Enhanced Tracking and Guiding
With Dob-Driver II
Understanding The
Fundamentals
This
report should help you improve your understanding of critical functions in the
Dob-Driver II hand controller, mechanical subtleties in mount behaviors, and
how these relate to effective use of drive systems when you home in toward
“arcsecond territory”. Simply by
knowing some basics you can produce large improvements in both visual high-mag
tracking and in astro-imaging without resorting to major mount rebuilding or
contemplating the purchase of a “better” mount. Remember that this report is about extremist details, not about
“normal” enjoyment of using a telescope with a Dob-Driver. It comes directly from the engineers instead
of grapevine-political positioning or marketing rhetoric so common with
competitors on the internet. So if you
consider yourself to be the obsessive type – dive in -- read on!
Telescope motion is a
system: Electro-Mechanical
Accurate tracking in any
telescope system can be separated into two primary categories… electrical and
mechanical. This paper will discuss
them as such in separate sections.
An altaz mount is far
simpler to use than an equatorial mount.
It has no setup requirement like equatorial mounts and does not
even need to be level. Equatorial
mounts, on the other hand, almost always use worm-gears for control of the main
axes, tilt the tube at compound angles while tracking, experience polar error
similar to altaz mounts though to lesser extent, and end up with a host of
subtle dynamic deviations. So in the
end, though a Dob-Driver II will operate either mount method with very high
motor precision, each mount choice has its give-and-take, both mechanically,
and from the convenience of use perspective.
Driving with a serious
intention: What’s your vector Victor?
Any uniform velocity in any
given direction can be defined with only two points. This is a vector, and is the way equatorial mounts work –
with only one heading (Westward in Ra) required to counteract the planet
rotation. There are no other vectors,
just a slight occasional Dec axis adjustment, and the speed is relatively
constant such as sidereal, solar, lunar, planets – unless Mr. Atlas decides
someday to put a different record on and pirouette in a new fashion – the world
will turn the same for all! But
Dob-Driver II is NOT just a vector machine although it does that too -- it is a
trajectory machine! A trajectory
needs more than 2 points to define the curved path that is actually a
continuous +/ – change in velocity slope profiles.
So… let’s get started… show me the key elements!
The basic Dob-Driver
givens: 1) Since these use a stable-cut
quartz crystal, and the hand control microprocessor executes exactly 1-million
instructions per second with not much else to do but be totally focused on
motor behavior, we know the time progression to a very high precision. 2) Since the motor rotor itself is fully
digital, we also know the position to a very high precision at any given time on
a continuous basis. 3) Both of those
things allow us to set a velocity on both axes in any direction and maintain it
or bend it in a curvilinear fashion to our will with absolutely insane
regularity at anytime we want, even changing in thousandths of a second as
requirements dictate. 4) The only
remaining uncontrolled parameters are the operator precision in centering an
object (or an autoguider that is doing the same), and the mechanical
aberrations described in the next section following this one.
Entering Track Mode: The act of entering track mode identifies
the first point in the trajectory curve.
We call it the “anchor point”. Every
time you enter track mode it will start a new anchor point, even though it
continues at this time to use a previous tracking data set if you tracked
something (or even the same thing) before, until more data is accumulated. Prior track data is cleared when you have
recently powered-up and are tracking the first object, or when we detect the
telescope has been slewed to a new position that is largely far away from the
last position tracked, or you entered track mode again and new data is being
collected from your pushbutton adjustments while observing. Note that you have likely just used pan mode
to place the view where you want it to be before entering track mode.
Re-Entering Track Mode: What if I don’t like the view position – I
want to shift the view a bit and do precision tracking again? You re-position the object or tracking star
where you want then exit track mode (the T lamp blinks) then you enter track
mode again. A new anchor point is
established anytime you do this.
Fine-position shifting of an image is a lot easier in track mode than
pan mode so you see this is an important thing to know – try it on the Moon
sometime like I did when moving at 500x from one crater over to a mountain
ridge – you reposition the view first, then exit and re-enter track
mode.
Re-Centering in Track
Mode: Once track mode has been entered,
all subsequent corrections with pushbuttons or by an autoguider camera
device must return to the exact same position that was there when track
mode was entered originally, else the time/displacement calculation that is
supposed to adjust for tracking errors will not be correct. Note- an autoguider camera input is normally
used in guide mode, but can be also used in track mode, though track mode is
not recommended for imaging purposes it can be nice to track objects for very
long periods unattended. For the vast
majority of Dob-Driver uses, visual errors of re-centering an object or star
reference point to the exact location when track mode was entered is not very
important, since there are some very-fancy proprietary algorithms combined with
“fuzzy logic” principles that are there to deal with that. In the case of very-high-mag observation and
duration imaging however, accurate re-centering becomes more important. The Dob-Driver II will calculate, store, and
immediately use new trajectory data when it knows you are done re-centering –
this is 7 seconds with no button pushing.
