I used the Polar Alignment module to refine the alignment given the change in temperature up here in Canada. I've been guiding for about 20 minutes and phdlogview reports a polar alignment error of 0.1 minutes. This took less than 10 minutes to get. The Alignment modules is one of the gems in EKOS!
I have an unobstructed view of Polaris and NCP, and a clear view to the east, and the meridian to about -5 degrees south. (Presumably you're asking about drift alignment if you are asking about the east/west star? I was using the EKOS Polar alignment bits, not drift alignment).
I changed my imaging telescope, so I re-did the Polar alignment as the imaging axis would likely be different between the two telescopes. The value I reported above is an estimate obtained from phdlogview after about an hour or so of guiding.
If I recall, EKOS polar alignment reported that the alignment of off by about 5 arcmin or so, so not much and the adjustments were minor. I was happy I could get it that close with little effort, compared to drift alignment.
I used the most recent stable KStars, not a nightly KStars build. If you recall my original post on this module back in July where we discovered the correction vector direction issues, I've encountered the problem perhaps about 25% of the time and simpy redo the alignment if I pick the wrong direction.
I saw your request for feedback. I've had lots of cloudy nights lately, but it is supposed to be clear again tomorrow night , so I'll have another crack at it if I have time. Did you want me to test using the latest nightly?
I asked about the east/west views not about drift alignment, but because Step 8 in the Polar alignment workflow (www.indilib.org/about/ekos/alignment-module.html) needs a star "as close as possible" to the east or west cardinal points. I have a restricted view and have tried the polar alignment in the past, but have not found it to do better than using my polar scope. However, this *is* something I intend to try again, because I need to be able to make longer exposures.
Also, knowing what focal length you are imaging at and your exposure times would be helpful to get a sense of how the alignment quality helps you.
Ubuntu 18.04 and Raspbian Jessie; INDI 1.7.4
Mounts: CEM-60 chiefly; iEQ45
Cameras: Atik 383L+, QHY5-II-M
If you have a clear view at Polaris (or the southern celestial pole when in the southern hemisphere) then there is a much simpler polar alignment procedure nowadays. Simply point your telescope with camera at Polaris, go to the Alignment tab, select Polar Alignment and click start. This will take an image and plate solve it, rotate the mount 30 degrees westward (or east if you selected that), take another image, rotate 30 degrees more and take a third image. Then the three images are used to compute the polar alignment error, which will show up in the last image as a pink arrow. Click any star and the pink arrow will start there. Then use the alignment knobs of your mount to move the star to the other end of the arrow and you're done.
Wouter van Reeven
ASI6200MM and 7 slot 2" filter wheel with a SkyWatcher Esprit 80 ED on a SkyWatcher HEQ5-Pro
ASI1600MM-Pro Cooled and 5 slot 1.25" filter wheel with an 8" TS Ritchey-Chrétien on a SkyWatcher EQ6-R
Do you have to eliminate any cone error before using the alignment module tool? I'm just very weary about using this alignment module without shimming out the cone error of the OTA first. Any thoughts?
I wasn't following the process you describe - I think the documentation is out of date. My understanding is the new process uses plate solving to get an accurate measure of the centre of rotation by calculating 3 points on a circle, and then a vector is computed to move the centre-of-rotation to NCP. The direction of the vector was the only point in question when I ran into problems last July - not sure this has been resolved yet.
For this alignment, I was imaging at a focal length of 900 (f7.5). Take a look at
for more details. When it works, it works well. The work-around for the vector direction issue is to redo the alignment and move it in the opposite direction.
The problem with reversed direction for the correction vector could still be present. But I wanted to solve a more pressing matter, and change the behavior to use W/E motion commands exclusively without using "Goto" commands.
I've never done that and the few times I've checked my cone error, it's been pretty small. My maximum exposure is typically 10 minutes, and I work backwards, using my polar alignment to compute the amount of rotation I would see for a given exposure.
My biggest issue tends to be related to cabling, especially when it gets cold and the cables become rigid - my guiding RMS will increase by .1 to .2 arcseconds.
If you are using PHD2 for guiding, they have a paper that recommends a polar alignment that is slightly off to avoid DEC oscillating back and forth. I have been successful setting a larger DEC min move and using the Z-filter option to guide DEC. I do pay attention to the difference between DEC and RA guiding as it can lead to oval stars - they should be close.
If you have a long focal length and intend to use long exposures, then you will at least need to use some form of off-axis guiding. The EKOS alignment module works very well to get very good alignment and typically won't be the limiting factor for your imaging requirements.