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

antsrealm

Member
Join Date
Dec 2010
Location
Brisbane
Posts
207
Hi,

I have configured a project with the following hardware.

5069-L30ERM
842E-CM-S Absolute Encoder.
Stratix 2500 1783-LMS5 (Is this suitable?)
5069-OB16F

It's all working, the axis is setup under the MG. I have the correct units and operation.

The only issue is I'm not sure it's working as fast as it could be.

I believe this is done from the update period.

When I have this on 6ms it time syncs and everything is ok.

When I lower it down and currently I have it on 0.5ms I get loss of communication with the encoder. It drops in and out every few seconds and wont time sync.

Can anyone give me some advice on what to expect here and maybe what update period I should run at.

The actual PLC program is fairly small so the utilizations are fairly low.

I could email the program to someone if you want to have a look. It's a basic application. Just an encoder giving machine position feedback and triggering certain things at certain positions. No servos.

Thanks.
 
Did you try bypassing Stratix 2500 (plug encoder direct in PLC ethernet port)?

It's been a while but I've been told that CIP sync enabled switches are preferred for encoder applications.
Even then, 0.5ms might be pushing it.
 
Sunday morning, all my manuals are at work. 0.5 ms is way too fast if you ask me, why would you even want to go below 6ms when your average servo update period is 8ms?

Think about your PLC scan time, as well as the response and cycle time of the components of your machine 6 ms should be more than adequate.
 
Last edited:
The Coarse Update Rate is just the rate at which you can do PLC-commanded things with the axis, like read the position into your program or issue a MAG or MAM or MCD instruction.

If you're using a CIP encoder and a fast output module to fire outputs based on a Motion Axis Output Cam, then the Coarse Update Period defines the approximation points between which the motion planner assumes a given speed/accel/decel.

Are you using MAOC, or trying to compare your .ActualPosition to a value and use an ordinary output instruction ?
 
Can you post an example of how you are using the feedback? When I build anything with CIP motion/Time Sync, the Sync is usually in the uS range.
 
Ahh ok. I was just worried that maybe I was doing things wrong. I did try direct to the PLC with no switch, didn't see any difference.

I am back in the office with this project in two days time so I can post some more information then.

I am comparing actual position for non critical stuff and I used a MAOC and the fast output card for the high speed stuff if it turns out I need it.

I trended the actual position against the firing of the MAOC driven fast outputs and they seemed less accurate then just using a comparator on the actual position. But then I suspect the trend is probably slower then the fast outputs and not accurately representing what's happening.

I will give you some more info regarding scan time etc when I'm back in the office but i'm sure the scan time is very quick at the moment.

I will begin the install of this PLC into the old machine in 4 days time.
 
I trended the actual position against the firing of the MAOC driven fast outputs and they seemed less accurate then just using a comparator on the actual position.

It's very likely that you're getting the status of the discrete Outputs up to a full RPI later than they actually occur, since those status values aren't going to be updated in the PLC's tag database until after they turn on.

Combine that with the update of the Actual Position at the Coarse Update Rate, and the update rate of the Trend object, and you get a visualization that looks more like Picasso than Hopper.
 
It's very likely that you're getting the status of the discrete Outputs up to a full RPI later than they actually occur, since those status values aren't going to be updated in the PLC's tag database until after they turn on.

Combine that with the update of the Actual Position at the Coarse Update Rate, and the update rate of the Trend object, and you get a visualization that looks more like Picasso than Hopper.

Yeah I had a feeling.

I'll update with how I've used it and see what you think in 2 days time. Thanks.
 
Ok sorry for the delay coming back on this. Ken if you could let me know what you think about this that would be great.

I have installed the PLC on site now and everything is working except I have issues with consistency of the encoder position.

The encoder runs at a top speed of 30 rpm.
The encoder operates 0 to 360 degrees for 1 product pitch and 1 revolution.

I am currently comparing the Axis.ActualPosition in the PLC to a window set by a LIM instruction.

If the actualposition is within the window and the photoeye is on then the machine will fault.

Problem is the position it sees the photoeye at 15 rpm and the position it seed the photoeye at 30rpm is significantly different.

I've had this perform well at times but then terribly later on.

I was hoping that at a relatively slow speed of 30rpm it would be fine without using MAOC instructions as there is at least 15 windows like this across the machine.

I have noticed that on power up the encoder is time synced with the PLC but after approx 1 minute it is not synchronized any more and never syncs again until a power cycle. Does this matter?

I have tried 4ms update period to 10ms update period with no real noticeable difference.

PLC scan time from memory was 0.1ms

One of the guys on night shift was playing around with MAH instruction and claims its working a lot better after he executes the MAH isntruction once after power up. Does this really have any impact?

Thanks for the help, not many people seem to know much about this stuff.

Thanks,
 
My logic was if the speed is 30 rpm max then that's 360 degrees every 2 seconds and 1 degree every 5.5ms so if the update period is 4ms and i'm using the actualposition in the PLC program then shouldn't that be good enough?

I don't work in increments less then 1 degree.
 
As a contingency I have written what I think is the correct way to do the MAOC. I will be using it mostly internally checking the timing of sensor inputs from product detection against these windows set in the MAOC. Would this be an acceptable way of getting sub 4ms response times ??

It will also operate two standard outputs for now but I have the option to go over to the fast output card but I shouldn't need that.

This code is currently just under the main task getting scanned continuously.

I have just created it offline and have not tested it yet. At the moment the machine is running ok without it today. If it plays up again I'll put it in and see what happens.

Image Link as the attachment doesn't seem to be appearing.
https://drive.google.com/file/d/17UYm2KXlAPSDmU81C92KjtQWwtImlQxS/view?usp=sharing
 
Last edited:
You may be having latency problems with the program scanning.

The MAOC is a black box. If you want to put logic in then create a task that is event driven using MAW axis watch.

You can then trigger at axis positions and record the sensor states in your logic.
 
Thanks Geoff.

Current scan time is 0.1ms I would of thought that was fast enough ?

The MAW event is new to me.

I am just getting my head around this course udpdate period.

It seems like this encoder cant go lower then a 4ms coarse update and that's the time it updates the actualposition data to the PLC but also seems like the MAOC outputs would also be updated at the coarse update period unless im using the fast output card. A lot of my error checking is internal and against inputs so I'm not sure how i can get it any quicker then 4ms. Any thoughts on the time sync saying "not synchronized" for the encoder. This seems like a red flag to me.
 
At 30rpm, you're looking at 5.56ms per degree. What accuracy are you actually chasing for your application? Also don't forget that physical I/O has associated filters and RPI settings.
 

Similar Topics

Hello Everyone, Some of you may know I am working on conveyor project and I am about to finish it. I need some advice regarding my project. I...
Replies
3
Views
2,211
Hi Everybody, I need your advice regarding pros and cons of using DeviceNet. We have an application with 17 VFDs and 33 fixed speed devices and...
Replies
1
Views
5,099
Hi everyone id like to start by saying im not a PLC programmer so my technical terminology is limited. i would like advice on what software is...
Replies
4
Views
269
Connected Components Help Hi there everyone, I’m recently new to the PLC world and was hoping somebody might steer me in the right...
Replies
3
Views
394
Hello, I'm struggling to learn something on Wonderware, and the distributors are taking days to get through the email chains. I was hoping for...
Replies
1
Views
343
Back
Top Bottom