Posted 13 July 2012 - 11:12 AM
"After you do your first 2+4, you can read the polar misalignment by ALIGN/Polar Align/Display Align. Write that down, I will come back to it. If you display align again after ASPA it will show your polar misalignment to be exactly zero (or sometimes just a few arc seconds off). How can it be EXACTLY zero? Surely there must be some error. The software assumes that all parts of the ASPA went perfectly. That you synced perfectly. That the pointing was perfect while adjusting the ALT and AZM axes. That your adjustments in ALT didn't tweak your RA axis a little bit in the AZ direction. That there was zero mount flexure on your ASPA star. These and many other assumptions are forced onto the software because it has no way to quantitatively measure those errors and account for them. Because none of these assumptions is true, the mount model now has data that is not correct. In particular, the mount models assessment of your polar misalignment has an unspecified amount of error. Errors in tracking have been improved, but the mount model (pointing) was degraded (slightly). If you redo your sky align, the mount model will once again have an accurate measure of your polar misalignment. It is my belief that if you had checked your polar misalignment after your second 2+4 alignment, you would have found that it was not exactly zero, but rather on the order of a few arcminutes. This was the error you introduced into your pointing model on objects near the pole.
A reasonable question would be "How big are the errors introduced into my mount model as a result of ASPA". The math is complex, and involves many mount/OTA errors that most people don't know and can't measure . But using strictly empirical data derived from re-checking my polar misalignment after re-doing my 2+4, I would say that ASPA will usually get you 90% closer to the pole than you were when you started the ASPA. That is, it converges on the pole with repeated execution until you get within the pointing accuracy of the mount. Despite being only 90% closer, the mount model will now report you are EXACTLY on the pole. (BTW it doesn't always miss by falling short, it could overshoot by 10% just as easily and there could be some mingling of errors between axes).
Finally, consider this, the measurement of polar misalignment is not going to be more accurate than the pointing accuracy of your mount model. For your own mount/OTA this seems to be a few arcminutes. Once you get within that limit, or even close to that limit, continued application of ASPA is not going to cause further convergence. In my experience, if the polar misalignment is less than 10 arcMinutes, then I will see less than one minute of Declination drift in 100 minute exposure set, and for what I do that is usually good enough. If my measured polar misalignment is 15 arcminutes or higher, or if I need to do several 100 minute sets, then I will usually do one more ASPA. As I mentioned, pointing accuracy is not critical for me. But if I redo my sky alignment anyway, I can usually measure a polar misalignment of less than 4 arcminutes on both axes, at least half of of which I would attribute to my mount model not knowing exactly where the pole is. If this level of polar alignment was not enough, then I would do drift alignment.
For my particular application, I do not need ultra precise pointing. As long as I start within 1.5 degrees of NCP, I do not feel compelled to redo my sky alignment after ASPA. I know my polar alignment is "close enough" and the errors I have put in my pointing model (if any) will not affect my work. Sometimes I do it anyway though, just to satisfy my curiosity. Your mileage may vary."
By the way the service reps on the phone don't really have much knowledge on the subject. You don't want to know what they told me to expect out of the CGEM for pointing accuracy!
Posted 13 July 2012 - 01:02 PM
Thanks for the posting (and starting the new thread).
Posted 13 July 2012 - 01:54 PM
1. Selecting the correct alignment star for the ASPA - Choose a star as close to the meridian as possible, and around or below the celestial equator.
2. Using an illuminated cross hair eyepiece when aligning on stars for the 2 + calibration or All-Star alignment star - I was blown away on how much more accurate the polar alignment & pointing accuracy improved after I started using a cross hair eyepiece for alignment. If you are not using a cross hair eyepiece, you are introducing additional inaccuracies every time you align on a star (because you are "guessing" where the reference point of the field is each time you align, instead of using the repeatable position of the cross hair). I was able to go from 4-6 arc minute alignment errors in Alt & Az before using a cross hair EP to 1-2 arc minute alignment errors in Alt & Az when using one. Note: this assumes that you are also using the "up" and "right" motions for the final centering approach on every star.
To anyone that doesn't currently use an illuminated cross hair eyepiece for alignment, I strongly recommend getting one - you may be surprised at the difference.
Additional notes: When doing the ASPA, after doing it the first time I always turn the mount off, turn it back on & redo the 2 star align + calibration stars and check the "display align" to see how close the polar alignment really is. If it is more than 2 arc seconds off in either axis (probably because I didn't do a very accurate job adjusting the Alt & Az axes on the scope mount), I will repeat the ASPA and recycle/realign the mount to verify. I have never had to redo the ASPA more than once to get really good results after I started using an illuminated cross hair EP.
