Drift align questions..
Posted 17 May 2009 - 04:44 PM
Would up with consistent readout from Autostar of less than 5' from pole.
I then shot 90 minutes of a high dec target (NGC3953) just of the SE corner of the dipper's bowl. During this time I had about 1/3 frame drift in dec and about 1/3 frame in AZ. (About 5' in RA and 3' in dec)
1) Is less than 5' polar error enough to cause this over this time period?
2) Are there areas in the sky, such as polar regions, where misaligment will result in more visible drift over time than another position in the sky, such as the equator? In other words, are some directions more of a stress test on an alignment?
3) Does mislignment ever manifest itself as RA drift as well as DEC error? I ask this because during classic drift aligns, you're always looking for DEC error.
4)Obviously the holy grail is zero polar align error...realistically, what is usually acceptable? For now, my typical setup is my SN8 (800mm FL) and my DSI-C.
I've suspected I need to play with the custom tracking rate a bit as i think the LXD is running a tiny bit fast, so I'm trying to sort out how much of what I see could be polar error and how much is due to tracking speed.
Posted 17 May 2009 - 09:08 PM
First off: what type/magnification are you using for an illuminated reticle EP in that scope?
I'm using a 5mm Orion IR EP, and I've done 10 minutes drift alignment w/ no motion....
The Polar alignment routine for my MI-250 says that I should be able to get within 2' using it...
I routinely got within 2-3' with my old CGE Polar alignment routine, so 5' isn't particulary "good" IMHE....
So, if Autostar is reporting 5', I'm thinking either:
1) Autostar isn't particularly adept at polar alignment/reporting, and/or
2) your IR EP is too low a magnification to get the precision you really need?
In thinking about it, I would think that the closer you are to the NCP, the LESS the issue with drift would be... (same angle error is much less actual distance... - but am quite willing to admit this logic is seriously flawed!)
Separately, RA error is just as bad as DEC error; just less noticeable (unless/until your mount can't guide it out)...My *understanding* (quite willing to admit I'm wrong on this one - only from what I remember I've read) is that Meade does guide/tracking at 2X Siderial (versus the sub 1X available on other systems)...
Anyways, on my 200mm F4 R200CF (same 200mm, same f4), I find that getting to ~2' or better is well worth it....
clear enough skies
Posted 17 May 2009 - 10:36 PM
During align I am using the DSI, barlowed at 2x, and the reticle screen in Nebulosity. I believe this give me about 9'x7' FOV or so and about 2" per pixel. Last night I ran for 15 minutes or so in the E position and another 15 in the meridian position and had good results, so I decided to go ahead and start shooting again and got what I described.
In thinking about it today, it seems as if higher latitudes should not only show the misalignment worse, but it should translate to RA error as well. I'm beginning to suspect the reason the standard drift method is standard is that near the meridian, error is nearly all DEC and so it's easier to deal with.
Posted 18 May 2009 - 02:44 AM
Posted 18 May 2009 - 11:46 AM
Last night I did 113 minutes on NGC3718 with the custom tracking rate set to -100. It did slow down the RA drift, but i'm still not sure wether I'm compensating for how much NCP misalign and how much slightly off tracking rate.
Posted 18 May 2009 - 01:10 PM
Oh, and I've also got a 3x barlow on the way from the shop and swap, this should increase my leverage on getting the error to show up faster for some fine tuning.
Posted 19 May 2009 - 01:54 AM
Posted 19 May 2009 - 06:31 AM
Edit, does it work in simulation mode?
Posted 19 May 2009 - 06:46 AM
Posted 19 May 2009 - 12:25 PM
NO software to install and nothing hard to do. Just keep your finger on the button and watch the screen.
Posted 19 May 2009 - 01:02 PM
there are so many ways to skin the cat on this, it's kind of fascinating in it's own right.
Now, on EQalign...
I installed V22.214.171.124 and couldn't get past the start wizard..I'd get in site, scope and when it went to start, i'd get an 'instance not found' error I think it was, and a chain of ASCOM barfs. Can't run the simulator, can't do anything. Can't imagine that I must be connected to the mount just to get the application up.
I went to their wiki site and someone else had has this problem, and the author recommended a beta version upgrade which supposedly solved it.
