Time spent looking at the stars is added to your life
Bashful, Misty, and little Ralphie - my heavenly stars
A-P 105 Traveler ~ TEC 140 ED ~ TEC 6 MCT
A-P Mach1 GTO ~ Losmandy GM-8 ~ TV Gibraltar
Quote:Orthogonality of the setup. Are the two axes truly perpendicular, or is there some offset? This includes some looseness, or even some manufacturing or assembly misalignment.
Quote:Movement or shift in the setup if it is a dobsonian. Many dob OTAs slip side to side in the rocker box cradle, and this can introduce some error. When I have used a tube-type OTA, I've put furniture bumbers on the side of the OTA to keep it from shifting in the cradle.
Quote:Stability of the alignment of the optical axis. Does the refractor tube move a bit in the cradle or on the dovetail, or does the SCT mirror "flop" in going through the meridian, or does the mirror shift during focusing of an SCT, or does the focuser shift during the alignment process, or does the instrument need collimation? These all can lead to a warped sense of the sky in the computer. Ways to improve the performance are to make sure the optical tube is secure, check the collimation, never change focus once the alignment process is started, and if you have an SCT, check the user material or vendor to see if there is a way to either reduce the mirror flop or actually lock it down during alignment. And once the alignment process starts, if you have a newtonian on a GEM mount, don't rotate it in the rings no matter how uncomfortable, if at all possible.
Quote:Finally, the biggest error cause I've seen is in the human at the eyepiece. Many people don't actually align their pupil with the center of the eyepiece. This can give some parallax error such that if you change sides of the meridian in a two star alignment, you introduce a double offset from the optical axis.
Quote:Another potential problem is too low a power causing a variance in position. The way I cured my alignment troubles was going to a high power reticle eyepiece that nailed down the center very well.
Quote: Also, I use a defocused star, like a bagel. This is easier to center accurately than a pinpoint star.
Quote:I think that collimating, minimizing mirror shift on SCTs, and using a defocused star in a reticle eyepiece at very high power are great ways to improve the accuracy of star alignments.
Quote:Durval:It sounds like you have system up and running reasonably well, not perfectly but pretty good and you just need to fine tune it. You have done a good job.
Quote:I think your questions are beyond the scope of the beginners forum and would probably be better suited for the mounts forum where people who have real experience with encoders and DSCs hang out...
Quote:I agree with Jon and will move your thread.
Quote:at 1024 steps the best you can expect is about 0.35 degree accuracy. To increase resolution you have no choice other than using higher resolution encoders.Dale
Quote:There are a number of different sources of flexure and inaccuracy in both dobson and equatorial mounts. At 4096 tics (the old equatorial standard) the best one could typically hope for was about 20 arc minute pointing error.
Quote:In my current setups I have pointing errors of around 2-4 arc minutes (on the AP900 with 10k tics) and 3-6 arc minutes (on a G11 with 8192 tics).
Quote:Higher resolution encoders do "buy" some increased accuracy but they typically are about third on the list of things to do.
Quote:1.8 degrees pointing error is not a system I would consider working very well: I didn't like 20 arc minutes
Quote:However, how fine you need to go is a function of aperture and field of view, field of view compensates for a lot of pointing error and aperture helps making things more obvious to the eye (but limits field of view). In any case on my 5" refractor 3-5 arc minutes of pointing error is overkill: twenty arc minutes is plenty of accuracy.
Quote:You have several ways to approach the problem I've tried them all. You can try to nail down the hardware accuracy as Floyd points out the orthagonality is important. So that means shimming and tightening etc.
Quote: You can extract more accuracy from the encoders by upgrading them but if your primary source of error lies elsewhere it won't work too well.
Quote:You can get an Argo Navis computer or hook your encoders up to a laptop with software bisque. What we're getting at here is a pointing model which takes data from the sky and then analyzes how the star/object positions that you report differ from reality, and constructs a computerized compensation. Argo navis offers this. Software Bisque has a program that does it. It is built into Gemini. Other go-to systems have partial versions (not as many variables are modeled, such as Celestron). In any case by far and away the most effective thing I did to improve pointing accuracy was to dump NGC Max and dump Sky Commander and get Argo Navis. You can go to their web site (wildcard-innovation.au) and download the pdf of their instruction manual and then look at the mount modeling/pointing error sections of the catalog. If you zap a few stars the computer will model the various sources of error and then compensate the coordinates. Argo Navis' pointing model got my pointing error out of the 18 to 25 arc minute range down to the 5 to 10 arc minute range, and upgrading the encoders shaved those values approximately in half. So the biggest single thing I did was to upgrade the computer.
