Sign in to follow this  
Followers 0
TimWilborne

Coordinated Timestamped Data

20 posts in this topic

I was sitting here playing with a Controllogix and accidentally clicked the Coordinated Timestamp Data input format. Now I know basically what it is but can someone explain to me a realtime application of it? Also are there any disadvantages to using it? Thanks TW

Share this post


Link to post
Share on other sites
You have never wanted to know the time between two events? My customers are usually trying to shave every millisecond off their machine cycles. We have event logging with millisecond resolution time stamping Customers love it for optimizing their machines.

Share this post


Link to post
Share on other sites
I have never willing used CST Data format. Does it require more bandwidth / take more time to update than other formats?

Share this post


Link to post
Share on other sites
I guess I can see that angle, but I just feel there has to be more. Not that optimizing machines isn't important. We all try to squeeze a little more. But it seems that would be more of an extra your getting for free from a bigger feature

Share this post


Link to post
Share on other sites
CST is used in most of the I/Os, mainly applicable in some application like Motion, Sequence of events, diagnostics & in redundance 1756Sync Synclink is used to synchronous between racks

Share this post


Link to post
Share on other sites
Yes this is kind of what I am reading. My motion applications are just a bit too rough and must not require the precision to realize the advantage of this feature.

Share this post


Link to post
Share on other sites
Hey gurus I've got a motion application where the position is varying ever so minutely from cycle to cycle. We think it is scan times and propogation delays at this stage in our debug. WOuld CST help give as better idea of what is happening?

Share this post


Link to post
Share on other sites
We normally try to blame those small variations on the mechanical guys, have your tried that? Any osculation when it is idle? How much variation are we talking about? What type of controller are you using? What type of resolution does your system have? BTW, you know you are jacking my thread...but that's ok, this would be the only method I would probably use it for

Share this post


Link to post
Share on other sites
Share the evidence. So BobLfoot, do you have are graph or some way of capturing data so you can share it with us? I will be going out next week to install a rotary shear. Actually the customer already has it wired up and working in simulation mode. We can compare the motion profile our controller will use to the motion profile used by the other controller. For the first time they can see how the other controller is not matching speeds very well so it pulls the sheet metal instead of matching speed and just cutting the metal. Diagnostics and high speed data capture is crucial in motion applications.

Share this post


Link to post
Share on other sites
Here is an example. ftp://ftp.deltacompsys.com/public/jpg/EventLog.PNG Note, the picture is big. You must down load it and use your best viewer to see this and even then you may need to scroll. The most recent events are at the top. You can see where the states make their transitions, when the commands are issued and when the calculations are done and what the results are. The time stamps have 0.0005 second resolution. So you can see I am a big believer in time stamped data. There are filters so that only the things you want to see are logged. With this one fears no evil because the debugging tools show all. This is from a rotary shear application I have been working on and will install next week. You will be forgiven if you drool a bit.

Share this post


Link to post
Share on other sites
Does the size of the variation exceed the position lock tolerance for that axis? You may need to add some test logic to get some variables at the right times rather than try to monitor it with a PC, depending on the speed of the process... Edited by OkiePC

Share this post


Link to post
Share on other sites
Thanks for all the feddback so far fellas. TW - Apologize for the Hijack, but it seemed to fit. And yes the present variation is being blamed on mechanical by the Rockwell Motion Expert we had in. I'd just like to know my own machine better and be sure thats all it is. Peter - I'll look at the posting as I begin to digest my machine. I plan on having data before changing a single line. OkiePC - Can't say for sure right now. I should explain a little further. THis is a package wrap application and the axis of concern is cutting the wrap film. There is a variation of 1/2 to 1 inch in cut as the machine runs. It appear cyclic in nature, but cannot say it fits a pattern yet. BTW this pattern is not enough to mess up normal operation according to everyone except the operator on my shift {sort of a perfectionist}.

Share this post


Link to post
Share on other sites
Wow, 1/2 to 1 inch. I just came out of the steel industry and type of error would have everyone pointing fingers trying to figure out what was wrong. Find it hard to believe that ones mechanical unless you have a coupling getting ready to fly apart. Does it always lead, lag, or back and forth?

