Hi folks,
Last night I got to test the two nights of calibration work I spent on the mounting earlier in the week. Put simply, the mounting knows what angle it is pointing at with incredible precision due to some very clever encoders embedded in the works but that doesn't mean on its own that the mounting can point at and track any star you care to name because of mechanical misalignments, most commonly in the way the telescope is attached, mechanical flexure of the telescope and atmospheric refraction which bends the position of a star the closer it gets to the horizon. To calibrate the ASA mounting I told it to point where it thought a star should be, I then fine tuned that pointing using my camera and, with the star centered exactly on the electronic crosshairs, I then told the mounting's software that that was where the star actually was. Repeat many times (I went over the top and repeated 82 times) with stars scattered all over the sky and then tell the software to build a pointing file so that in future it can compensate automatically for all those repeatable errors.
All fine and dandy and it's very nice to have a star appear where you expect when you tell the mounting to point the telescope at it but, of course, the real purpose is to allow the mounting to track the stars with high precision. How high? Well this is a 400% crop of a few stars with an
unguided exposure time of 1,000 seconds (16 minutes 40 seconds). The scale is 0.475 seconds of arc per screen pixel.
Photoshop smoothed out the pixels during the 4x enlargement but it didn't make the shape any rounder. This was cropped from pretty near the corner of a single 4096 x 4096 pixel Hα sub (the ML16803 sensor has a diagonal of 52.1 mm) in the Gulf of Mexico region of the North America nebula so it will show up tracking errors, image rotation caused by polar misalignment and optical flaws in the TEC 140 'scope and field flattener. The only processing done was image calibration (bias, dark and flat frames applied) and a histogram stretch to show the stars at reasonable brightness while de-emphasising the nebulosity as all I wanted to test was star roundness. For context here is a 100% crop of the region taken from a stack of eight 1,000 second Hα subs with a more normal histogram stretch and a little contrast enhancement and noise reduction applied.

For this image the scale is 1.9 seconds of arc per pixel. Quick and dirty processing done just for the purpose of this post. The data will eventually be combined into a four panel mosaic of the area which I'll be working on this summer with, I hope, the addition of SII and OIII narrowband data as well as Hα and RGB. Based on what I'm seeing I think the mounting will comfortably be able to run 1,500 second exposures for the SII and OIII subs so I'd better order those filters!

Bob.
P.S. Please note that this is not some April fool's joke - no deconvolution was applied to either image!