MC / MCR in detail pls!!

[asmdude]

The MC / MCR is listed by mitsubishi to use for high efficient ladder switching. So, I would expect it to skip devices that are not active due to a resetted MC. But then, I read the following:
"The devices between the MC and MCR instruction are even processed after the input condition of the MC instruction is reset. Therefore, the program scan time in this case is not decreased."

That puzzles me, why is this so "high efficient" if the instructions are still processed even when the MC is reset? Or do they mean that the instruction are processed ONLY for the first scan after the MC is resetted??

Does anyone know that??
Michel

[Goody]

MC / MCR is listed by mitsubishi to use for high efficient ladder switching

Not sure what you mean by ladder switching. The MC (master control) and MCR (master control reset) is used for virtually turning off a section of the left hand bus rail between the MC and MCR instruction.
This disables all outputs within this section. (there is more to the MC instruction but that is what it is foe)

[asmdude]

Originally Posted by Goody

Not sure what you mean by ladder switching. The MC (master control) and MCR (master control reset) is used for virtually turning off a section of the left hand bus rail between the MC and MCR instruction.

Hello Goody, "ladder switching" is a term used by mitsubishi and it does exactly what you just described. The part I dont understand is why it is called "highly efficient". A conditional jump for example is efficient because it saves execution time when the jump instruction is active. So, I would like to know, do I save execution time when using the MC / MCR, and if not, then why is it called highly efficient??

[rdrast]

That is incredibly stupid terminology.

They probably claim that an MC/MCR pair is [Choke] "Highly Efficient" because it doesn't actually scan the logic in between the MC and MCR instructions if the MC is false; rather, it just sets the outputs off.

Don't get hung up on words... Or, for that matter, the ethereal "Efficiency". Write a program that is safe, and performs the desired task. Only if you see that you have a scantime issue later, should you try to optimize it (and probably not even then).

[asmdude]

Originally Posted by rdrast

Only if you see that you have a scantime issue later, should you try to optimize it (and probably not even then).

Yes, that is exactly my point, I actually do have a scantime issue... So you suggest to "just leave it that way", you wouldn't worry much about a scantime issue?

[Steve Bailey]

I can't speak from experience with Mitsubishi, but on the PLCs with which I am familiar, using the MCR instruction minimizes the difference in scan between when the function is active and when it is not. In other words, if the scan time is X milliseconds when the logic controlling the MCR is true, it will still be X milliseconds when the logic controlling the MCR is false.

What is your "scan time issue"?

If your goal is minimizing scan time, then you should be using JUMP instructions or subroutine CALLs with logic controlling when to JUMP or CALL. If the scan time would be too long if all of the JUMPs were false and all of the CALLs were true in a single scan, then your challenge is to design the logic so that doesn't happen, yet still execute the subroutines (or the Jumped rungs) often enough to keep the process under control.

Also, remember that in the absence of an instruction turning it on or off, an output will retain the last state it was told to be in. That means if you don't call a subroutine or if you continually execute the JUMP instruction, outputs turned on or off by the rungs in the subroutine or in the JUMPed zone will not change their state.

You should also read the documentation for your PLC to determine what happens to timers within a JUMP zone or subroutine.

[panic mode]

I agree, if you have problme with scan time, you can try skipping
parts of the program or call parts of program conditionally.
personally I would also rather use CALL Pn and FEND
but this has to be verified for particular CPU.
I'm not sure if they all support it (anything reasonably recent
should). what PLC and programming software you have there?

[asmdude]

Originally Posted by panic mode

I'm not sure if they all support it (anything reasonably recent
should). what PLC and programming software you have there?

I am using medoc with an F1-30 (do they still work? yes they do....) :-). It supports jumps, but no calls... I actually thought that the MC/MCR worked similar as the jump with the only difference that it would disable the outputs in case the MC is false. I don't have the docs of this dinosaur so it's a bit of a guessing game...

[bob1371]

Originally Posted by asmdude

I am using medoc with an F1-30 (do they still work? yes they do....) :-). It supports jumps, but no calls... I actually thought that the MC/MCR worked similar as the jump with the only difference that it would disable the outputs in case the MC is false. I don't have the docs of this dinosaur so it's a bit of a guessing game...

If your a member of www.mrplc.com you can download manuals from thier site.
http://forums.mrplc.com/index.php?dlcategory=20
or if not, just sign up. They have both the F1 hardware and programming manuals listed.

Bob

[asmdude]

Originally Posted by bob1371

If your a member of www.mrplc.com you can download manuals from thier site.
http://forums.mrplc.com/index.php?dlcategory=20
or if not, just sign up. They have both the F1 hardware and programming manuals listed.

Bob

That is exactly what I needed! Thanks bob!