CLogix JMP > LBL instructions

PeterW

Member
Join Date
Jun 2006
Location
Calgary, AB
Posts
2,531
Has anyone used multiple Jump to Label instructions in Control Logix and if so, have you noticed a scan issue the more you use.

a colleague was testing a theory in an effort to reduce scantime and found the scan of the PLC (an L63) raised to above 200mS when jumping code, instead of the expected lowering of the scantime.

His test including jumping multiple times in 100 routines, which gave approx 900 jumps altogether.

He stated that the more JMP>LBL pairs the worse the scan got.

So has anyone else experienced this and what's the likely cause?
 
I have never played with any of this, so I don't personally know. But what is the whole context of his test? One great big huge program file, or is it broken into subroutines?

I would expect that if there's whole sections of code that you're interested in skipping, you ought to have that in a separate subroutine, and just JSR into that subroutine when you need to. I.e., instead of hitting a JMP and having to search for the corresponding LBL, don't even evaluate any of the undesired code...
 
200ms? I've got some huge programs in CLX, and the cyclic task scan time still will not go above 18-20ms.

Off hand, I'd say that your colleague is jumping BACK, rather then jumping forward, and executing more code per scan then the thinks he is.

If he's really concerned that JMP/LBL is a problem, put the conditional code in a subroutine, and only call it when needed.
 
200ms? I've got some huge programs in CLX, and the cyclic task scan time still will not go above 18-20ms.

Off hand, I'd say that your colleague is jumping BACK, rather then jumping forward, and executing more code per scan then the thinks he is.

If he's really concerned that JMP/LBL is a problem, put the conditional code in a subroutine, and only call it when needed.
I agree with RDrast.
200mS is wayyyy too long.
I've also done some large programs including motion and never got to see anything above 15ms or so.
Maybe he is "looping" through parts of code?
 
Our scantimes are normally 20-30ms but with customer insistence of hot stand-by's this can easily rise to 60ms or more.

Our process needs small scantimes, less than 50ms is ideal, the shorter the better.

The process is scan heavy but the idea is that not everything needs to be processed every scan but portions of the routines do need to be.

A typical PLC can contain 200+ routines, 100+ can contain scan heavy processing.

It appears that if you remove the 100 routines that did not contain JMP instructions, then the PLC scan (with jumps) drops to about 17ms, the whole program (all the routines) with the jumps removed (so all code processed) takes about 30ms, the whole program with sections of code jumped over, takes 200ms+.

The jumps do not go back, there is no looping.

It appears the number of routines effects the scantime when jumps are used!

and its a L62 I'm told not a 63.
 

Similar Topics

I am a beginner on the AB platform and I have a question. I have a 1769-l32e with scanner devicenet,I am presented with the offer to buy a point...
Replies
2
Views
1,714
Hi!. I have CompaqLogix 5069-L330ER project that I want to simulate with the Studio5000 Emulate. I usually create a copy of my project...
Replies
0
Views
898
Hi, I have a Compaclogix L33ER that from time to time loose the program. I haven't been able to find why is doing it. I am trying to make the...
Replies
2
Views
1,408
Hello, I am pulling my hair out trying to setup Modbus Master Read/Write Commands on a Prosoft MVI56E-MCM attached to a Clogixs 1756-L82E PLC. It...
Replies
1
Views
1,541
I use plclogix plc training /simulation software to get accustomed to the Alan Bradley Rslogix 5000/500 Platforms;it very good learning tool...
Replies
8
Views
2,166
Back
Top Bottom