4/29/2010. Second night with the A-P Mach1GTO. Ah, much better! This morning I remounted the telescope and started the mount. I just selected "resume from park." I love the auto-park feature; turn off the mount and it immediately saves its data and then shuts down, so unless it matters for other reasons (to clear an obsevatory roof, allow easy dismount of the OTA, let a cover fit properly...) there is no need to remember to park the mount at the end of a run. Then I slewed to the Sun. I missed by several degrees. I tweaked the aim using the shadows of the mounting rings and pressed "recalibrate" (not sync). Next, I slewed to Venus. The telescope pointed into the trees. An hour or so later, when the field rose above the pines, I still did not see Venus. I used the buttons to begin a search for the planet... and immediately discovered that Venus was hiding behind the crosshairs. After admiring that for a while, I parked the mount in position 3, covered the telescope and waited for nightfall.

In the interim, I elected not to relocate the ST80 further up toward the A-P's objective. Rotating the imaging camera 180 degrees moves its viewfinder eyepiece out to an easily accessible location, so why bother?

In evening twilight, I slewed to Capella on the opposite side of the sky (aim was off by about 1.2 degrees owing to either poor polar alignment or, more likely, to the fact that I had not taken care to align the guide telescope with the main instrument or to insure that either is orthogonal with the mount). I centered Capella and recalibrated (just for drill). Then I polar aligned with the newly focussed boresight, returned to Capella and recalibrated (this time for real). Twilight was deep enough by then to take test frames to confirm focus and aim. Best focus is four short tweaks of the FeatherTouch fine-focus knob from full in -- more than a quarter turn, less than half. I'll soon use the Netbook and Nebulosity to nail this down, but that's good enough for tonight's practice session.

I slewed to M81 and M82, then took time out to watch an unexpected pass of the International Space Station. I could see M81 in the guide scope. A 100-second test frame confirmed aim in the main instrument. I used the parfocal eyepiece LVI supplied with the Smartguider to improve its focus. The crosshair eyepiece focuses near 5.19 on the GSO focusser, but a better focus point for the LVI is 4.9 or thereabout. This can still be improved. The guider immediately locked onto 4.5 magnitude 24 UMa a couple of degrees west of M82. I got the point spread down from 8 to 4 pixels but then lost the star to clouds and mechanical motion when I tightened the lock screw. I got the star back with the parfocal eyepiece and calibrated the LVI guider wtih the A-P's autoguider speed set to 0.5x sidereal. This time I left the focus alone at a relatively messy 7 pixels.

Autoguiding corrections in declination were constant and not small. I watched the graph through the first 10 minute exposure and most of the second and noted that the pattern was a persistent sawtooth. A correction N was immediately followed by another to the S of equal size, then another to the N, and so on forever. I could tell by listening to the dec motor that it was in constant motion. I thought the guider was overcorrecting, trying to correct 100% of every excursion by the guide star and apparently overshooting. I puzzled about how to access the LVI's aggressiveness settings without stopping guiding (or without stopping it for long) and decided instead to cut the correction rate on the A–P drive in half, to 0.25x sidereal. That worked superbly. Both axes immediately settled down. The graphs of their corrections were nearly flat.

M81 and M82, 12nm Astronomik H-a filter, Canon 20D-HG @ ISO 1600, 5-inch F6 A-P on a Mach1GTO, autoguided by LVI Smartguider on an Orion ST80. 5 subs, 10m54s each. Through clouds and full moonlight. At a guess, approximately equivalent to a single 10 minute exposure on a clear night.

The guider held onto the guide star for well over an hour through surprisingly thick clouds lit by the rising Moon. It was a very impressive performance. I'm thinking the Smartguider will indeed guide down to 6th or 7th magnitude with the ST80 under more reasonable conditions for astrophotography. With the focus further refined, it will likely reach even deeper.

I used an Astronomik 12nm H-a clip-in filter to suppress both light pollution and moonlight; also, I wanted to make long exposures to get some experience with the guider. You really can't believe how cloudy and bright the sky was. I could usually see the entire Big Dipper. At best, for only a few minutes at a time, I could see stars as faint as 3.4 near my targets. Generally, the telescope was aimed into a featureless gray void.

I shot a dark frame while shutting down the mount and putting the electronics away. I had no particular use for it as tonight I just wanted to confirm tracking; anything that comes out of this for show and tell would be a bonus.

Note: Just before packing up, I reset the autoguide speed in the A-P keypad to 0.5x and reduced the aggressiveness of the LVI on both axes to 2 (from 5). Following many large corrections to return the guide star to its assigned point, the graphs looked good. Try it that way next time without the kludge of changing the correction speed between calibration and guiding. (Note from tomorrow night: 2 appeared to not be aggressive enough; 4 worked better.)

I brought the flash card in and took a look. The first two frames which were made with over-aggressive autoguiding are trailed in dec. No surprise there. At a glance, it was clear that subsequent frames were tracked very well. They suffer from the noise inherent in 10 min (actually 10m 54s) iso 1600 frames shot through glowing clouds, but everything worked.

I wanted something to show for the night's efforts, so I spent some time cleaning up the image data. Using Maxim DL5, I extracted the red Bayer plane from each image and from the dark frame. I saved the extracted data in FITS files. I loaded the results into PS and applied the dark to each as an exclusion layer. I took a one pixel median to suppress remaining noise. I saved the images as TIFF files which I aligned and stacked in Deep Sky Stacker, averaging them to further suppress noise and to help capture faint details. I used PS to crop and clean the result and managed to make it recognizable if not quite presentable. There were simply too many clouds, too much scattered light, too much attenuation to get a good signal. Over-stretching very thin data emphasized banding in the 20D.

At full scale, the H-II regions in M81 are quite sharp and trace the spiral arms delicately. There is the slightest hint of the elaborate H-a clouds apparently being expelled from M82. All the pieces play well together. Stay tuned.

Next time: add a 30s delay between exposures to try to reduce glow from the amplifier in the camera.