Share this post


Link to post
Share on other sites
I believe that CST data can be used with FBD PIDE and TOT. I could be wrong, but i thought i read this once

Share this post


Link to post
Share on other sites
More Info about CST The Coordinated System Time (CST) is a free running clock that counts micro seconds. All modules have a CST that starts counting when they power up. This clock is synchronized to all the local modules by the processor when its declared to be a master. The CST in the processor starts at 0 when power is first applied to the processor that previously had no battery. If a battery is already installed and the processor had been powered up then the CST time will already have a value in it. Thus to reset it you have to remove the battery when the processor is powered down. The CST is a 64 bit number that is held in a array of dint[2]. The CST gives the ability to schedule outputs at specified times after an event occurs. This is done by adding a value to the CST time when the event occurred and sending the CST to the output module via the timestamp. The input modules return all 64 bits but only the lower 24 bits are currently used in the output modules when scheduling is used. Thus time stamping is limited to 16.7 seconds into the future. This is based on the 24 bits being used (2 raised to the 24th give 16,777,216 micro seconds). Since this could change in the future its recommended that the ladder maintain both dint's and use the carry bit when adding a value to the lower 32bits as an overflow into the upper dint. When adding values to the lower 32 bits the number will become negative at some time and sometime later rollover to zero. When the value rolls back to zero is when the carry bit will get set. This will not effect the output time since only the lower 24 bits are looked at. The output module actually looks for its CST time to be equal to the scheduled output time (lower 24 bits) to set its outputs. Once an output module is configured to use timestamping all of the output bits will go at the timestamp. The timestamp must be refreshed in order to get the outputs to go through again. If not refreshed the outputs will wait until the timer rolls over and back to the time interval. Without refreshing this rollover will occur every 16.7sec. In order for modules to have a common CST they must be in a rack that has a processor that is configured to be a master. As long as this is true even a remote module can have a scheduled output based on an input from that rack. In this case you must make sure that the i/o update time is low enough to not exceed the CST scheduled time. Eg the reference time in the future is far enough that the i/o time doesn't exceed it. Once it gets to the modules it will work ok since the offset is based on the local CST time. http://rockwellautomation.custhelp.com/cgi...amp;p_topview=1

Share this post


Link to post
Share on other sites
rpraveenkum, so in the Rockwell workd the CST is also used to turn on outputs at exactly the right time?

Share this post


Link to post
Share on other sites
Yes of course CST is the backbone to activate the Modules in logix systems

Share this post


Link to post
Share on other sites
Even the motion module like the M02AE, M02As and HYD02? I have never heard of CST and these motion modules mentioned at the same time. I am very familiar with how these modules work. I do think it is great the Rockwell this capability. Ok, now what modules can take advantage of this? From what I have seen the I/O bus is asychronous to the scan.

Share this post


Link to post
Share on other sites
this what Tech note says about Motion Motion is controlled by the 1756-M02AE module (M02). One ControlLogix processor in the chassis must be configured as a Coordinated System Time (CST) master. CST enables coordinated motion. The M02s won't even configure if there isn't a CST master. All M02s controlled by the processor MUST be in the same chassis as the ControlLogix processor processor. The M02 is not a very intelligent device. The ControlLogix processor executes the motion instructions and does the coarse path planning to ensure coordination of the axes, and downloads the coarse path to the M02. The M02 generates the fine path. In order to meet the coordination specification, the communications from the ControlLogix processor to all M02s must be simultaneous. Right now, even ControlNet can't guarantee the time requirements. Thus, all M02s must be in the same chassis as the controlling ControlLogix processor. http://rockwellautomation.custhelp.com/cgi...amp;p_topview=1

Share this post


Link to post
Share on other sites
I worked on a wrapping machine a few years ago. The poly film was pulled continuously from a reel and cut by a rotating knife. The film had pre-printed registration marks which we used to correct the cut length. What I noticed was that from full reel to empty reel, the correction changed from about +10mm to - 10mm. Actually, I can't remember if was plus to minus or minus to plus. However, it was consistent and fairly linear with zero correction at mid-reel. A possibility?

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0