Numbers never lie, do they?

[wimvb]

Near perfect polar alignment with the synscan? I doubt it, but the proof will be in the proverbial pudding. There's too much of a moon for serious imaging, but I need more training in guiding. Hope tonight there will be opportunity for that.

Btw, 2-star alignment with a 2 x barlowed 17 (?) mm eyepiece with homemade reticule resulted in these numbers. The second alignment star was pretty much spot on without adjustment, so there's hope.

[wimvb]

And here's the "pudding" from that same session. The guiding graph from Lin_guider data (RA and Dec deviation over time). Over nearly a 12 minute period, there is a declination drift of 15 arcseconds (which for my rig translates to roughly 11 pixels on the imaging camera), and a similar drift in RA.

After about 700 seconds I turned on Dec guiding (one direction only) and about a minute later I turned on RA guiding as well.

Apparently the polar alignment wasn't "near perfect" after all. Surprise, surprise

Guiding with Lin_guider works ok, but I think I'm going to experiment with OpenPHD on my Raspberry Pi. The problem with Lin_guider is that there is hardly any documenation, and it's difficult to tune in appropriate gain parameters for guide control. PHD has more tools to analyse the guiding graph and correct guiding (mis-)behaviour.

[wimvb]

The topping of the pudding: M38

Skywatcher 150PDS on AZ EQ6 GT, Pentax K20D 9 x 4 minutes at ISO 400 under the light of an 80 % full moon

guiding: ST80 with ASI 120MM and Lin_guider on Raspberry Pi (1 second exposures, RA averaged 3 frames per guiding pulse, DEC+ averaged 6 frames per guiding pulse)

9 out of 30 images stacked, cropped and stretched in PI. bias and dark frames, but no flat frames (didn't bother for this test). Even dark frames were from a previous imaging session.

The other images were good enough as far as tracking was concerned, but I think there was some high cloud wich spoiled the seeing. I decided to only stack the frames with visually the best FWHM.

The light from the moon gave a horrible amount of light pollution with associated noise, but fortunately the cold kept the camera noise and hot pixels under control.

[wimvb]

Some more data from the same session (images shown here are resampled crops from the original frames)

The left panel shows the first (red) and last (blue) image from this session. Approximately 2 hours between subs, the offset is 23.81 pixels in x, and 9.55 pixels in y, or 33 arcseconds in RA and 13 arcseconds in Dec. Apparently the guiding can't completely eliminate the drift due to PA error. Per sub (4 minutes) the drift is 1.1 arcseconds in RA and 0.4 arcseconds in Dec.

The right panel shows the guiding calibration. As I align my camera sensor with the RA axis, the horizontal lign is RA, and the vertical line is Dec. The increase of intensity at the highest Dec position indicates dec backlash.

 

[MattJenko]

Pardon my ignorance, but what has a 2- star alignment got to do with polar alignment?

[wimvb]

2 hours ago, MattJenko said:

Pardon my ignorance, but what has a 2- star alignment got to do with polar alignment?

By itself, nothing. But after a 2-star alignment, the synscan will report the polar misalignment numbers, which can then be used for the (synscan) polar alignment. These numbers and the polar alignment can never be more exact than the mount's pointing accuracy, which is 5 arc minutes.

The subsequent drift data showed a larger polar misalignment. Hence the topic: numbers don't lie, but you have to be aware of how you get and use them. It got me thinking about how accurate alignment is, and how accurate it needs to be.