Do you want to improve GoTo accuracy?
Posted 08 April 2013 - 03:49 PM
In the past I have always aligned my mounts by just estimating visually when an alignment star is in the center of the eyepiece. I would sometimes find the star with a wide field EP and then switch to a more powerful EP to try to get as close as possible. But no matter what I did I always found myself having to nudge the view just a little one way or another after every GoTo operation. It was rare to get something centered well enough that I did not feel the need to move it a little or a lot.
Then I bought the 32mm Illuminated Cross Hair EP from University Optics last month. I have now had a chance to use it with all 3 of my computerized mounts and I have been blown away by the difference in accuracy. Over the weekend I used it with the CGEM for the first time. I made sure the mount was perfectly level before starting and then did a 2 star alignment with the addition of 4 calibration stars. Normally this would have put nearly every object in the FOV of a 22mm 82 degree eyepiece in my C11, which I considered good enough. But after aligning the CGEM with the new EP, I did around 50 GoTo operations before shutting down for the night. On all but 3 of those GoTo's the objects were perfectly centered. I know the coordinates of some objects are not exact so that is what I think caused problems for those 3 objects, but they were still within 10 arc minutes of the center.
I only wish I had thought to do this a long time ago. It saves so much time not having to center objects after every GoTo.
Posted 08 April 2013 - 04:04 PM
Posted 08 April 2013 - 04:17 PM
Posted 08 April 2013 - 05:37 PM
Posted 08 April 2013 - 06:26 PM
If the mount is made to good precision (RA and Dec axes mutually perpendicular), and the scope's optical axis reliably passes through the mount's polar axis (departure here is called cone error), then a polar alignment via a borescope, followed by a single star alignment, would likely be better than obtained in most actual situations.
The requirement for multiple stars is so that software/firmware can calculate and compensate for the various and sundry errors in execution.
Perhaps the biggest one is cone error, and is the one you usually can do something about. One can significantly reduce this by positioning the scope in Dec until the minimum field motion is obtained when swung in RA. Then (for GEM mountings) adjust the OTA so that either the front or aft end is 'lifted' (by shimming, if necessary) until the optical axis coincides with the polar axis.
I would not be surprised if cone error in some cases approaches 1/2 degree. That alone would be an error most worthy of reduction. The result might well be a good reduction in the number of stars required for decent pointing accuracy.
For a friend's 14" LX200, I reduced his cone error, and what formerly were typical pointing errors of 12-15 arcminutes have been reduced to 3-4. And that's not using the "high precision" mode which requires first syncing on a nearby star. For his imaging activities, which he conducts from inside his house, this has made a world of difference.
Posted 08 April 2013 - 08:22 PM
Levelling the mount makes polar alignment a lot easier. Imagine a mount whose North side is lower than the South side. As you adjust the azimuth adjustment, you'll also be changing the altitude. Remember, we're looking at wanting to make adjustments on the order of seconds of arc and on many (amateur astronomer, 8" diameter baseplate) mounts, 1/8 of a turn of the azimuth adjustment can easily move the mount about 0.7 minutes of arc. (Hopefully, I got my math right on this one.)
Adjusting the mount's position on a tripod, where there's a bit of play in the mount's rotating on the tripod just complicates polar alignment. Higher end mounts have finer threads on the alignment adjustments which facilitate accurate polar alignment. These mounts also come with a "fixed" baseplate that doesn't move on the tripod or pier when doing polar alignment. It's a lot easier to machine rotating parts to tight tolerances when you don't have to plan on them being disassembled and reassembled every time they're used.
Posted 08 April 2013 - 10:18 PM
Posted 08 April 2013 - 10:37 PM
I can throw down my GEM mount on any ground and have Polaris dialed in within 5 minutes, 10 at most. A tripod tilt of even 10 degrees (or more, as long as stability is not compromised) imposes no difficulty.
It must be stressed that a polar alignment to an accuracy measured in arcseconds would require a lengthy drift-alignment session. And even then, no one polar altitude will apply over all parts of the sky due to refraction. In the real world, a polar alignment to 1-2 arcminutes is about all that can be expected from a borescope. But this us better than the 5 arcminutes a Celestron engineer stated is possible when relying only on the mount's controller routines using multiple stars.