Guide Mode: Move over to guide mode when the
magnification is very high, or you wish to guide astrophotos. It is somewhat similar to track mode but
with some very big differences for this purpose.
Entering Guide Mode: It will continue to use current “vector”
data obtained. There will be no new
anchor-point or curve calculations, as all is now relative velocity to where
you are currently. If you did not track
first in track mode then there will be no tracking vector data and this mode
will not be useable.
Re-Entering Guide Mode: No difference from the above “Entering Guide
Mode”.
Re-Centering in Guide
Mode: Very different from the track
mode though the operation and intention may seem quite similar. Any push of a button will alter velocity at
one of two speeds that is calculated by your setting in config mode (the T
lamp) while the button is held down.
Release of that button will then add or subtract an appropriate very small
amount of tracking speed in that axis, always less than +/-½% as calculated by
the proprietary fuzzy-logic. During the
button push on-time the speed of movement will be altered as described above to
“catch-up” or “fall-back” toward the intended position - there will be no
backlash jerks (absolutely zero backlash deviation) and the observed speed
change while holding a direction button down will be very subtle as set by your
program settings in Config mode – only discernible at high magnifications with
a crosshair reticle or on our Tech2000 autoguider units with a PC screen
connected. This is where the Dob-Driver
internal firmware really shines. Because
the drive velocities in each axis and the direction of changes dynamically and
independently adapt so that the duration or frequency of each following
correction will be less. In effect
– the motors will “evolve” to meet your mount gears and structural flexing
particulars in each short time frame like a child learning, then soon arrive at
the perfect velocity in every axis that is needed to obtain maximum tracking
precision. This is always a
“hunt-and-seek” function. It will
dynamically adapt to your mount, drive, thermal shifts, gravity orientation
shift, and telescope dynamics on the fly whether you are doing it manually with
a crosshair or an autoguider camera device is doing it. This is a very important feature typically
only available on major observatory mount setups.
Meridian crossing in Guide
Mode: In altaz mount perspective, any
object will ascend to Meridian, at which time the altitude velocity is
zero. Then direction will reverse and
the object will commence descent. The
opposite is true to the North when below Polaris – the object will descend to zero
velocity and then begin ascent. The
Dob-Driver II will realize that zero velocity is approaching and will reach a
threshold well before the zero point is reached and simply stop doing that ½%
velocity adjustment since it becomes meaningless to approach zero speed by ½%
at a time for each correction. The next
correction that reverses the motor will invoke a backlash move. Most of the time it is not noticed by a
telescope operator. But in precision
guiding I do not usually want to bet on the result and therefore avoid
exposures that might cross the meridian – I’ll wait if it looks that close
until it reaches meridian and then expose from there-on. When using a Dob-Driver II on equatorial
mountings – ignore this consideration.
OK – Did all the easy &
quicker direct stuff above with electrical controls and figured that out. Now I’m a control freak. Now what do I look for when that is not
enough to solve the remaining tracking aberrations?
The basic Dob-Driver
givens: 1) Since the digital motors
move at a steady velocity thanks to the quartz crystal clock, regardless
of load or friction changes, we can be sure the speed at any time currently
assigned to a motor will not vary until the firmware tells it to do so
at the appropriate times. 2) The
position resolution at any main axis on the sky in any instant of time is
calculated within 2 arcseconds or better.
With HRG motors this can go down to .5 arcsecond or better. So why are the remaining deviations from my
camera data showing larger than 2 or .5 arcseconds? The answer is in the mechanical structures between the main motor
rotor and the telescope main axis. How
about a good encoder on the main axis?
Find one that gives over 5 million encoder counts for one revolution
with a certainty of +/- .5 arcseconds and then we can use that. As long as you are willing to pay for 2 of
them, one for each axis, and are willing to accommodate such (very large)
encoders. Gearing encoders to increase
precision does not just add significant complexity or delicacy to mount
handling – it adds back-in the same mechanical flaw problems you wanted to
avoid in the first place! No wonder
large observatories cost many millions.
