Jump to content


Photo

Questions about guiding: PHD2, drift & eggs...

  • Please log in to reply
10 replies to this topic

#1 Jon Rista

Jon Rista

    Vanguard

  • -----
  • Posts: 2198
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 04 March 2014 - 05:28 PM

I am hoping someone here with the knowledge can help me solve a problem I've been having. A couple, actually.

First is with PHD2 and guiging. I imaged flame and horse head nebulas about a week back. During those nights, my RMS error for RA and Dec in PHD2 was pretty good, considering that I've had practically no experience so far polar aligning and drift aligning. My RMS for RA was ~0.08 for several nights, and DEC RMS was ~0.1. Two nights ago, I was trying to image Pleiades. I set up a single 40-frame 210s f/4 ISO 400 sequence in BYEOS, told it to dither with PHD, and just let it rip.

I checked on it a few times during the approximate 3-hour period it was imaging, and I noticed that my RA RMS was between 0.25 and 0.29, and DEC RMS was between 0.27 and 0.3. That is about three times worse than I was getting previously. I always check and fine tune my polar alignment each night before I start imaging, and it is never off more than a tiny fraction of a turn in either alt or az. The only difference between my prior imaging sessions with flame and horse head, and the recent session with pleiades, is that I used AstroTortilla to plate solve on the western side of the meridian.

I should also note that I have not yet programmed PEC. I have not yet figured out how to do that with EQMOD, which seems to involve getting guiding logs from PHD2, and I have yet to produce any.

My questions are, first, could AT have possibly done something that would have affected the guiding precision? If not, what would cause the issues I was experiencing? Earlier on in the night, when pleiades was mostly overhead, the subs seemed to be a bit better, most of the stars were pretty round. As the night wore on, and pleiades sunk towards the horizon, the issue seemed to get progressively worse, and the latter half of my subs (about 20 or so) have small but clearly egg-shaped stars.

Are there any tutorials that cover how to fine tune guiding with PHD2?

One other question I have has to do with drift alignment. I'd done a little bit of drift alignment prior to my flame and horse head nebula session, not much. It seemed as though the stars I was drifting with had very little movement off the retucule over a period of about three and a half minutes, which is as long as my exposures have been so far.

I did some more drift alignment last night, as it was lightly clouded and not good for imaging, but stars were still plenty sharp in BYEOS. I noticed that over a period of about 2-3 minutes, there was indeed some downward drift when using a star near the meridian, so I started tweaking azimuth. I must have tweaked my azimuth back and forth across the entire range of freedom possible with my Orion Atlas mount, and for some reason, all the stars near the meridian and celestial equator that I tried drifting with over a period of about an hour ALWAYS drifted downwards. It did not seem to matter what adjustments I made in azimuth, they just kept drifting, and at about the same rate.

What error am I making here? I'm sure the error is between mount and lawnchair here, but I'm not sure what it is. I thought that it was important to angle the reticule such that when you moved the mount east and west in RA, the star would travel horizontally through the center of the horizontal pair of rules in the reticule. That was the first adjustment I made, to reorient my camera such that that was true. My lens is 600mm, so magnification is not very high, but with BYEOS it is easy enough to make sure the drift star is bisected by one of the reticules horizontal rules, and after about three minutes it was always pretty clear that the star was drifting downwards, albeit slowly. When running the BYEOS drift exposure mode, which renders the drift star's trail, I could even barely see the slight oscillation of periodic error in my worm gears, which appeared as slight left-right movement of the star trail.

But for the life of me, I simply could not get the star to stop drifting down no matter how much I adjusted in azimuth. I literally took it to the extremes in both directions, and I really couldn't tell that there was much change in drift. This is kind of making me go crazy, because I did this before, and it seemed like a synch, and I am just dumbfounded as to why it no longer seems to be working. :question:

Thanks for any assistance!

#2 SergeC

SergeC

    Apollo

  • *****
  • Posts: 1011
  • Joined: 23 Mar 2012
  • Loc: Gainesville, FL

Posted 04 March 2014 - 05:35 PM

This site has an autoguide calculator and a very detailed discussion of the factors affecting guiding. Although not specific to PHD(2), the information will help you optimize parameters.

#3 shawnhar

shawnhar

    Fly Me to the Moon

  • *****
  • Posts: 5308
  • Joined: 25 Jun 2010
  • Loc: Knoxville, TN

Posted 04 March 2014 - 07:52 PM

Eqmod will record and apply the pec easily, click the little plus symbol in the upper left and it will change what you use on the screen, one of those is PEC. Get guiding going, click record, let run for 40 minutes. You can do this while imaging if you don't use dithering. The push play and it will apply, it should load it next time you start eqmod. You have to park though, loose power or don't park and you loose pec.

#4 shawnhar

shawnhar

    Fly Me to the Moon

  • *****
  • Posts: 5308
  • Joined: 25 Jun 2010
  • Loc: Knoxville, TN

Posted 04 March 2014 - 08:00 PM

That is really weird about adjusting your azimuth and not seeing an immediate change, there is no way it was actually moving the entire range if you saw no change.
I can do 4 or 4 turns in az and see the difference in 30 seconds at the equator.

#5 Jon Rista

Jon Rista

    Vanguard

  • -----
  • Posts: 2198
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 04 March 2014 - 08:27 PM

Sarge, thanks! Excellent article.

Shawn, thanks for the tip about PEC. When I started to read the EQMOD PDF about PEC, it appeared as though they were doing a fairly complex procedure using PECPrep. It seemed a bit complicated, as programming PEC with the hand controller seemed easier.

