the update period back to 4ms to make it as quick as possible and we'll see how we go.

Thanks for your help. I called rockwell about this and spoke to the reps and the drive / motion guys. They all had no idea when I started talking about motion. This was a far better way for me to get quality help

Photo of time sync now after connecting direct ...
https://drive.google.com/file/d/1-VYLjyYuMmNrKU2gdt1l5NK-rm_5Wgjh/view?usp=sharing

It sounds like you are doing open flap detection. You can cheat a little on that and just measure the length of the box instead of checking windows.

Yeah that was open flap detection but there is quite a few other checks and then there is some film feed outputs as well that require some accuracy from the encoder. Seems to be holding up ok for now