Always do the most accurate job possible on every step. If your alignment star is even a tiny bit unentered (because you slewed a tiny bit to far), back off to the left & down and re approach the star.
After adjusting the Alt & Az axes of the mount for the ASPA, make sure the star is still centered AFTER you tighten down the mount bolts - I have found that my mount will shift a little when the bolts are tightened down, so I have learned to compensate for this by placing the star in a slightly off center position before tightening the bolts.
Beyond the specifics, the most important things to get good results (with ASPA or anything else in life) are: patience, precision, and practice. Never rush the process or cut corners on any single step of the process, ever, and the more times you do it, the better, faster, & more consistent the results will be. Just pretend you are a droid performing a maintenance task, or a physicist in a laboratory conducting an experiment.
Posted 13 July 2012 - 02:19 PM
Posted 13 July 2012 - 02:47 PM
When I set up for imaging there are never any eyepieces of any sort involved, so the framing crosshairs in BackyardEOS, Nebulosity and PHD become extremely handy.
Posted 13 July 2012 - 07:07 PM
Posted 14 July 2012 - 09:38 AM
Posted 14 July 2012 - 09:49 AM
Posted 14 July 2012 - 10:10 AM
People often talk about using a star "near the equator" but really it should be a star near the horizon, but not so low that it is strongly refracted. 20 degrees is probably a good value. The adjustment being made to the mount is effectively alt/az and the equator has no particular relevance. If you lived near the equator and you used a star on the equator - i.e. directly overhead - it would be a disaster. So - I would use a star near the meridian and low in the south (or north for people in S. hemisphere).
The other point is to center the star carefully, but try to do it in one az motion and one alt motion - and don't sit there fiddling with each to get the star centered. This could cause the mount to rotate a bit in angle - allowing the star to be centered, but no longer have the mount polar aligned.
As for turning the mount off - I would avoid it because after ASPA the mount still retains the calibration info from the prior 2+4 process. If you want to improve pointing, then just replace the two alignment stars carefully. This should not take long since you don't need to redo the calibration stars.
Posted 14 July 2012 - 10:19 AM
Posted 14 July 2012 - 10:31 AM
For people at 40 degrees latitude the equator is 50 degrees up - which reduces sensitivity to az motion by a factor of about 0.7. It should still work fine, but a star lower down would let you dial in the az motion more directly. So for most people this stuff doesn't matter much and any star would work pretty well - but for a better star to choose I would use one down low.
For someone at the equator with a star directly overhead, az motion would do nothing to center the star and ASPA would not work at all. But if they used a star about 20 degrees above the horizon it would work fine.
Posted 14 July 2012 - 11:31 AM
Posted 19 July 2012 - 10:40 AM
Posted 19 July 2012 - 12:07 PM
I'm looking over procedures and had a question. When aligning Polaris through a polar scope, I've read to adjust the tripod legs, but wouldn't this throw off the leveling? Does that matter? I've always adjusted using the elevation knob, but that sometimes takes me off my latitude mark by a few degrees. Which is the proper way to do it?
It's hard to obtain a precise adjustment via the tripod legs but if the elevation adjuster lacks sufficient range you can use the legs to get closer. "Level" isn't a word with any meaning in this context so you can ignore any effect on that. The latitude scale on the mount isn't meant to be precise so you can ignore any effect polar alignment has on that, too. Both are used only to get into the ballpark before performing the polar alignment. Polar alignment with a polar scope or ASPA is the goal and once that is achieved neither level nor the latitude scale matter.
Posted 19 July 2012 - 01:14 PM
Posted 27 August 2013 - 08:47 PM
I found that you don't have to power cycle the mount after ASPA. What happens (at least on my CGEMDX) is that the calibration stars are cleared after ASPA. Re-entering them allows me to measure the introduced error after ASPA. This saves a little bit of time in the alignment procedure.
Is this correct? I dont see an answer for this.
Posted 27 August 2013 - 10:00 PM
I find that to be really really good.
Posted 27 August 2013 - 10:50 PM
After doing 2+4 and then ASPA I checked with PEMPRO and it registered as less than 2 arc minutes on both axis. After fiddling with it I was able to get it down to about .6
I find that to be really really good.
What is pempro?
Posted 24 September 2013 - 01:07 PM
Here's another data point, for anyone searching for info on ASPA in the future. I used ASPA on the current version of NexRemote (controlling a CGE), and then tested the result with PemPro. As 'Whichwayisnorth' found, my results in both axes were around 1-2 arcmins off, which allowed me to take 30 minute subs (focal ratio = 1.4 arcseconds/pixel).
Posted 24 September 2013 - 06:44 PM
Posted 24 September 2013 - 09:03 PM
Posted 24 September 2013 - 09:47 PM
Posted 24 September 2013 - 09:53 PM