Jerry Pinter's construction notes
Thanks for your compliments and interest in my scope. First of all,
I need to acknowledge and thank Mel Bartels for his generosity and
patience in providing his system plans, software, and support. We had
a daily e-mail dialog for about 6 months while we worked out all the
bugs in my scope and even a few in his software. I had severe backlash
(about half my CCD field of view) which rendered my slewing/finding
ability useless, and after many unsuccessful iterations Mel came up
with a great backlash compensation feature that completely removes
any backlash no matter how large. Now my ol' light bucket is a pleasure
to use and can slew, find, track and image with the big boys!
Before I answer your specific questions, I'm forwarding you a
general description of my scope which addresses most of your questions:
---------------------------------------------------------------------------
My scope is a Newtonian reflector with a 13.1" f/4.5 primary
mirror and 3.1" secondary mirror. The optical tube is supported
by an altitude-azimuthal ("Alt-Az") Dobsonian-type mounting.
A 16"-diameter fiberglass tube is held by cork-lined plywood
split-rings that gently squeeze the tube to hold it in place. The
rings can be easily loosened to allow the tube to slide back and
forth if necessary. The tube is counter-balanced by lead weights.
The tube rings have two 8" circular wooden disks that serve as the
altitude bearings, the 8" disks (one on each side of the scope) have
an aluminum ring for the bearing surface. The tube/ring assembly
sits in a rocker box (which is somewhat similar in design to Coulter
or Orion dobs). Instead of teflon pads which were great for visual
observing, I replaced these with 2 ball bearings on each side for the
altitude axis. Thus the aluminum-surfaced 8" wood altitude disks sit
directly on ball bearings. The result is an almost frictionless
altitude axis (very sensitive to balance and thus practically unusable
when the altitude drive is not engaged). On one of the 8" altitude
disks, an 11" wooden disk (concentric with the 8" disks) is used as
a pulley for the belt-driven altitude 200-step motor. The stepper
motor has a 150:1 reducing gearbox with a 1" toothed pulley. Thus
the altitude drive has a final reduction of 11x150x200. A clutch
can quickly disengage the drive by loosening the belt for manual
movement (the motor housing has an adjustable spring-loaded tension).
The original drive apparatus is from "Dob Driver II" by Tech2000
which sells their own dob drive system (I am using their motor
housings with release-clutches, gearboxes, ball bearings, pulley,
and toothed belt.) The azimuthal axis of the rockerbox is made from
an axle-bolt with brass shoulders (no side-to-side slop). The box
rotates on 2 ball bearings (again, no more teflon pads for very
low friction). The third support is the 2" knurled wheel of the
azimuthal stepper motor gearbox. A clutch can quickly disengage
this wheel and let the rockerbox rotate on a teflon pad for manual
movement. The ball bearings and motor wheel roll along a formica
sheet bolted to the bottom of the box. Four rubber wheels underneath
the box allow it to be rolled around. When I set up to observe, I
drop three 5" carriage bolts through holes in the baseboard which form
three legs that the mount stands on. This provides a stable base on
unlevel ground. The base does not need to be leveled since the
drive software automatically compensates for any tilt in the base.
The tube/ring assembly simply lifts out of the rockerbox cradle,
so it is easy to transport and set up. I have a Telrad finder and
a cheap 60mm refractor (with a 12mm illuminated crosshair eyepiece)
that I use for aligning alignment stars.
The scope is controlled by PC using Mel Bartels' software as
explained on the Internet at http://www.efn.org/~mbartels. The
software uses the popular 2-star alignment method which initializes
the software to the sky without the need for polar alignment or
leveling of the base. In the software I have object libraries of
250 bright stars for alignment and a 5000-object library of NGC
and IC deep-sky objects. The scope can slew to objects from the
libraries or to any input coordinate position or by using the
handpaddle pushbuttons to manually slew. The scope automatically
tracks wherever it is pointed and displays it's position in both
RA/DEC and ALT/AZ coordinates.
The CCD camera is the Cookbook 245 with the Low Dark Current
modifications in place. I have replaced the water-cooling system
with a forced-air convection cooler built by my friend Kurt Mihalco
of Coherent Systems. His cooler is nice and compact, and replaces
the bulky plastic tubes, copper coil, bucket, and noisy pump.
I also use Kurt's custom power supply for the camera, it runs off
of a single 12V battery. I use this single 12V battery to
power both laptops, the scope, and the camera (about 6 amps total).
I use a 95 amp-hour marine "deep cycle" battery that provides
power all night while I'm out in the desert or on Palomar Mountain
where I usually observe.
------------------------------------------------------------------------------
Q. What did you use for the az bearing(s)?? A flat metal plate like Mel??
If yes, please describe, along with where you found it and what cost.
