22-inch Autoscope, as seen in the BYU machine shop. OTA is not the original version. These first few photos give an idea of overall construction and layout.
Individual components, from the ground up. The base is two rings of steel. Bottom ring is 1-inch, top ring is 3/4 inch. Heavy! The tape measure has six inches of tape exposed for size reference. The disk standing on edge is the bottom bearing race of the thrust bearing for the azimuth cone. The race appears to be a spherical, and the roller bearings are not cylinders...they are 'barrel' shaped..fatter in the middle. This bearing disk has oversize holes to allow lateral adjustment, and the mounting bolts allow vertical adjustment by rotating the support nuts.
This platform sits on top of the base, and supports....
...the azimuth cone's upper ring. The ring is machined, and about 7/8 inch wide. Also shown is the other half of the spherical race thrust bearing. The thrust bearing mounting holes in the azimuth cone are not oversized...which means you can't adjust the bearing laterally.
Close up of spherical race thrust bearing.
This bracket sits directly on top of the pedastal, and holds four roller assemblies that constrain the upper ring of the azimuth cone.
Each corner of this bracket has a hole that carries a spring loaded, adjustable plunger. The plunger has a pivot at the attachment point that carries....
...this roller assembly, which consists of two vertical rollers to constrain the azimuth cone's upper ring, and two pairs of horizontal rollers that keep this roller assembly properly oriented against the azimuth cone upper ring. In this photo the roller assembly is engaged wtih the azimuth cone's upper ring.
Other side of roller assembly. Now we can see the two pair of horizontal rollers mentioned in the previous photo. These two pair do not tightly clamp the azimuth ring...the gap is about 1/32 inch, perhaps less. This is a friction drive, and the original plans called for stepper motors to drive two of the eight total vertical rollers to provide azimuth motion.
The fork is a box frame that bolts onto the top of the azimuth cone.
Top of fork arm, looking down.
Fork is made of 1/8 inch steel.
Mounting plate on top of fork appears to be 1/2 inch thick.
Mirror box, mirror cell (for axial loads), baffle tube (also mirror's lateral support), elevation shafts and pillow block bearings.
One side of the mirror box has machined mounting faces for the elevation disk/friction drive.
Elevation shafts are slightly less than 1 1/2 inches.
Baffle tube is bolted in place...here removed to see it's overall length.
Friction drive disk for elevation.
OTA, minus the secondary mirror holder.
Main mirror. 22 1/4 inches diameter (approx 22 inches effective aperture), 3 7/8 inch thick. Central hole diameter is about 6 7/8 inches. Central hole is lined with a sleeve that feels like Teflon, 1/4 inch thick. Optics fabricated by Paul Jones/Star Instruments. Primary f/ratio 3, overall system f/ratio 8. Optical design: RC. Substrate: zerodur.
Secondary mirror in cell. Approx 8 1/2 inches diameter. (Thickness unknown...I have not taken cell apart yet.) Secondary has small center/collimation circle (etched?).
Other views of secondary holder.
Stepper motors, controller cards, computers, and a stack of documentation.
Fast forward 3 1/2 years! I've gained experience and confidence by refurbishing many smaller scopes. Now it's time to return to this project...and it got a kickstart thanks to John Manford's donation of a hefty engine hoist!
This was an initial fit test, and I needed to make a short extension for the hoist arm....I can't lower the azimuth cone into the frame in the current configuration.
The extension arm was the next day's project.
Here's one of the mods to the scope frame - a simpler roller bracket. There are plenty of degrees of freedom to allow adjustment so that the roller can seat properly against the azimuth disk.
The bottom of the azimuth cone is bolted to a spherical thrust bearing. Today I added some X-Y push-push brackets so that I can properly align the thrust bearing with respect to the azimuth disk's axis.
With the engine hoist's new extension arm I could easily lower the azimuth cone into the frame and start adjusting all the bearing geometries. It's evening, and I'm getting tired. I'll finish the job tomorrow...but one close glance shows me one adjustment that needs to be done....
...this roller bearing bracket needs to be tilted so that the entire roller face contacts the azimuth roller. Some trial and error with shims at the bracket base will fix this....tomorrow.
Time to measure wobble/runout of the az. disk as it spins. Initially it was peak-peak 0.035inch! That's a bad tolerance in wood shop. Adjusting the push-push bolts to move the spherical thrust bearing reduced wobble. In about four iterations peak-peak is only 0.001 inch. Success! After that I adjusted rollers for tilt in two axes. Now azimuth motion is much smoother and only needs a few ounces to overcome friction. Next step - evaluate the friction roller drives...maybe they can be reused?
Gear reducer is a harmonic drive, 100:1 ratio. Note the toothed periphery of the ball bearing assembly.
Here's a closer view.
With all the metal work, new adapters, modifications, etc....I'm doing lots of drilling. I've learned to sharpen bits because I'm not close to a hardware store when they get dull or break.
Servo motors and encoders to replace the original stepper motors.
Incremental encoder with a ribbon cable. I'm used to four wires from encoders, not ten. Gotta find a data sheet for these encoders.
I used spare parts to fabricate a spindle that will allow me to grind large disks into precise circles. This was an initial test with Plexiglas...a horrible material, but the results were promising. (I don't need an exact outer diameter. I just need a round circle of an approximate diameter. I can make disks 33 inches in diameter with this rig.)
I need large disks on which to mount encoder tapes for high-resolution measurement of telescope shaft position. This will improve pointing and tracking performance.
This precision ground steel disk is 1 inch thick, 20 inches diameter - heavy! I am cutting away the outer 1.5 inches for an encoder ring. The resulting ring will be much lighter than the entire disk.
First pass with thin abrasive cutting wheel.
A couple minutes later (ha ha)...half way through...must flip disk and continue on other side.
Noisy and dirty work.
First daylight coming through from other side! Almost done.
It took two grinding wheels to finish the job.
But is the outer diameter still a precise circle? I had my doubts, based on how the disk behaved when the last cuts were completed and the ring separated from the rest of the inner disk.
Burrs removed - ready for measurement of outer diameter.
Simple rig on heavy plywood...two blocks to set position of outer diameter/surface...dial indicator to measure deviation from a circle as disk is rotated 360 degrees...and the results were not good.
Note the little dot on the wood...then I rotate the disk 90 degrees....
And the diameter is about 1/32 inch different?
Not only that - the outer diameter is not an oval, but an irregular shape. The amount of internal stress in the steel is amazing...it warps this 1-inch cross section by so much!
Lesson learned - cut the ring from the parent disk first, and then do the final machinging/grinding.