R.S. Logix Trends

[Bulletin_Blues]

Hello all. Im getting a call every coupla weeks that a mill is just "shutting down." I have Very serious suspicions that a normal stop is being pressed, seeing as though most stops fault the mill etc.etc.

Last night i tried using R.S. Logix Trends to view the pushbuttons
compared to the run bit. Unfortunately it took me most of the night to get it tweaked, and they never shut down again. Now Im at home trying not to think about work, but... Heres some silly questions I`m pondering.

1. I can adjust the refresh rate on the display to up to 50 milliseconds. Is that the refresh rate on just the "display"? or is that the refresh rate on... well... what Im going to be actually monitoring? The reason I ask is the computer started getting really weird at anything under 1 second. I did hook up a laptop as well that was able to monitor at 50 milliseconds so I think Ill work with that.

2. If I Trend the 8 hour shift and then print it out at the end of the shift is it going to give me like 900 pages of trending, or can i just save this to a location on a drive?
This is the only mill that does not have any SERIOUS data logging going on, and Im just trying to work with this (and learn) while I can, so... any comments on R.S.Logix Trends would be appreciated.
Thank You All Very Much! I learn so much on this sight they actually gave ME a Tech Job, ouch!! :}

 

[Beryl]

Trend in RSLogix is just for monitoring in real time. I would not recommend updating a trend every 50ms for 8 hours. My system will start to have issues with a 30min trend (8 pens). To my knowledge you cannot save this anywhere. It works great for looking at points in real time but is definately not SQL or Historian.

Good Luck

 

[msinclair]

One thing that I have done on occasion is to add a trap bit to catch short duration events when I don't have data logging on the line.

Something like:

 Button                   Trap_Indicator

--| |------------------------(L)

I agree with Beryl - RS Trend is great for real-time (with events that are true for several scans). I don't think that it will help you here though.

 

[Rick Densing]

You could also have the normal stop increment a counter.

 

[jimbo3123]

You could also have the normal stop increment a counter.

You could do this, or a programmed in fault-logger

Say, when someone pushes the normal stop this happens
MOV (time info -> N11:0)
COP (N11:0 -> N11:100), 100 words long
COP (N11:100 -> N11:1)


This will give you a log of the time that the normal stop button was pressed for the last 100 times that it was done.

 

[Beryl]

Jimbo,
I not sure how your fault trap works can you explain further? It does not work the way you have posted.
Thanks,

 

[msinclair]

Beryl,

Allen Nelson did something similar in this post:

http://www.plctalk.net/qanda/showthread.php?postid=1603#post1603

Marc

 

[Beryl]

Thanks Marc,
I have never thought of doing it that way. Works pretty slc. (pun fully intended)

 

[Mickey]

The Histograms tool might work for this.

Mickey

 

[Bulletin_Blues]

Thanx for the ideas

Thanx for the replies.

Are you saying i can log TIME! Never even thought of somethin like that. That sounds sweet. Where is the time value found?

Can't say I ever used a histogram either. Im at home right now with my laptop and histograms wont open (greyed out), guess I have to be online?

Another problem is alot of permissives are in the run
rung and when the mill stops alot of these permissives shut off as well, such as coolant pumps turn off, choppers come up, stuff like this
and I have to verify that these are not causing the stops themselves.
So I`m running into alot of chicken or the egg type chasing. With the Trends I was actually able to see P.B. pushed THEN coolant off. But I agree I`m not sure if using trends is the best idea now.

Anyway time stamps sound cool and I will look into incorporating these into my grey matter.

The mill uses R.S.logix 5 along with an Automax max pack for drive control.

 

[Begolfin]

The idea of using a latch bit from the stop button is the easiest way I've found to catch a intermiten problem. I have also used the latched bit to move the A-B plc-5 registers S:21 hour, S:22 minute,S:23 second,, with a oneshot to a register to pinpoint the time of the problem. If ones available, send the information to the MMI ,,the operator might be able to get a quick read when the event happens and spot the problem. Hope this helps.

Later
begolfin