Quote:Pointing models for dobs are less complicated than pointing models for equatorial mounts. If you zap stars only in the areas that you can see the compensation model will be most accurate in the areas that you can see. If you have all-sky access you would get a better statistical all-sky performance if you zap all-sky stars.
Quote:How many stars you choose to start with is up to you, it is possible to add additional alignments as you observe (smaller tighter objects are best) so if you do some alignments and then do some observing on small planetaries or tight globs you can do some observing and improve the model as you do so.
Quote:there is a Yahoo group Argo_Navis_dtc and the aforementioned home page for wildcard innovations.My friend Pete used to have pointing accuracy of about ten to twenty arc minutes with his Obsession servoCAT go-to mount. Part of the issue that affects pointing accuracy in go-to systems is backlash compensation. I was able to get greater accuracy because I tended to build 10 star models and because CNC machined aluminum is intrinsically more rigid than wood with inserted truss poles.
Quote:If your encoders rotate at a 1 to 1 relationship with the axis involved, then 360/counts per encoder rotation, will be the the finest angular measurement. 360/4096=0.08789 degrees = 5.27 arc minutes degrees per step. The pointing accuracy depends on factors mentioned by others.
Quote:On my mount the dec encoder is driven directly, and my dec encoder gives me 8096 counts per revolution of the dec axis. The same model encoder is geared to my RA axis at a 2.5 to 1 rate resulting in 20,240 counts per revolution of my RA axis.
Quote:I have been using a skycommander DSC using 1 star alignment and usually point to within 20 arc minutes. If my polar alignment is closer to being on I point to within 5 to 10 arc minutes.
Jc ATM 10" F6.1, 1/25th wave spec (max wavefront error +/- 1/12.6 in zone 4 of 6, sodium light ), 6" F7 spec, 127mm F9.4 achro Refractor, Criterion DX8, 10 x 50 bin, SP mount/Synscan goto. ETX80 (finder) Canon 20D, PST DSI 1, Butenschön 125mm F13.4 refractor, diy spectroscope and curiosity
Quote:Hello CN folks,After a few weeks getting my DSCs operational (long story), I'm still having problems... maybe someone here can help point me in the right direction.
Uncle Rod Uncle Rod's Astroblog: http://uncle-rods.blogspot.com/
Quote:Hello Durval,You can also try PalmDSC from http://palmdsc.dougbraun.com/It is a lot easier to use it than Astromist if you just want a DSC. If you get the same kind of precision with PalmDSC then you will definitely need to address your telescope's mount fabrication issues.Best regards,Lucy.
Quote:a couple of points not mentioned so far, are the encoders perfectly concentric with the axes? does your optical axis coincide with the ota mechanical axis?
Quote:My two workhorse telescopes are 10" f/10 SCT, and 18" f/5 newtonian, so upon initial thinking I was a bit concerned about too small a field of view to get the target star into the field, balanced with the more power, the better. But my decision was taken away by my best friend and spouse for 44 years, who bought me an 8mm reticle eyepiece for Christmas four years ago. I guess that's why I married her! She saw I was looking over choices, and made it for me.
Quote:Since I don't use this eyepiece for any purpose other than star alignment, it works fine at 286X in the big newt and 318X in the SCT.
Quote: One lesson I learned with the one I have is that when there is a reticle position adjustment capability for off axis guiding with the reticle, you need to make sure that the reticle figure remains centered. If it moves off center through use or any other reason, then the alignment suffers as the tube rotates to move from star to star. I find I need to eyeball the back side before I put it in the focuser to make sure the reticle plate is still centered and hasn't shifted.
Quote:Assuming your encoders are OK (nothing slipping, encoders properly mounted and concentric), the probable reason is your alignment procedure. It's hard to say much more since you don't mention the brand of DSCs, but, in general, stay away from stars near the horizon. 20 - 30 degrees up, minimum. And you want the two stars to be as far apart as possible. This time of year I use Polaris and Procyon.
Quote:Leave Astromist out of the mix till you get going. And be aware that DSCs, on most scopes, won't always put an object dead center. If you are getting 10' accuracy, that's about what to expect.
Quote:Nice! Can you mention the make/model? Also, what TFOV does that EP get on your scope?