Learning to use setting circles
Posted 10 December 2013 - 11:28 PM
I have reviewed your details and concur with your assumptions.
In reality there is always some deviation in production and assembly of components into finished products. At Questar we strive to produce the highest quality product every time on every unit.
There are many possibilities that could cause a deviation of 1 degree.
The first I will discuss is easily depicted in your photograph. Depending on the amount of clutch friction and spring tension of the AZ axis could tilt away from vertical due to telescope and level weight. The Questar drive system is a fine balance of drive friction and clutch pressure. There are several arc spring washers that apply a downward pressure on the Questar azimuth axis under the turntable cover plate. If this pressure is not correct or the arc washers are not working correctly the weight of the telescope and level could make the telescope flex past vertical.
The second is the actual attachment of the declination circle. This circle is pinned in place and usually glued. If the glue adhesion breaks down this circle can move slightly, more than a degree even if pinned. You can usually see this if you lock the declination clamp and then try to move the telescope in the declination axis. You will feel the telescope move but the declination remain in place. This is generally checked on all service and repair.
The next item is the declination line on the sidearm. This line is manually inscribed and could be off by some small amount. The line thickness could also come into play and could easily be off by .5 degree.
The last item is the actual way we square the optical axis. The telescope assembly is placed on a granite table. The base is checked for flatness. The telescope tube is then paralleled to the table and at 90 degrees to the azimuth. We use a dial indicator to indicate along the tube top diameter from one side to the other until it is exactly the same on both ends of the tube. This would indicate the the tube is perpendicular to the bottom of the base. Depending on the skill of the person and reading the indicator this angle could be off by a small amount. Once the tube is in this position the line is then inscribed on the sidearm to match to the declination circle.
If you put together the possible accumulation of errors it is feasible that the circle could be off by 1 degree.
If you have determined the declination reading to be in error it is relatively easy to compensate as long as you know the error amount.
Posted 01 August 2014 - 05:52 AM
I kept wondering why you, Michael, concentrated on such earth-centered methods of polar alignment (spirit levels, latitude measurements, protractors, etc.) instead of cutting to the chase and using star-based methods like I prefer. Then I realized that you're in the Southern hemisphere with no blatantly obvious pole star. Even in the northern hemisphere I sometimes find myself in a location where Polaris is not visible, and I don’t want to be “stuck”. So in the last few months my brain has been simmering away trying to think of a fast method for star-based alignment without using a pole star, which would work in either hemisphere.
The drift method looked time-consuming to me, so I thought a two-star recursive method would be better. I think I've got such a method perfected now, at least in my mind, though I'm probably not the first. I'd like to see if others think it will work. (It’s hard to test it because the Powerguide shortage has delayed delivery of my scope, but I digress.) The method requires either a tripod with adjustable alt-azimuth, or the Questar’s table-top legs on a sturdy table. These instructions assume that you know the crucial difference between the Right Ascension (RA) and Declination controls (on the scope) and the altitude and azimuth controls (on the tripod or legs under the scope).
1) Turn on the RA drive.
2) Do a rough polar alignment with the alt-azimuth controls, using a compass or other local knowledge.
3) Dial in the Dec coordinates of the first star, and move the scope in RA manually to get as close as possible to the first star. Then use the alt-azimuth controls to center the star in the field of view.
4) Move the RA ring with your fingers to dial in the RA coordinates of the first star. Don’t touch the RA ring again, as we’re going to align against this calibration position.
5) Dial in the RA-Dec coordinates of the second star, and adjust it to the center of field using the alt-azimuth controls.
6) Dial in the RA-Dec coordinates of the first star, and adjust it to the center of field using the alt-azimuth controls.
7) Repeat steps 5 and 6, flip-flopping between the two stars, until you don’t need to move the alt-azimuth controls any more to center the stars in the field of view.
That’s it. It seems to me that this method would recursively converge on perfect alignment, rather than flip-flopping between two bad alignments. Am I correct?
Obviously the two stars should be in different parts of the sky, preferably far apart. But best not get too close to the horizon to avoid diffraction effects. (Stars, moon, and sun at the horizon are shifted by 1/2 degree!)
Ideally, if you could find two bright stars that have the same Declination (or the same Right Ascension) then flip-flopping between them would be faster as you would only have to change one coordinate each time. Surprisingly, I don’t think this would lose any accuracy, as you are still using two coordinates for each star and adjusting in both altitude and azimuth to re-center each time.
Comments please, especially from any southern hemisphere folk who might find it particularly useful.