This does not even start to address gravity flexure of tube and mounting
structures. There are some bits of
knowledge and devious techniques to save you a couple million bucks – read on!
Convenience Day/Night
tests: On a new mount, particularly
larger ones that take time and effort to set up, we usually will set up at a
test location that allows close examination of behaviors over several days of
time. The philosophy being – why move
it out and set up each night and take it down again for each test? Also when the mount is moved each test run
then how could you be too sure about changes you have made to improve it? We will normally set up the mount
just-inside a garage door bay, a window, or door perhaps, that allows us to
focus “down the road” in daytime, follow stars above the horizon in the eve or
mornings, and simply close the door anytime we want to stop and leave the
premises without having to set it all up again. It’s useful when making changes then evaluating them at leisure
over a number of days time without disturbing the mount basic position or
adding lots of setup time under an outdoor sky. Does it pan smoothly, track smoothly, accelerate and decelerate
smoothly? No jerks? Backlash look good when motor direction
reverses? Look below for tips.
Quality dollars: Bowing to the makers of very fine mountings
and telescope equipment… there is no consumer mount made on this planet that
can hold arcseconds for significant durations even in highly-steady sky
scintillation conditions. Sorry to
inform you. It’s not just about perfect
mechanical machining (as if that were possible), but is also about temperature
and gravity effects on structural systems.
All imaging work except on bright objects should include guiding as a
planned prerequisite whether you stack or not.
Eyeball with crosshairs at high mag or via a Tech2000 autoguider. No human fabrication can be assuredly capable
of the required precision on its own mechanical merit and the required
emplacement details and setups themselves could be quite intimidating also with
no feedback such as “guiding”.
Worm gears: There is the “worm” that turns against the
“wormwheel”. The worm is helical in
design and this cannot ever be cut with precision – hence the PEC problem. Each rotation of the worm gear “worm” shows
variance that is repeating each revolution, so that can be profiled with
“Periodic Error Correction” training sessions.
But that is only trained on one tooth of the worm wheel, which contains
between 144 to 360 teeth or 10 to 4 minutes respectively. What about the (astronomically precise?) cut
positions of the other 359 teeth on that wormwheel? If you have been accurate to extreme with setting your PEC
training efforts, it will be of only some help on the next tooth and others
clear around the wormwheel since each will most-assuredly vary in true
position. Runout is another serious
problem. How “engaged” is the worm to
the wormwheel? The wormwheel
most-assuredly will never be perfectly centered on its hub – that’s the part
your telescope is mounted to. And it
also will certainly not be of perfect radius at every one of the tooth
locations. Better mounts use spring-loading
to maintain full contact of the worm at a near-nominal pressure as the
wormwheel rotates, some use the wormshaft itself as the spring-load method by
making it thinner and/or longer but we do not condone this (though its better
than nothing!). Humans made these
things and they will never be error-proof.
You must guide out these errors.
Balance: This is the most common problem in telescope
accurate tracking both on Altaz and EQ mounts.
Make absolutely sure that both axes are always in balance prior to
imaging at the present angle of the telescope tube. Many experienced imagers take special care to check that the
drive motor is the thing in-charge, such that imbalance will not pull the
telescope ahead of the motors’ intention on that axis, and that not so much
opposite imbalance will hold the drive motor back from doing its job to drive
forward accurately and uniformly. There
is another technical paper on our site that delves into balance analysis separately
in far more extreme detail.
Belt drives: The belt can creep-along on a slip pulley
surface when tracking. Not easily
visible when looking at the belt even when slewing (and it can be no problem
while slewing but is when tracking due to static friction of bearings). Balance can be a problem with this showing
one way is OK and the other way is not.
Mark a pencil mark or use masking tape to monitor the belt creep when
this is suspected. Rub some violin
rosin on the belt to reduce creep or adjust instrument balance to suit. The best solution is to reduce bearing
friction or imbalance that causes the belt creep in the first place. Another common aberration with belts or gear
teeth and shafts is the interplay between belt microstretch or gear tooth flex,
shaft torsion, and bearing friction.
Teflon pads (Dobs) and axle shafts (ie- giro) at slow speeds can have a
tendency to grab when the static friction is too much for the load being
carried – then release when belt or gear tension, or driveshaft torsion,
becomes sufficient to pop it free. You
will see this ‘pop’ at high mag with a crosshair or on a camera image when a
target slowly backs off from the intended guide point then suddenly lurches to
catch up. The next part about bearings
gives some solution pathways for this.