If the borescope's rotating scale is used, the mount *should* be reasonably level in the E-W direction. But even if the mount were to lean to the L or R by 10 degrees, the induced error in azimuth is a pretty small 0.7 arcminutes. That's tantamount to being 'in the noise.'
My main point is to assure all that *precise*mount levelling by itself is in no way a prerequisite to good alignment and pointing accuracy. It's really extra effort for little to no gain, except for the more compulsive among us.
Posted 08 April 2013 - 11:02 PM
Incidentally I did not use the cross hair eyepiece on Polaris I just roughly put it in the center of my 40mm Pentax. I only used the cross hair eyepiece on the alignment stars. This is visual viewing not AP so I really care nothing about getting a good polar alignment.
Posted 08 April 2013 - 11:21 PM
Get yourself a wide field illuminated cross hair eyepiece. ......
Then I bought the 32mm Illuminated Cross Hair EP from University Optics .....
I’ve found this to be true using my old 12mm guiding eyepiece to set up my CG-5 GoTo mount.
I’ve wondered if the same improvement would result from using a cross-hair eyepiece when initializing the NGC-MAX on my MI-250 GEM and my Dob??? However, that would definitely require a wide-field unit like the University Optics eyepiece.
Posted 09 April 2013 - 03:15 AM
Centering Polaris is perhaps as important as any other star if used like other stars in the computational routines.
Actually, I would prefer if all mounts could just not use Polaris at all, or at most only as a way to first get a rough alignment. It's so near the pole that mount and (especially!!!) cone error can result in the mount 'thinking' its polar axis is closer to the pole than it is. Of course, the numerous subsequent star alignments (nominally) compensates for this by modelling the very error. This is OK for visual, where pointing accuracy is paramount, but the potentially large polar misalignment could result in not insignificant Dec drift, limiting the duration of unguided astro-images (including piggybacking with telephoto lenses, not just through the scope.)
My take is that if something can be done to assure reliably better polar alignment *and* improved pointing accuracy, why not do it? Cone error is one area in which it pays to at least assess. If it's found to be no worse than about 10 arcminutes (5 is a better limit), you may 'safely' overlook it.
Posted 09 April 2013 - 03:21 AM
But I do know that tracking will not be very good on the CGEM if you are not pretty close to leve
Is this something you have tested? Polar alignment does not require that the scope be level, it just requires that the RA axis of the scope be aligned with the earth's polar axis, the axis of rotation.
Posted 09 April 2013 - 08:33 AM
I've read somewhere here a couple of months back a relaying from a Celestron engineer who stated that a mount-assisted polar alignment (presumably relying on what I consider to be a dubious use of Polaris) could be expected to not be better than 5 arcminutes. After using the all-sky (or some such name) alignment procedure. A polar borescope can handily best this. Levelling not required.
Does this mount permit the option of selecting alignment stars *after* you've boresighted on Polaris, and thus already have a pretty decent polar alignment? I ask because if the alignment procedure requires you to UNDO YOUR GOOD POLAR ALIGNMENT obtained via the borescope by now adjusting the mount in order to bring Polaris into the field center, that's not so nice.
(The just-outlined condition is what the LX200 mount imposes when you want to align it. After setting up a very accurate drift alignment, when it slews to where it 'thinks' Polaris is, it's best to hit ENTER without moving *a thing*; certainly not the alt-az adjustments, for then the hard won polar alignment is buggered.)
And so the best procedure, if available, is to polar align with the borescope, then run through the star alignment routine, *perhaps* leaving Polaris out of the equation, all while without adjusting the mount head whatsoever.
If the mount's axes are nicely perpendicular and the cone error is small, it may require to use no more than two strategically selected stars to realize good pointing. The very requirement to employ numerous stars in order to finally enjoy decent pointing is a clear sign of one or more of the three principal errors rearing their heads; polar axis alignment, axis perpendicularity and cone error. Not to say other gremlins aren't present, such as flexure/sagging/looseness and awful mirror flop on a Cat. But they can be easily checked for and minimized.
In order to use a polar borescope with confidence, it must be itself boresighted to the mount's polar axis. Instructions cannot be far away.
I've described the essence if the procedure to assess and correct cone error.
