OK, Enough about the idea of a flat helping anything.
As far as the laser, I don't think aiming at a centered dot on a paper mask is going to get me very accurately centered but I suppose it is better than nothing. The Glatter laser I have projects concentric circles but these circles do not reach the edge of the objective over the path length. Now, if they could somehow be spread out more they might prove useful in centering.
1. I will take the rotator to my local machine shop. I'm hopeful the tech there will help me measure the runout as Josh suggests. If not I'll need to wait until the tool I've ordered arrives.
2. I will post a photo of the setup.
Repeated and got same values pretty much going the opposite direction (at 360 and 315 measured 9), but ended up at 180 = 1.5, so the measurement is probably only accurate to ca 10 microns or so. But, I think I can safely say the thing is tilting by 80-90 microns.
I put the dial indicator on the side to measure concentricity:
Those angles, are they the rotator PA while being moved by the motor, and the dial gauge is fixed? And what is being used as the reference surface, the rotator body I presume?
So I see two possibilities:
-the moving tube of the rotator is rotating true (no eccentricity) and the end, or mounting surface, is cut crooked
-the moving tube is cut square but rotating with eccentricity
In either case the other end of a 'perfect' adapter attached to the moving tube will draw a circle around the optical axis.
I assume you measured the surface at a diameter of 100mm
The angle is 80 um / 100 mm = 0.8 mrad
and the orbit radius will be 0.8 mrad x 190 mm = 152 um
The side to side motion will add/cancel to the above with an amplitude of 80ish um.
Not too bad really. A 0.5 mrad tilt of the image plane gives 0.5 mrad x 35 mm = 17 um of defocus in each corner of your KAF16200 chip (plus in one corner and minus in the other). No big problem at f/7.
I appreciate your calculations! This is reassuring and confirmation of your assessment of the video of the laser. This, at least, represents the best I can hope for as long as I can get the rotator mounted properly.
Yes, the dial indicator was sitting on a table with the probe on the moving part. The angles are just arbitrary degree readings from the software during rotation. I did take care to keep the angles the same for both tests.
My Gemini now fails to rotate at all. I was doing a simple experiment with essentially no weight on the rotator. It rotated about 1.5 times without any issue but suddenly stopped and indicated "not homed." I tried to home it. Motors spin up but rotator doesn't rotate. After awhile it announced "homing failed." Hoping it was software I disconnected power and USB. Reconnected power and motors start spinning but no movement. After a while they stopped so I connected the software. It connected and indicated homed at 0.00 degrees. Tried to rotate and nothing. Strangely it counts up or down as it tries to rotate and the numbers stop on the assigned degree of rotation even though there is no physical rotation. I tried the hand controller too. My guess is that something is rotating internally to activate sensors of position and that something has broken that links that mechanism to the actual rotation.
Anyone here have any experience opening or servicing a Gemini? I fear I will be packing it up for a trip to the USA. At this rate I will never get to image.