Bearings: Teflon, axle shafts, rollers, load carried,
and their related surface materials all play a part in microfine motion
effects. Though in industry the usual
solution is plenty of oil and grease which can make your “Tin-Man” seem to move
about OK, on telescope bearings this is usually not acceptable since wet lubes
will pick up dust or lint or particles or at the very least for exposed
surfaces like altitude bearings on Dobs there is always the risk of incidental
contact and distribution with fingers or other objects used during operation or
transport. With Teflon pads you should
at least compress fine sandpaper between and move the parts so that the pad
surface will have the high areas removed and thereby obtain a better psi
distribution for the load carried at each pad.
Then score the pad face with a utility blade scraped sideways in a
cross-hatch pattern to allow release of air molecules that will typically form
a vacuum lock. On axles, like Giro
altitude shafts for instance, we drill the plastic pressure-slug that is under
the shaft tensioner knob and inject some grease periodically into the bolt
threads there. Then when the tensioner
bolt is screwed in again it pressurizes the grease to flow from the center of
the shaft outward toward the ends of the bearing sleeve. Simple thinking in these ways can pay off
large dividends to get around mechanical flexures in drivetrain components and
eliminate or strongly reduce those ‘popping’ releases from static friction. Commonly referred to as ‘sticktion’.
Dob Azimuth Rollers: When using rollers instead of Virgin Teflon
to support a Dob structure, they can exhibit the “pie-dough problem” that
produces a ‘pop’ or small-scale lurching like described above for belts. When rolling out a pie dough in the kitchen
you always notice that the rolling pin does not want to move in a curved path –
it really wants to go in a straight path.
The wider the roller the more true this is. Oh they will certainly move and it will not be obvious in manual
moves or normal tracking, but the details when it comes to arcseconds of angle
can be much more subtle when guiding.
Sometimes you may hear a ‘pop’ occasionally on a heavy scope when it is
quiet late at night, then it will repeat after some seconds. It is not the drive motor but is instead the
slo-mo screech of the azimuth roller on the formica, etc. surface. This problem will disappear by rubbing some
graphite (pencil lead) on the path of the rollers, we use the Tech2000
“Micro-Glide” lube, or I have also used a very thin film of grease on the
roller path as it dries up soon and does not collect grit and dust particles
yet remains present as a film like the dry graphite but it must be re-applied
occasionally for imaging precision work.
Dob azimuth pivot bolt: On more rare occasions the azimuth center
pivot bolt could introduce some influence on fine guiding. If it is wobbly and tilting around then
there can be positions where it wants to bind up the free rotation, and result
in a ‘pop’ sort of action again due to pressure-flexing and sudden release of
wood mount structures -- more so than the actual slipping of the azimuth
traction drive wheel instead that would occur from severe pivot bolt
lockups. Not that this needs to be
agonized over with exotic precision shafts or any such thing. I have readily got away with nothing more
than wrapping the bolt threads with Teflon pipe thread seal tape from the
hardware store to provide a more vertical snug fit and lowered friction, in the
least limiting the screw threads from gouging into the wood bolt hole and
making things worsen over time. Its not
hard to scheme up a better and more rigid bolt-pin-stud mounting method with a
bushing on the free turning part anyway as we have done that too -- many times. There is one interesting story where at
Astrofest in Illinois I had a customer approach with his arms folded sternly
and he flatly stated “It don’t turn”…
“Hey let’s go see!” As it turned
out, it was a (heavy yes to be sure) Mead 16 Dob. A quick examination revealed that the scope could not be moved in
azimuth, even by hand, even pushing hard enough to tilt the (very heavy) mount
off the ground! Loosened the pivot bolt
nut and all worked fine. He had it that
tight! The lesson is – decide which
axis is the problem and look for simple mechanical corrections.
This is only a rough glance at the large world of
precision astronomy. No doubt I have
missed something or you need embellishment on some topic line. Since this broad content could be expanded
into book size it is probably best at this time to keep it a bit general and
broad-brush.
Please direct questions or suggested expansions to
Perchance then on a clear and quiet night, when less
burdened with the paltry circumstance that this world proffers upon our own
human nature, we will speak again of these means that we hope… should
ultimately expand our mind.
Dave Masters / Tech2000 www.tech2000astronomy.com
~ Political Election Adden-dumb ~
Paul:
“Quantum positions of two opposing points could lay indeterminately with
respect to one another.”
Dave: “Two
opposing ‘party-cles’ place points that always ‘lie’ indeterminately without
respect to one another.”