Mount axis perpendicularity can be assessed via creative use of a digital level (clinometer/inclinometer) resting on the OTA, while flipping and rotating the mount about. Or...
With a good polar alignment, find a pair of stars sharing the same RA (for the current epoch!) and which are very nearly equidistant from the celestial equator (e.g., one at +22.5 and the other at -22.5 Dec.) While the mount tracks, sight one star under the crosshairs, preferably oriented N-S/E-W. then slew to the other star, noting how far E or W of the crosshairs it lies. Go back and forth a few times to ensure consistency.
You MUST slew in Dec only! Therefore press just the Dec buttons, or perhaps better yet, unlock thecDec axis and move manually (if this can be done without inducing any unwanted shifting in RA and such.)
You can estimate the offset of the star from the crosshair, or if significant, drift time it (in seconds). The actual angle of this offset in degrees equals
Drift time / 239.3 * Cos(Dec)
For example, suppose the stars are at +/-22.5 degrees Dec, and the offset was drift timed as 4 seconds. The angular offset is
4 / 239.3 * Cos(22.5) = 0.0154 degrees, or 0.93 arcminutes.
Because we are sampling over a limited arc on the celestial sphere, this angle does not reflect the actual error in axis perpendicularity. We need the *true* angle at which the scope crosses a line of RA while passing through the equator. A rigorous treatment requires spherical trigonometry. But fear not! A simple approximation suffices for our purpose. The formula is
ATN (drift angle / Dec range), where ATN is the inverse tangent
We calculated our drift angle (for 4 seconds) as 0.0154 degrees, the stars' Dec range is +/-22.5, or 45 degrees. And so the approximate departure from axis perpendicularity is
ATN ( 0.0154 / 45 ) = 0.0196 degrees, or 1.18 arcminutes.
This approximation becomes more accurate the smaller is the Dec range for the chosen stars. But of course you want a reasonably large Dec range for increased sensitivity. I find stars at Dec +/- 10 degrees are fine, as you can still detect non-perpendicularity down to the hyper-fine level of better than 10 arcseconds (which requires estimating the star's offset from the crosshair, it being too small to drift time.)
If the axes are perpendicular to within 1-2 arcminutes, you're doing fine. You could predetermine the conditions in a pass/fail scenario, where for the selected stars (from planetarium software, and while reasonably near the meridian, preferably) you have precalculated a maximum drift time for your permissible departure from perpendicularity.
I know this is all details that might seem to be far beyond anything necessary. But you never know, do you? You might discover you have a gem of a mount, or one that warrants replacing...
Posted 09 April 2013 - 09:38 AM
Posted 09 April 2013 - 10:10 AM
Dark skies for all beings.
Posted 09 April 2013 - 03:40 PM
Posted 09 April 2013 - 03:44 PM
Posted 09 April 2013 - 03:46 PM
So just having a rough alignment on polaris is good enough. And by rough I mean that it was probably off by less than 30 arc-minutes. Combined with a level mount, that was good enough to keep the object in the center for way longer than I needed, over 10 minutes at least.
This isn't AP I am doing here just visual viewing. I only need the GoTo to center the object and then have the mount hold it there for at least 5 minutes. By then I am off to the next object anyway.
Posted 09 April 2013 - 03:54 PM
The really big improvement was not actually on the CGEM but instead using this new cross hair eyepiece with my Sky Commander and push-to mount. You only get 2 stars to align with on Sky Commander and getting them both dead on is very important. It still is not nearly as accurate as the CGEM but it gets much closer now and allows me to keep a higher power eyepiece in the diagonal as I push the mount around from object to object.
Posted 09 April 2013 - 05:16 PM
It works really well, especially with obstructed scopes because if you are a bit defocused you get a doughnut with a tiny bright spot in the centre - it's a diffraction effect. The doughnut shows the cross hairs and the bright spot gives an accurate reference. The trick is not to move the EP or refocus during the alignment process, this way errors in the positioning of the cross hairs don't matter.
Posted 09 April 2013 - 10:01 PM
Posted 09 April 2013 - 11:29 PM
Posted 10 April 2013 - 02:08 AM
This is what I'm talking about!
Posted 11 April 2013 - 02:33 PM
Posted 11 April 2013 - 03:01 PM
Thats what I use and it works a charm. Improves GoTo no end.