The Ohio State University College of Mathematical & Physical Sciences Department of Astronomy

Attendees: Pat Osmer, Paul Byard, Tom O'Brien, Mark Derwent, Darren DePoy, Bruce Atwood, Jen Marshall, & Rick Pogge,

Grating Tilt Prototype Tests

The main topic of this meeting was to hear a report by Mark Derwent on tests of the grating tilt mechanism in the mechanics lab. Clicking on the images below will load a full-resolution version.

Grating drive test rig in the lab. The actual grating drives will be made by cutting the large aluminum gear in half to make drives gears for two gratings. The drive is provided by a stainless steel worm gear connected to a standard stepper motor.

For the tests, a benchtop Compumotor controller and laptop were used. An absolute encoder strip (which will not be needed for the actual drives) was used for test metrology.

The tests sought to evaluate the following properties of the grating drive system:

1. Home position repeatability
2. Target Position accuracy and repeatability
3. Backlash measurement and recovery
4. Limit switch activation and recovery
5. Drive system life cycle

Grating Tilt Motion Tests

The grating tilt motion tests were setup with the following:

• The drive resolution was setup as 200 steps/revolution of the worm gear.

• An absolute encoder strip was used to measure the angle of the main gear. The encoder resolution was 1 count = 0.815 motor steps.

Home-Target Cycling

• The first test evaluates how close the grating can be driven to the home (zero-reference) position. The "in home" position is noted by the assertion of the "home" proximity switch. Multiple cycles were performed, using the absolute encoder to evaluate the true position relative to the proximity switch assert that stopped the motion.

  
  

  The grating home position is highly repeatable, with the home position being recovered to +/- 1 motor step. An oddity, however, is that the home position drifts after the first couple of cycles then stabilizes for all subsequent tests (up to 100 cycles). This is not understood yet.

• The second test was to send the grating home, then forward to a target position (here ~49100 encoder counts from home) and test the repeatability. The drive motor was advanced a fixed number of steps from the home position, and the encoders read to see how well the target was recovered.

  
  

  Again, the target position is highly repeatabl, her to +/- 5 encoder counts, corresponding to +/-3 arcseconds of rotation of the main gear. There are two remaining effects not understood, the slow ramp-up at the start of the cycle of tests (reflecting the home position ramp-up above), and an oscillation with a period of ~16 cycles.

The conclusions are that despite some oddities we need to explore, we are well within the specifications of grating tilt repeatability.

Backlash

• 301 motors steps are needed to overcome backlash.

  
  Backlash hystersis test results.

• Backlash proved to be highly repeatable and stable. Typically the difference between forward and backward going positions was under 2 motors steps (rms).

  
  Backlash recovery repeatability results.

Limit Switch Tests

Limit switches are non-contact inductive proximity sensors.

When activated, the limit condition is recovered by reversing the drive direction.

If a limit is encountered during a "grating home" routine, the motor will revers direction and continue the search for the home switch.

Conclusions:

• The inductive proximity sensors will work as limits switches, and did not fail to activate when the system was sent to the limit positions.
• The limit recover procedures are straightforward and will work in practice.

Grating Life Cycle Test

Test setup was for 9000 cycles of the grating drive, derived from 150 nights/year with 3 cycles per night, over 20 years, with relubrication of the system every 2 years (using MoS₂ grease as the lubricant).

1 "cycle" is "go to home" followed by "go to target" commands.

Results:

• Gear wear on the main aluminum gear was 0.002-0.003 inches after 9000 cycles.

• Gear wear was smooth and evenly distributed.

• There was no visible wear on the stainless stell worm gear.

Conclusions:

• With proper alignment, lubrication, and a 2-year maintenance cycle, the gear can be relied upon for 9000+ cycles.

• An aluminum gear with a stainless steel worm gear is the correct combination.

• MoS₂ filled lubrication is better than E.L.T. lubrication.

There was some discussion about the basis of the maintenance cycle used for the tests. If the numbers are right, it looks as if it may not be necessary over the life of the instrument, even to changing the lubricant.

Overall, the grating drive concept has passed our tests with flying colors. We will proceed to the next steps of developing the full grating drive system, and the grating mounting system.

Optics

We discussed quotes for the imaging flats, including both polishing and coating. The red channel flats will be coated with the overcoated silver, and the blue flats coated with aluminum. Prices look competive (we can't quote online...). Delivery time is 14 weeks minimum. We will also ask for witness samples of the coatings for life testing.

We received quotes for the imaging filters. After much discussion, we decided to order an SDSS filter set (5mm thickness, 86x86mm square) for the imaging filters. A complete set means the filters within that set should be parfocal.

Spectroscopic filters for order separation will need to be 182x86mm in size. We still need to determine which ones we will need, but there will be no problem obtaining standard Schott filters (e.g., GG385 and such like) in this size. This decision is coupled with the parameters of the gratings, and comes later in the project.

For the "open" position in the blue camera, a 2mm piece of polished fused Silica will permit use of MODS unfiltered (imaging or spectroscopy) without the need to refocus for the filters. No similar analysis has been done for the red camera at this time.

The next MODS meeting will be in September before the start of the Autumn quarter, but after the Labor Day holiday.

R. Pogge, 2001 August 30