Re: Re: 3 axis controller

[Ray <rehenry-at-up.net>]

On Fri, 06 Apr 2001, John wrote:
> Ray, thank you for your input.  How I interfaced the PLC to the Mektronix
> PC based control was quite simple.  The entire tool change logic was
> contained in the PLC.  When a tool change command  is given from the
> Mektronix PC, a T command TTL logic is outputted to the current sinking
> input of the PLC (isolated input) along with a strobe command.  The PLC
> instantly outputs a feed hold command to the PC (Mektronix board in the
> PC) at which time the tool change commences.  When the tool change is
> completed, the feed hold is removed and the program in the PC continues
> to the subsequent blocks thus proceeding to machine until a new tool
> change is called for.

Well the interpreter has the ability to issue those kinds of cannonical
commands to NML.  Since the tool changer is basically a stand alone unit
those minimal signals could be implemented as a part of tkio rather
easily.  Fitting the signals onto an I/O port someplace might be the
biggest trick. 

There will be some problems with spindle orientation if it is a motor
thing.  However if the spindle controller has an orient command pin and an
in position signal you should be okay to take those from your PLC.  My
hydrapoint activates a solenoid that rams a dog into the decelerating
spindle and uses a limit switch to see that the dog has found it's place.

What I would do to start is list all of the auxiliary I/O that you need. 
Keep the limits and homes grouped with each axis.  All the rest should be
listed along with any timing and logic that might be required between them. 
Be sure to include stuff like the nature of estop both internal and
external.  Lube, coolant, any other solenoids, brakes, lock pins, and all
of the sensors on the machine.  

One of the shortcomings of the EMC the way it is right now is that there is
no adequate spindle speed system.  The interpreter issues speed commands.
Manual buttons and language exist to start, stop, reverse, increase and
decrease spindle speed but we don't have a speed reader and a system to
ramp and hold the spindle to the desired speed.  There was some of this in
the early versions but the code would need to be updated and an input
system worked out for the spindle tach or encoder.

> This simply requires that the PC based control has provision for
> outputting a tool (T) strobe command along with a BCD or binary tool
> number command in TTL logic.  Then the complex logic of the tool changer
> can be written in an external PLC.   Of course this same functioning can
> be done internally in the PC, but it would require a fairly powerful and
> user friendly software PLC.  A very nice feature of a PC based control
> would allow for user friendly input/output commands such as these, and
> also some spare M or miscellaneous output commands to do special machine
> functions that may arise. 

Extra M functions like M45 for an indexer would be really nice on some of
these older machines.  I understand that Tom Kramer, the interpreter
author as made it fairly easy to add this kind of code, but I am no
programmer so if I tried it, I be about as successful as a cow trying to
stand on one leg.

 I hope this makes sense.  Does the emc allow
> for these features. 

What you are asking makes a great deal of sense and yes most of it is
available with a little effort.  I don't believe that these kinds of
changes will make it on to this weeks list of NIST to do's but the EMC is
moving more and more toward being applied to complete milling centers and
not just hobby or retrofit machines.  What is there right now is an awesome
base from which to build.

I guess my bottom line advice to you is, if you need all of it this week,
buy the control.  If you have the time and can work with like-minded people
converting similar machines it can and will all be done.

Even I could add the manual tool change popup to tkemc.

Ray