Re: [Fwd: ATM steppers & microcontrollers]

[Mel Bartels <mbartels@efn.org>]

Jack Bindle's design of microcontrollings operating steppers with a
current limiting circuit is an excellent concept and worth close
attention.

My system does call for individual adjustment of each microstep's
voltage.  This is typically done with a lever attached to the motor
shaft and a protractor.  The software has a test mode where individual
microsteps can be moved through so that you can increase or decrease the

spacing between the microsteps.

One advantage of this method is the handling of heavier torques.  A
stepper motor's rotor will deflect 0% when directly on a winding and
deflect by the % of torque loading when inbetween windings.  So, if the
torque loading is 20%, then the rotor shaft will be accurately
positioned when on a winding but 20% behind when inbetween windings.  By

having control over the rotor's position at each microstep, one can make

up for the extra molasses, so to speak.

Implicit in most amateur computerized designs is overbuilding to conquer

certain problems.  For instance, on my 20", the torque required to move
the scope is very minimal, about 1 in-oz.  I expended some effort to get

the shaft that the worms fits on that came with my surplus Byers gears
to move as freely as possible without backlash.  So, torque loading is
not really an issue for this scope.  It also means that I can reduce
current going to the motors while tracking to a very minimal level,
indeed, 0.1 amps at 12 volts.  Very tiny steppers can be used, and I
have heard of the steppers in floppy drives being used to drive average
sized amateur scopes.

Stepper motors have tolerances in their construction that limit
microstep accuracy to about 1/20 of a fullstep.  More microsteps per
fullstep might mean smoother motion but no greater accuracy.  For smooth

motion, experience says to keep microsteps no greater than 1/2
arcsecond.  This is not only for absolute positioning accuracy on the
sky, but also for vibration's sake.

Most amateur scopes vibrate at a frequency of several Hertz.  You can
see this in the eyepiece at high power by giving the upper end a rap of
the knuckles.  If the motor step rate matches the scope's natural
frequency, the scope's motion will be jittery.  So, a jittery image at
high power when tracking can be caused by too coarse of a step, and by
inadvertently matching the motor stepping with the scope's frequency.

By adding more weight to the upper end, or by using softer materials, or

by extending the length of the upper end, one can get a scope to have a
lower vibrational frequency.  This is usually bad news because the
damping time is lengthened, and can even reach the point where modest
wind gusts can keep a scope in a state of perpetual vibration (I have
seen this).  A higher stiffer frequency is desired, but only up to a
point, because if too stiff, the scope will resonate with the higher
frequencies found in stepper motors.

My 20", when rapped on the upper end, has a quick jog to one side then
back again, of about 10 arcseconds.  This occurs in about 1/3 second.
The damping is excellent on my 20", and is because the altitude bearings

are rather large, lowering the height of the rocker box's side height,
making a shorter path from the bearing contact points to the ground.
This is my 3rd iteration of this mount, so one (at least me) cannot
always hit it right on the first design.

So, a stiffer mount means a quicker vibration that die out quicker.  You

can get vibrations to dampen very fast by judicious use of material.
One excellent but little used idea is to place very thin rubber under
the teflon pads.  The rubber strip stops the transmission of the
vibration through the mount.  A mount vibrating actually consists of a
wave that travels from scope component to scope component.  The tube
assembly will transmit its vibration to the rocker and base, which then
reflect the wave back into the tube assembly.  Slowly this dies as the
wave is absorbed by vibration stopping material like rubber.  Putting
rubber between the feet and the ground also stops the reflection of the
wave back into the mount.  You can tune the rubber to best stop
vibrations in your mount by trying different thicknesses.

So, with some care you can turn a scope of good vibrational
characteristics, namely damping in 1 sec, into a great scope that damps
in 1/2 sec, and a so-so scope that damps in 2 seconds into a good scope
of 1 sec.

With a stiff mount you can contemplate correcting for wind gusts like
the pros.  In fact, this already comes up in my control program thanks
to the higher count encoders now available.

For no extra cost really, one can get encoders that count to 8192 counts

per rev from US Digital.  And interfacing boxes like David Lane's MGIII
and the new Bob Seagrest unit can handle up to 8x counts if you gear
down the encoders with a gear or a belt.  8192 counts translates into
2.6 arcminute resolution and 8x that is about 20 arcsecond resolution.
Probably greater resolution is meaningless unless you go to the trouble
of software correction of encoder misalignment and other mechanical
issues.

When the wind gusts strongly, Chuck Shaw's 14" actually resolves the
changing scope position thanks to the wind pushing the scope around
enough that the encoders register the change.  The mount attempts to
keep up, but since the software was not written for this fast of
repeated response, the mount doesn't.  Perhaps in the future I will
allow the software to do rapid updating and motor corrections for wind
gusts.

Now that encoders can resolve very tiny angles, the issue of mount
accuracy comes up.  Many mounts have misalignments in the 3 principle
planes of: 1. altitude vs azimuth perpendicularity, 2. optical axis
misalignment in the altitude direction, and 3, optical axis misalignment

in the azimuth direction.  These errors are often 1/4 degree to a degree

or more, so they are a significant barrrier to accurate finding.  You
can buy Bisque software for $250 or Chandler's program that compensates
for these errors.  Bill Gray of Project Pluto Guide and myself are
working on our own version of mount alignment software.

We talk alot about optics, and that's great, but I would sure like to
hear more stuff about experiments and tips on mounts.  I know that
adding tracking to my 20" has enabled me to see about as much as I used
to see through a 30", so I view mount enhancements the equivalent of
adding aperture.

Well, enough rambling!
--
Clear skies, Mel Bartels    Programmer/Analyst, amateur astronomer
Eugene, Oregon, USA         homepage: http://www.efn.org/~mbartels
mailto:mbartels@efn.org     atm, atm-digest list-owner
Motorize A Dob: http://zebu.uoregon.edu/~mbartels/altaz/altaz.html