RE: Close loop steppers

I believe that several aspects of Stepper and DRO technology may have a
relationship which can address problems with each other. Obviously we have
had large discussions on both. A major concern with steppers is lost steps
and a major concern with some ideas for DRO's is accuracy, resolution,
speed, cost and repeatability.

Consider an approach. A black box to interfaces 3 stepper motors and 3 DRO
inputs (X, Y and Z). The box has some smarts and functions to solve two
problems. First, many of the ideas I have read and had for DRO technology
look good for at least four out of five (accuracy, resolution, speed, cost
and repeatability) but often misses in one catagory or another.

By supervising the Steppers with the DRO channels, it is ovbious that missed
steps can be detected, however consider the value of the steppers in
calibrating (on the fly) the DRO technology employed. In other words, take
idea A, B or C for DRO technology. Every step should cause a change in DRO
signature. A full step should result in a fully articulated DRO signature.

To illustrate what I am getting at, Let us say you ran a program through the
air without anything on the table and without a tool in the collet. You
might expect that you will NOT miss any steps and that for every step the
DRO signature was meaniful (whatever DRO method you like). From one point of
view, the DRO sinature is now calibrated to the stepper signature without
load.

As you push the concept, many machine dynamics come into play from
vibrations and so one, however, let us say that we only push the concept
enough to acheive two simple objectives. 1st to calibrate (continously) DRO
technology and second to detect and recover from missed steps.

To accomplish this, a micro controller within the black box, cooridinates
the moition and validates the motion with the DRO data and visa-versa. This
will preserve the open loop nature of stepper drivers from the stad point of
the computer. If the black box can recover from the lost step, things may go
on normally. If the black box can not revover from the lost step, things
halt.

The recovery attempt is simple enough. Pause the other motors from advancing
(if they were called to do so) and attempt to move the stalled motor. If two
or more motors were called to move and less than all moved, an error has
already occured and you have to decide if it is serious. When a single motor
is to move and it does not move, recover is striaght foward. If the stalled
motor is moved on the second attempt, continue cooridinating motion. If the
stalled motor remains stalled, shut down (abort).

An odd example of the value of the stepper to the DRO component may serve to
illustrate how the stepper may validate and calibrate the DRO technology. If
you wanted to, you could fabraice a home made DRO with a small stepper
motor. The stepper motor can drive something (Magnet, LED, whatever) and a
sensor mounted on the moving object (table) can provide an analog signature.
The stepper moves to obtain the same analog signal (or singals)to identify
where the moving object is located. The best technology for analog tracking
is differential in nature. That is, one sensor is moving away and another is
moving toward and the difference between the two will calibrate out any
error (for the most part). The reason this is significant is that magnets,
LEDs and other devices loose their strength over time. If the loss in
strength over time has an equal effect on two sensors where you take the
differece between the two sensors, you can avoid the problem.

In the spirit of the present discussions, the analog or digital signal from
the DRO technology only needs to show reasonable cange to validate the Step.
In this spirit, the DRO technology need not be very increadable at all. In
fact the simplest way to verify that a stepper motor has actually moved is
to use another (smaller) stepper motor as a signal generator. If you connect
a large stepper motor to a smaller stepper motor and connect the smaller
stepper motor to a scope you will see a bi-directional pulse for every step.
It is best to use a smaller stepper motor with the same number of steps per
revolution for this. This does have some limitations however. If the large
stepper motor moves foward and bounces back, the firmware reading the
signature from the smaller stepper (generator) must recognize this
sophificated fault.

The usual challange with the old idea factory is getting something
constructed which gets put into service. Some of this involves serious
electronics. Perhaps there is something for someone in these thoughts. Once
you go too far, you're almost back in the servo area again.

Perhaps one diamond in the rough here is the Stepper driver with feedback
from a second smaller stepper. The second smaller stepper is cheap and
solves the first order of problem of completely missed steps. The micro
controller code is pretty easy. For one thing the Micro controller program
can initiate a learn pass for each axis of motion and will recognize digital
or the polarity of analog feedback for every step between the limits of
motion. You know, Record what it should look like, and freak out if it
doesn't look right.

David Howland
dhowland@...




-----Original Message-----
From: root@... [mailto:root@...]On Behalf Of Arne
Chr.Jorgensen
Sent: Wednesday, December 15, 1999 11:44 PM
To: CAD_CAM_EDM_DRO@onelist.com
Subject: [CAD_CAM_EDM_DRO] Close loop steppers


From: "Arne Chr.Jorgensen" <instel@...>

Hi,

I have presented a few ideas about a simple closed loop kind of
stepper system, - in my "crazy stuff" inputs. I have a few other
solutions, but would just like to hear if any found this
interesting at all, and if someone has some comments. My idea
was just some stupid simple solutions, better designs can be
made. It all depends if anyone think this could be of any help.
One of the things I had in mind with all of this, was just a
dissable function if the lost steps grew out of hand. But I don't
know how you opereate steppers. The way I have done it before, did
not involve any real variable added loads, from the tool. So I
just wonder.

//ARNE



--------------------------- ONElist Sponsor ----------------------------

ChickClick.com - girl sites that don't fake it! Free email with Sass.
Tons of hip, free stuff too.
<a href=" http://clickme.onelist.com/ad/ChickClick ">Click Here</a>

------------------------------------------------------------------------

Welcome to CAD_CAM_EDM_DRO@...,an unmoderated list for the
discussion of shop built systems in the above catagories.
To Unsubscribe, read archives, change to or from digest.
Go to: http://www.onelist.com/isregistered.cgi
Log on, and you will go to Member Center, and you can make changes there.
For the FAQ, go to http://www.ktmarketing.com/faq.html
bill,
List Manager