As described above, I use a piece of a formica countertop with a
"pebbly" surface. It was a bit of a challenge to get this as
flat as possible, which is important for slewing accuracy. I ended
up using numerous bolts to squeeze the formica down to the plywood
underneath, and then "tuning" the bolts to get it as flat as possible.
A rigid flat metal plate would be a big improvement, I'm sure, and
some day this would be a good upgrade for me.
Q. Did you have any problems with the getting the taper ground on the az
bearing drive rod??? How did you calculate the taper?
As described above, the Dob Driver hardware uses a 2" diameter
knurled metal wheel to drive the AZ axis. The knurled wheel
rolls along the formica.
Q. Did you make your altitude bearings from wood, surfaced with aluminum???
If not, what?
Yes, but I didn't make them. They are original Coulter ALT bearings.
Q. What do you use for the motors/gearboxes (this is probably our toughest
problem). Where did you get them? Gear ratios, etc? Costs? Are you happy
with the periodic errors and backlash they have??? Details on the motors,
etc..... (steps, voltages, etc). How about the drive rods themselves???
Bearings for them, etc......
As described above, most drive hardware is from the Dob Driver.
I replaced the Hurst motors with Jameco steppers (200 steps/rev)
for $10 each. I run them at 24V in order to get the slew speed
up to about 1.3 degrees/second. These motors fit onto the Hurst
gearboxes fairly easily. The gearboxes have backlash, but this
is compensated by the software. I am now purposely keeping my
ALT axis slightly off-balance which removes all the ALT backlash.
The AZ backlash is about 7 arcminutes and is removed by the software.
Q. Clutches....what do you do for clutches for the drive rods?????
I use the Dob Driver clutches which are easily engaged and released
by hand.
Q. I see a ccd camera on your picture of your scope.....How long of
exposures can you make??? (unguided) without image smearing?
All the pictures on my Web page are single 30-sec exposures. I
have taken some nice 60-sec shots, but I don't get consistently
good results like I do for 30-sec shots. This is consistent
with what Richard Berry reports, as he recommends "track and
stack" multiple short exposures.
Q. What construction "Lessons Learned" do you have?????
By far the best advice is to make sure all your bearing
surfaces and axis are flat, perpendicular, concentric, or
whatever they are supposed to be! Any errors will result
in slewing and tracking problems. After spending a lot of
time tweaking my mount to remove sloppiness during original
construction (I never dreamed I would be using a Dob to
do astroimaging!), I still have to work within the limits
of the scope's capability. I can't do long across-the-sky
slews and expect to land on a faint object in the middle of
my CCD chip. Instead, I choose an area of the sky and
pick my initialization stars maybe 30 or 40 degrees apart max.
Then when I want to slew to a faint object, I'll first slew
to the nearest "alignment star" (I have over 250 bright stars
in the database, so they cover the whole sky). After centering
the alignment star, I "reset" and then slew to the object. This
keeps the slew less than 15 degrees or so, and I will nail the
object every time on my CCD chip (which has about a 15 arcmin
field).
Q. Do you get much backlash with a belt drive??? (Yes, I know Mel's
software addresses backlash)
Yes, I'm sure there is some small amount of belt stretching, as the
belt has different tension on one side of the pulley from the other
(easily demonstrated by plucking the belt like a guitar string).
The belt backlash is probably very small, but even if it were large
it would hopefully be consistent and easily removed by the compensation
software. A bigger concern with belts is probably non-uniform
stretching during tracking which would introduce image smearing.
With my scope, construction errors in the alt axis (not being perfectly
concentric, etc) are the primary source of error and probably not the
belt.
Q. The third support is the 2" knurled wheel of the
azimuthal stepper motor gearbox. A clutch can quickly disengage
this wheel and let the rockerbox rotate on a teflon pad for manual
movement.
How wide is the knurled wheel? Any appreciable width would introduce a
tracking error due to the differences in the ID and OD described by the
width of the wheel, I think thats why Mel suggests using a tapered drive
roller?????
The knurled wheel is about 0.25" wide. Ideally, I should have
tilted the AZ motor/gearbox/wheel housing a bit so that the edge of
wheel rolls on the formica. Although I must admit that I didn't
do this purposely, the wheel height is a bit higher than the
two ball bearing heights, so the wheel does tend to ride only on the
inside diameter (I can see this by looking at the worn track in
the formica, it's a narrow line and not 0.25" wide).
Q. The scope is controlled by PC using Mel Bartels' software as
explained on the Internet at http://www.efn.org/~mbartels. The
software uses the popular 2-star alignment method which initializes
the software to the sky without the need for polar alignment or
leveling of the base.
Don't you have to initially tell the scope where the ax axis is aimed (I
think Mel's software actually looks for an initialization point with the
tube at 90 deg to "vertical" (vertical being aimed parallel with the azimuth
bearing bolt, not necessarilly actual vertical)
Yes, the software needs to know the initial ALT angle w.r.t. the base.
