Re: What a mess 4

["Arne Chr. Jorgensen" <instel-at-online.no>]

On Tue, 25 Jan 2000, you wrote:

>NIST in EMC works within the idea of no hard-wired toggle switches, pots,
>or other devices that hold emc in a specific state.  They do this so that
>all aspects of machine control can be made to work over a network or
>distributed control system.  
>

Yes,  I know - and I do like that, would not be without that, but on the other
hand:

Let say a mill, - you have to talk to the outside world. You do this with the
motors, and you do it with the keyboard.  So I don't see the big deal. 
But as far as the machine I have operated,  a PC keyboard would not do good
in such an environment.  You also would like some easy control buttons.  You
may even like some "big" ones, so that you would be able to work with gloves on
your hands :)  


>The RS232 idea really appeals to me because you could feed it with a wide
>variety of hardware.  A couple of chips and you could watch many momentary
>buttons and feed the results to a serial in. The rt routine would just have
>to change the state of some shmem bits and I think that we could get access
>to the information with the new tkio stuff.  

Yes, but I don't see why you can't make use of the game port too.  This is the
fastest and most easy interface to experiment with.  The kind of control, and
panel,- would in the end be up to the needs you may have in a specific machine
design.  

Make it all like modules,  you just add the ones you like.     

>My thought on panel lamps was that you'd read the result on the computer
>screen but you could make the RS232 a two way thing.

Yes, point taken. But again - there are other options.  

But, we are all at different levels here. Some have a machine running - much as
they want to,  other may have some very different uses.  They might have to
experiment a bit, - see how it is possible to make it work on their setup,
according to their ideas.

You might like some cheap way to fiddle with it, - before you do a custom
design.  And this may bring us to -

PLC:
---   

I am not sure what Fred meant at all here ?

1.  A PLC syntax configurator of the digital port ?
2.  Interface with common PLC system components ?

Let us just take the last possibility - would this be with a RS232 link ? Use
some, - common PLC control units ?
This could be nice, and a lot of the functions we have been talking about could
be done this way.

But then again - if you remember the stuff a called "Front-input" device and
"Back-input" device ?

What I meant with this, - is that the front-input, is from keyboard, network,
RS232, etc. - coming into the interpreter on linux.

Back-input on the other hand, - here I meant on the back-end of the control
system, the real time tasks.  A game board could be perfect for "playing" with
this, that is - it will run synchronous with any hardware.  Let just say you
wanted RPM monitor and control over a spindle. It should be sampled at a rate
with the other tasks. A "front-end" device would not be able to do this.

Here is a little other use of it. Wire a - say 16 to 4 decoder to the joystick
button switches.  You have 16 digital inputs. On a stepper system, this could
be limit switches, and a realtime task could be stopped at the single step
output.  This is not possible in a "front-input" device.

You could also "hack" the game board, and have outputs too. The thing is that
it is cheap,  available, have a reserved address space.  


By the way, - when I said an operator panel, - I did not mean that you would be
done with the keyboard - just a switch - and you have the control where you
would like it to be.

 //ARNE