Posted 01 August 2014 - 06:13 AM
Posted 02 August 2014 - 10:27 AM
And yes, Questar sells a cross-hair (cross line reticle) eyepiece (PN 19151 for a little over $500).
Posted 02 August 2014 - 04:37 PM
Posted 09 August 2014 - 05:00 PM
Comments please, especially from any southern hemisphere folk who might find it particularly useful.
I may be able to try this tonight. If/When I do, I'll report back.
One thing you left off is that the error in the Dec circle must be determined and used to set the correct Dec position. Otherwise you'll be moving the alt-az adjustments to try to fix this error.
Earlier in the thread, if I understood you correctly, so said that Michael could use the tripod adjustments to compensate for an error in the Dec circle. Not so. If the Dec circle is off by 1 degree, then it traces a 2 degree diameter circle around the pole. If he uses the tripod to compensate for the error, he will be off by 2 degrees on the meridian pointing south. The easiest solution is to determine the error and just subtract it when setting the Dec. The only real mechanical solution is to put some sort of sticker on the fork arm to replace the index mark. To make lemonade out of lemons, you could scribe additional marks to turn it into a vernier. But I don't know how to do that. Fortunately mine is very close.
I have shimmed one fork arm on my Questar to compensate for a slight variation, getting the error down to about 10' arc. That was with a slice of sheet metal I had on hand. A more accurate measurement and a precisely cut shim could essentially eliminate the error. I also use a chart of the pole to star hop from Polaris to the true pole for a more accurate alignment. It is a pleasure when an object stays put for hours on end.
Posted 09 August 2014 - 09:49 PM
I found a website for our Brothers and Sisters in the Southern Hemisphere. http://www.users.on....lignment method
It gives a really good discussion of the topic from a Southern perspective. One thing I noted, was that the author suggests moving only half the distance on each iteration. I believe this was so you wouldn't overshoot the mark but I need to read it for more detail. I have read the 1/2 the distance instruction in the past.
I hope to visit down there one day and put these suggestions to use.
On the discussions concerning the mechanical alignments of the telescope itself - I have found that the objects are usually within 1/2 degree and always within 1 degree of where I'm looking. Most of this I attribute to parallax issues with me reading the scales and my very liberal Kochab alignment tolerances . 1 degree resolution on a 4 degree RA increment is very achievable. 1/4 degree resolution on the Dec is too. But then I'm from the slide rule and manual vernier generation.... :-D
I've never done imaging but if I do one day I'll use the drift method to really true things up. For daytime observation of the planets and bright stars I use the method that Michael outlines at the first of this thread. use the level and a compass adjusted for magnetic variance, move the setting circles to the sidereal time. Get the coordinates and go to the sun ( or moon if its out) and begin to make adjustments to the mount.
Really good discussion.
Posted 11 August 2014 - 05:28 PM
The two-star recursive method is just for standard star-based alignment - not for correcting Dec circle mis-alignment. I, too, once shimmed a fork arm to correct a mis-alignment. It was on a Criterion 4000 (not a Q), and I then needed to remove the shim after I recollimated the optics.
Moving the Alt-Azimuth controls half the amount needed to center the object sounds reasonable to me for a recursive situation, as you don't know which side the error is on. Thinking about it further, moving the full amount could conceivably result in a flip-flopping between two bad alignments instead of converging on perfect alignment. I figured I wasn't the first to think up this method.
It also occurred to me that you can probably save time by using this method "sequentially" (on a series of objects) rather than "recursively" (on two objects). When you dial in the RA-Dec coordinates of each object during the evening, if the object is not in the center of the field then the temptation is to use the RA-Dec controls to center it. Resist the temptation. Instead, use the RA-Dec controls to move it HALF-WAY to center, then use the Alt-Azimuth controls to move it the rest of the way to center. Subsequent targets should then be ever more accurate. This would also tend correct for any periodic error, as any given setting would remain accurate as you explored that part of the sky.
All the RA circles that I have seen have increments of 1 degree (4 sidereal minutes). So a resolution of 1/4 degree (15 minutes of arc, or 1 sidereal minute) is not only possible - it's my goal. Accuracy within 1/2 degree (30 minutes of arc, or 2 sidereal minutes) means that the target object will still fall within the approximate 1 degree field of view that the Q has on low power.
Posted 11 August 2014 - 07:23 PM
I stand corrected on the arc minute vs. degree graduations. Thanks for setting me straight! My fingers getting ahead of my brain.
i like your sequential approach and will give it a try if it ever clears up.