I have a wooden dowel rod cut to the exact length where the scope
tube is parallel to the base, so I always start in this position
(ALT=0). However, the software does not need to know what the
AZ starting position is, this is determined by the software after
initializing on 2 stars. The software calculates the AZ angle
along with the observer's effective latitude and longitude.
Q. The CCD camera is the Cookbook 245 with the Low Dark Current
modifications in place. I have replaced the water-cooling system
with a forced-air convection cooler built by my friend Kurt Mihalco
of Coherent Systems. His cooler is nice and compact, and replaces
the bulky plastic tubes, copper coil, bucket, and noisy pump.
Do you get enough of a temperature drop just using air cooling?? Is it
stable??? The water really helps on both of these areas.... I live in
Houston, and 80 deg nights are not uncommon....Air cooling will not drop the
chip temp a whole lot I would not think.
I spent most of my time working on my scope, and not optimizing
my camera. I've been quite amused reading all the CCD postings
about people arguing over things like which metal alloy should be
used for the wires connected to the chip to produce the best tradeoff
in heat capacity vs. joule heating! I can only WISH (I think) that
my set-up was optimized to the point where I need to rethink the
alloy of wire I use. But your concern about air cooling is valid.
For my 30-second exposures with the LDC mod, dark current is not
a problem (I think the LDC mod reduces dark current by about
10X). But yes, the temperature does vary with the ambient,
so I may have to take a couple dark frames thru the night. I try
to keep the ref and reset levels the same for each image, which
I believe helps the images and dark frames stay calibrated,
but I may be wrong here (I need to brush up on CCD operation). Kurt
was going to add a closed-loop system to the air cooler which would
modify the Peltier current to keep the chip temp steady, but this
feature is not yet available. For now I'm satisfied with the camera
for my short exposures and enjoy the convenience of not dealing with
the noisy pump, hoses, and bucket.
I also use Kurt's custom power supply for the camera, it runs off
of a single 12V battery. I use this single 12V battery to
power both laptops, the scope, and the camera (about 6 amps total).
I use a 95 amp-hour marine "deep cycle" battery that provides
power all night while I'm out in the desert or on Palomar Mountain
where I usually observe.
As described above, most drive hardware is from the Dob Driver.
I replaced the Hurst motors with Jameco steppers (200 steps/rev)
for $10 each. I run them at 24V in order to get the slew speed
up to about 1.3 degrees/second. These motors fit onto the Hurst
gearboxes fairly easily. The gearboxes have backlash, but this
is compensated by the software. I am now purposely keeping my
ALT axis slightly off-balance which removes all the ALT backlash.
The AZ backlash is about 7 arcminutes and is removed by the software.
Q. Excellent idea on the off-balance config to eliminate backlash in Alt! What
voltage were the motors intended to be run at? When you slew, what rate do
the motors step at??? Why the 24 volts?? Why not just up the stepping rate
at the nominal voltage (12v??)
They're spec'd at 24V. If I did my math right, slewing at 1.3 deg/sec
corresponds to about 1200 steps/sec. Steppers run faster at higher
voltages, so running at 12V will cause the steppers to stall out
at less than 1 deg/sec, too pokey for me. I could probably increase
the voltage to 36V or more and get even faster slew speed, for
example Mel talks about 2K or 4K steps/sec at overvoltages.
All the pictures on my Web page are single 30-sec exposures. I
have taken some nice 60-sec shots, but I don't get consistently
good results like I do for 30-sec shots. This is consistent
with what Richard Berry reports, as he recommends "track and
stack" multiple short exposures.
Q. hmmmmmm most of my shots are 30-60 sec images too usig my platform. I was
hoping for a minimum of 60 sec, and perhaps 120 sec images reliabily.... so
I could take 20 or so and add them together (I currently add my 30 sec shots
as you mention). What is the predominant problem with not being able to go
longer than 30 sec???
Good question! Imaging in the field with a closed-tube (non truss)
dob requires almost windless conditions. Any gusts of wind will
smear my images. Sometimes I hold up a sleeping bag during the
30 second exposure! If there's no wind, I get my best images.
However, I rarely get drift-free images at 60 secs. I suppose its
a combination of mount errors (my alt axis is not perfectly concentric
so the radius changes slightly, and my AZ bearing surface is not
perfectly flat so the scope will have up-and-down AZ movement instead
of purely AZ). Also, the ALT belt may stretch as discussed above,
and the gearboxes are inexpensive, not "precision" as Mel recommends.
Taking all of that into account, I guess I should be pretty happy
with my images, and I am!
Q. Thank you ever so much for taking the time to respond!!!!
You're very welcome. I'm glad to offer any information that
will help you, and hope you have success with the project.
It was challenging, fun, exciting, and extremely rewarding
to convert my Dob into a scope that can perform alongside
commercial scopes that cost 10X as much. Go for it!
Hope all of this helps, and have fun!
Jerry Pinter
jerryp@amcc.com