Nope. Now it hangs even earlier! Now, I seem to remember that this ASCOM thing is more complicated than I thought..I assumed it was just a format call, but apparently it is a visual basic or something SW base that must also be installed or something?
I was hoping I'd just click "LXD75" or whatever, but that's not the case. I think mine is ASCOM compatible, but what do I know. I guess here's where I learn even more minutae.
Well, the EQAlign looks very potent and i'll still work on getting it going..but that site Paul put up looks good too, so I'll give that a shot as well first.
And still...can anyone explain if high dec targets will highlight polar error more rapidly, and also turn align error into not only DEC drift, but RA drift?
Posted 19 May 2009 - 04:51 PM
Posted 20 May 2009 - 03:17 AM
Posted 20 May 2009 - 10:50 AM
Was ready to try the neato CCD alignment method Paul presented last night, but of course, clouds. Maybe tonight.
Posted 20 May 2009 - 12:05 PM
Posted 20 May 2009 - 12:15 PM
Posted 21 May 2009 - 11:43 AM
some comments on the 'new' method
previously, i'd been using the barlowed DSI and watching for drift using the crosshair reticle in Nebulosity. the benefit to moving the scope in the new method, is that offsetting the star for a period of time allows you to see the delta in a much clearer fashion than simply trying to see if it's moved while it sits in nearly the same spot.
I had to do a slightly different method than from the linked site, but using the same ideas.
my imager has nowhere near the sensor size in the example so running for a couple minutes would take me well off the sensor, then running back to the same spot would only bring the star back in the last 30 seconds or so. no big deal.
but after determining the key is the time you wait, not the distance you move, i cut out holding down the guide arrow for minutes only to reverse and do it the other way. Now I just settle for moving the star to the other side of the screen, letting a fixed amount of time (2 min for now) elapse,and then moving it back, again at guide rate.
The other difference and the key to doing it this way is I am not running single frames, I am stacking, so I can see where the star is going as I move it. In the single frame method, the hold down (the button) times are crucial because you cannot see where the star is or was until the frame is complete. thus to get it all in the right place, the single frame method really relies on decent timing.
me, i was running one second frames and just letting them stack, with no alignment target. however, I found I had to turn off drizzle because it was still trying to align stuff and was doing so with noisy pixels at random, thus rotating and shifting the accumulated frames...messing up the entire point of using this method which relies on fixed frame geometry with no shifting, alignment, or rotation.
So i read up and found that the 'deep sky' method on the DSI didn't rotate or align or modify anything if you didn't pick targets, it just stacked frame on frame aligned at the raw frame edges. perfect.
here's what i do now
1) put star at left edge.
2) start capture
3)leave it there for 15 secs to accumulate a bright marker for the initial point.
4) move star at guide rate slew the star to the right hand boundary.
5) wait 2 minutes.
6) move star at guide rate slew back to near left hand boundary, ending as vertically aligned as possible with the initial point
7) let image accumulate there for 15 more secs to get a solid endpoint dot
now check the delta, adjust and do it again. before, i'd simply let it run and watch for drift up or down, or accumulate and watch for a streak. what I like about this method is that moving the target off the initial point after making a 'mark' there, and then moving it back to make another mark, leaves two very distinct points which are easy to accurately measure and see how you're doing. whn it was live or a streak, this was much harder.
Posted 31 May 2009 - 11:00 PM
By applying the simple steps outlined above, (modified from those presented in the link to suit my setup), i've achieved what is my best polar align to date. I ran an hour two nights ago with almost no detectable dec drift using the DSI at prime focus of the 800mm FL SN8. It is sure nice to see the stars stay put, except for a tiny bit of periodic error, when each fresh frame posts. I am totally open loop on this mount, no autotracking, I am relying on good polar and a drive that works. My frame times are 60 secs max, now with it looking good I may try bumping that up.
I did a three stage process, DSI alone to get in the ballpark, then a 2x barlow. The following night I switched over to a 3x barlow, and that's as far as I went.
At some point I will deal with the ASCOM drivers to try the other SW presented, but I'll save that for when I get my Skyview Deluxe pier header done and the mount tuned. I don't want to mess with the LXD75 now that I have it close to nailed.