Re: [Emc-developers] More ideas on Classic Ladder and HAL

[Ray H. <re...@up...>]

H Eric

I suppose that I should fess up and raise my hand here as the instigator
behind the existence of CL as a part of the EMC project.  This post will
offer a bit of that history and ask a couple of questions.  Please don't
take offense as I intend them to carry no emotional load.

My interest in "ladder" logic programming started not long after PLC's
showed up and began replacing RRL (relay relay logic)  I still have a
couple of 200+ relay logic panels out in the shop to remind me not to
ever go there.  After finding EMC and attempting a couple of machine
integrations with it, I found a real need for easy configuration of what
I like to call "machine logic."  PLC and ladder is one part of that
ability.  It is the PLC language of choice for machine tool builders.
(I'll explain the other parts in a bit because they can be related to
what you've started here.)

I watched the control.com list for several years as it spawned several
PC based PLC projects.  (We don't use the term softplc because that is a
copyrighted word.)  ClassicLadder seemed like the easier project to get
working with EMC and PaulC grabbed a copy and added it to the EMC
repository at SourceForge.  That really went nowhere and became dated.

JohnK and several other HAL developers had a long series of discussions
with the MatPLC and EU project folk developing motion control.  We
agreed that because of the way that real-time was being handled between
the projects we could not integrate their work with ours without giving
up the motion centric nature of the EMC project.  

IMO the HAL writers felt that HAL itself could handle logic elements
with a few AND and OR gates developed as HAL blocks.

Not long after that, PeteV picked up a M5i20 "anything io" card with 72
IO and decided along with others of us that he needed a logic engine to
use with it.  He grabbed the then current ClassicLadder code and created
what we have now for both ladder and motion code for this PCI card.  The
short answer to why CL is there in EMC2 is that we needed the ability it
offered and it works.  Thanks Pete.

All of the code in the EMC repository is a work in progress.  HAL and
EMC are evolving>  We expected CL to do the same.  One problem with
grabbing a snapshot or fork of a project like CL is that it will become
dated.  It is not easy to keep them in sync and since our interest back
then was to get a ladder logic engine that would work, and we have one
now, the priority to update that work is low.

The problems that you have identified and criticisms are quite valid.
It would be nice to bring this system forward to the point where it
looks better and works in easy concert with EMC, HAL, HALUI, HALVCP, and
other systems.  

You say, "I am also trying (and failing <g>) to resist the temptation to
start over without giving it a fair chance...." I note that you are wise
enough to temper this urge long enough to consider parts of the old code
as well. PLC is a pretty specialized rigorous thing that bears little
resemblance to a end user threadable set of logic gates.  My only
concern with respect to CL as we embark on this task is that we
carefully preserve the proven essence of PLC as we change things.  So my
question is, "Are you able to move us forward with the creation of an
EMC-PLC that we can trust?" 

After that question is answered in the affirmative then the XML approach
needs serious consideration and a path for expansion must be built in.
Rather than this beging just one more in a long string of configuration
files like parts of RUN, INI, VAR, TBL, HAL, it should move the data
from all of these into a coherent single configuration system.

My notions of machine logic stretch out to include all of the INI file,
all of the non-coordinate system variables in the VAR file, and a whole
range of machine parameters that control things like the nature of
operator door interlocks, tool change positions, preconditions and post
conditions for canonical commands, lathe vs mill and such.  I would very
much like to see these things included in a single configuration system.

My second question then is, "Is it possible to include an expansion path
as you develop your XML based system?  Can this file be made compatible
with and readable by common existing readers rather than only a text
editor.

Ease of machine setup and maintenance is my last concern.  I use the
user space ladder viewer to watch the state of the machine as I develop
a ladder program.  I know that I could start an instance of a
specialized HALVCP to show me the state of switches and relays on the
machine.  That display would be quite foreign to a machine tool
maintenance person.  The existing CL display is quite familiar to most
of us.

This leads to my third question, "Will your EMC-PLC have the ability to
display it's state while it works?"

Hope this helps.

Rayh



On Tue, 2006-09-19 at 08:41 -0400, Eric H. Johnson wrote:
>  Hi all,
> 
> It looks like I found out what the size limit is as far as posting to this
> mailing list. 

<bas that includes all of the substance of Eric's post so read back>