ANOTHER CRAZY IDEA

Hi, here is another crazy idea:

For some time ago, I described how you could wire up an old disk
drive as a stepper test rig. I have received reports off list,
that several has had this working. You could add some L/R with
power transistor and drive large step motors as well.
At the moment, I just can't remember who came with this, - but
someone wondered if you could parallel this with the step motor,
and see if you had lost steps.

Here is a few ideas:

If you could run a linear slide with a small step motor parallel
to your table, then you could use some opto sensors to see if you
where off by some steps, and correct it.


Here is a another crazy idea:

Mount a little second step motor on the back of the main step
motor, but let it run freely. Assuming you can get to the main
step motors axle, then mount a disk to this, and a disk to this
little freewheeling step motor.

Something like this ( a will use a "." just to have the
characters used come out even on each line. Just pretend they are
not there )

.+---------+...|.|
.|.........|...|.|...+-----+
.|.........|...|.|...|.....|
=|....A....|===|.|===|..B..|
.|.........|...|.|...|.....|
.|.........|.L<|.|>N.+-----+
.+---------+.M<|.|>O


A = main step motor
B = freewheeling step motor

Between them is two disks ( I will comment on these in a minute )

L,M < is light sources
N,O > is sensors.

Each disk is a wheel with "sprockets",with equal divisions for
dark and open areas.

Let us give the disk on the main motor the name of AA, and the
other BB.

If you could have opto sensors on One of them, rotating with the
disk, then this setup should
work. But it would be a bit difficult to arrange the electrical
connections. So how could you make
the sensors and light source stationary ?

Well, If you attached a polaroid filter on each disk, 90 degrees
change on the inner and outer track.
Then a lead or lag could be picked up, and with counting pulses.
This could then be used to correct the
lost step.


Here is another idea:

If these disk was made with a dark and reflecting surface, then
you could use opto proximity detectors.


Here is yet another idea:

..
......|...|....
.....<|.0.|>...

This is a little part of the outer circumference of the disk. The
light source "O" is put between the
disks, and "<" is the opto sensor. What will happen ? If disk AA
slows down compared to BB then you will have a quadrature signal,
just as a normal encoder. This would probably be the best and
easiest way to implement this kind of sensing. You would have
direction and pulses lost. You could in fact use the wheels from a
mouse, or something else, as this one don't need to have the same
index count as the steps/rev on the motors. Some circuit would
just insert enough step pulses as needed to align these again.

You may also have a counter, let say a 4017 or something. This
could take 10 counts, on the 11th, it would stop the motor before
any real damage is done. I don't know, there could be many way to
implement functions
here.

Well, just another wild idea.

//ARNE



[Non-text portions of this message have been removed]