As for the drift stuff, I'm completely mindboggled. The only thing I can think is that somehow the orientation of my sensor frame was wrong...is it possible that the drift I was seeing was in the opposite plane? What is the proper orientation of the frame when doing dift? When you adjust in RA, should the star move left to right, or up and down? Also, every article I've read about drift alignment only seems to mention newtonian scopes as needing to reverse the directions of the corrections...is it possible that with my lens/DSLR setup, I also need to reverse directions?

Either way, I would have still expected an obvious change in drift rate once I had moved azimuth to one or the other extreme. The drift of my stars, which I always placed in exactly the same position on the reticule before each test, always took about 270 second to become noticeably off the line, and by 300 seconds the star was usually dead center between the two horizontal lines on the reticule. By that time, I could usually make out the slight oscillation in the line, which I'm fairly sure was PE (it was a pretty regular oscillation), not drift in the other axis.

My expectation would have been to see drift in the other axis, and a more pronounced change in the primary drift of my star as I changed Az. It really seemed as though none of the adjustments I was making had any effect. I finally ran out of stars in close enough proximity to the meridian and CE that I gave up (and now, I'm pretty sure my once-excellent polar alignment is WAY out of wack. Doh!)

#6 SergeC

SergeC

    Apollo

  • *****
  • Posts: 1011
  • Joined: 23 Mar 2012
  • Loc: Gainesville, FL

Posted 04 March 2014 - 08:39 PM

Try drift aligning by watching the PHD graph (assuming your guider and OTA optical axes are parallel and concentric) by watching the direction the DEC line drifts with DEC guiding turned off, especially if you check Trendline in PHD2. You'll see drift much more quickly than by eye, and can adjust with much greater resolution.

EDIT: Don't mean to imply you can only do this thru a perfectly concentric guider. You can always just mount your guide cam to the main scope for drift alignment. The point is, the speed and accuracy of aligning with a camera is vastly better than with a reticle eyepiece.

#7 Madratter

Madratter

    Fly Me to the Moon

  • *****
  • Posts: 6073
  • Joined: 14 Jan 2013

Posted 04 March 2014 - 08:41 PM

Is it possible that you were pointing in the direction where drift isn't affected by azimuth, but by the altitude adjustment?

#8 TimN

TimN

    Apollo

  • -----
  • Posts: 1416
  • Joined: 20 Apr 2008
  • Loc: Ontario, Canada

Posted 04 March 2014 - 09:33 PM

If seeing is bad drift alignment is very difficult. That may be your problem.
You can also try a software solution for drift alignment. I have been looking at drift alignment and as a computer guy I think a computer solution may be easier, more accurate and less prone to problems. I've done some research and should I ever get clear skies again I will try this program:

http://www.alignmaster.de/

Free to try for 30 days and cheap to buy.

#9 Jeff2011

Jeff2011

    Surveyor 1

  • *****
  • Posts: 1805
  • Joined: 01 Jan 2013
  • Loc: Sugar Land, TX

Posted 04 March 2014 - 10:08 PM

I checked on it a few times during the approximate 3-hour period it was imaging, and I noticed that my RA RMS was between 0.25 and 0.29, and DEC RMS was between 0.27 and 0.3. That is about three times worse than I was getting ...



I recommend reading the dithering section in the BackyardEOS manual. Guylain does a good job explaining it. Basically you are going to get artificially bad RMS readings when dithering because dithering moves the searchbox which makes it look on the graph that the guide star has moved quite a bit. If you watch it dither you will see the graph jump sometimes in both RA and Dec.

#10 josh smith

josh smith

    Viking 1

  • *****
  • Posts: 866
  • Joined: 13 Oct 2013
  • Loc: Pittsburgh, Pennsylvania

Posted 04 March 2014 - 10:14 PM

Earlier on in the night, when pleiades was mostly overhead, the subs seemed to be a bit better, most of the stars were pretty round. As the night wore on, and pleiades sunk towards the horizon, the issue seemed to get progressively worse, and the latter half of my subs (about 20 or so) have small but clearly egg-shaped stars.


That portion sure sounds like differential flexure. Unfortunately that is hard to diagnose when dithering. You can do a capture with both cameras while pointed straight up and then with both when pointed towards the horizon and see the difference in the deltas between the center of fields. This should let you know how much flexure you have and determine if that is an issue or not.

#11 Jon Rista

Jon Rista

    Vanguard

  • -----
  • Posts: 2198
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 04 March 2014 - 10:53 PM

Try drift aligning by watching the PHD graph (assuming your guider and OTA optical axes are parallel and concentric) by watching the direction the DEC line drifts with DEC guiding turned off, especially if you check Trendline in PHD2. You'll see drift much more quickly than by eye, and can adjust with much greater resolution.

EDIT: Don't mean to imply you can only do this thru a perfectly concentric guider. You can always just mount your guide cam to the main scope for drift alignment. The point is, the speed and accuracy of aligning with a camera is vastly better than with a reticle eyepiece.


I actually had PHD running with guiding disabled. I did notice that the dec trend was at a relatively steep angle upwards towards the top of the graph. It also seemed even more apparent in the guide image that the star was drifting. I do think I need to correct the orientation of my guiding image sensor, however. It seems to be at an odd angle to the RA/Dec plane, and at the time I wasn't sure if the angle of the dec trendline was due to that, or something else. The RA trendline was also at an angle, also towards the top of the graph, however it was at a much slower rate than Dec.

I did notice that AstroTortilla has a polar alignment testing feature. I tried that a few days ago, however I was having problems with AT solving at the time. Has anyone tried using AT's polar alignment test? Is it informative/effective in helping you perfect your polar alignment?






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics