The Starry Night, 67

:: home ::




              <<   41   42   43   44   45   46   47   48   49   50   51   52   53   54   55   56   57   58   59   60c   60   61   62
              63   64   65   66   67   68   69   70   71   72   73   74   75   76   77   78   79   80   81   82   83   84   85   >>   SRCH


10/23/2011. A night of mixed success. I got a good, sharp focus (in the 1.5 arc second range) and unguided drift is reasonably well controlled, so both seeing and polar alignment are on the better side of fair. The first 900s sub of NGC 6960 was very promising. But the next three were garbage. PHD had somehow misplaced the guidestar. I suspect the OAG slipped. I refined its focus and tightened everything down yet again. A lot of play, but I resumed guiding. The next several images of the Veil were excellent and will find their way into the data pile soon enough. An overnight run on M31 was also excellent. See:



M31, Hubble's V1, dust lanes, starclouds
18x900s L


That wouldn't be bad as a wideangle view of starclouds, dark lanes, and star clusters in the summer Milky Way, but it's actually a photograph of the very same things in a galaxy 2.7 million light-years away. When things work, they work well.

It occurs to me that by the time I have a good record of V1 in all its phases, I'll also have ~50 hours of exposure on this field. If 4.5 hours looks like that, who knows what ten times the data might show.

Achieved PSFs are a bit over 2 arc seconds in this 4h30m exposure. Lots of good detail, and yet... I am not happy with this hit and miss experience. You don't buy the good stuff for it to work right "more often than not" or "most of the time." (Arguably, I did cheap out on the off-axis guider, so let's not press this line too hard.)

Overnight, I dwelt on bracketry to freeze the OAG guide-camera head in place. I finally imagined something that might work, but was pretty sure that it would take all day in the shop guessing at angles and lengths. And then it might preclude some adjustments, or demand even more fine tuning. So I settled for a set of pre-tensioners in each of three axes. Rubber bands. Stretched around the guide camera and fixed to the OAG or to the imaging camera. They'll do for now.

Guide corrections were (as usual) slightly wild, constant, always excessive. Polar alignment is not perfect, seeing is not perfect, etc., but something is making the mount work too hard and spreading out the light in the process. So I finally did some research on PHD software parms and how they might be adjusted when long focal lengths are used for both guiding and imaging. I reset several. If they make a difference, I'll post the settings below in case they're generally useful or in case I forget and misplace my notes (as seems inevitable).

Later that evening: the guide parm changes were a great success, but the OAG is falling apart on me again. First, here are the PHD parameters I adjusted to accomodate a longer than anticipated guiding focal length: 350ms calibration steps; 2x2 mean "noise suppression"; Dec minimum move 0.8 pixels. I also tried both lowpass dec corection and "resist switch" (the default). Both "auto" and directionally selective dec guiding worked well.

But that OAG. The rubber bands were an abject failure. They held it nice and steady, but they held it nice and steady far out of alignment. In fact, so little was square that I never got stars that were not streaks of light. The compression of the rubber bands made focusing almost impossible. I fiddled with the OAG in the dark until it developed whole new degrees of freedom. About midnight, I brought it and the CCD inside, took it all apart one more time, made sure I understood where the slack was developing, what moved, should move, and shouldn't move and then I followed some bold Internet advice.

Cyanoacrylate is your friend. I Superglued a few attachment points where slop seemed unavoidable. Maintenance of a few aspects of this OAG is now impossible (the pivot point for the mirror is utterly inaccessible, for example), but it does not move in ways it should not move. And a quick look under increasingly hazy stars showed me the tightest, brightest off-axs guide stars yet. It feels solid and focus should be much easier when there is little or no danger of tilting the guide camera, the mirror stalk, etc, etc. The wx guys are saying tomorrow night's skies will be good; I'm saying the hardware will be ready, too. Finally!


10/24-25. Good seeing, excellent tracking, but Robofocus balked. I used the momentary contact buttons to get close to best focus, but it won't talk to the computer. Next day, I reverted some Windows updates and Robofocus is happy again. Make a restore point and don't lose it. Figure this out later. The next night, all was good, except there were sudden RA excursions of unknown cause. I got one on the PHD log (if I can find where it's stored) so we'll see if there are any clues. I saw two, both at similar Alt/Az, so... is the weight shifting from east to west at that point and shifting the weight-bearing point from one face of the RA gear teeth to the other? That's my best guess for the moment. Look for slack in the RA train and correct it by reweighting the outfit for the moment. Then, for some reason, Nebulosity hung up 56% of the way through a 900s sub and went to the 2000ms delay between frames. Where it spent the rest of the night. At least that's what the screen reported when I found it dead this morning (and no new images in the directory). It's not always something, but it's too often something. That's progress for you.


10/29. I have no clue why Nebulosity froze up once, but it did not do so the next night. To address the sudden RA excursions, I moved the counterweight stack a few inches further out the bar to make sure the rig is always out of balance in one direction. That should keep the drive pressure always on the same side of the RA teeth. I've increased guiding exposures to 3 seconds (from 2) and added a 1 second rest period between corrections just in case there was some settling time to consider. These mods have elicited the best tracking yet: stars are round to within a few percent, and the tattletale plot within PHD guide says that all corrections and deviations are an arc second or smaller. It makes a difference on the chip, too: I'm getting decent images of V1 in single 900s sub exposures now; combining only two produces a pretty fair image.

I set up for photographs of the Iris Nebula, NGC 7023, and started an RGB sequence during the last of twilight. Tonight is our once-every-other-year Halloween party, so I also set up a casual studio on the porch to photograph guests. I kept ducking into my office to tend the astrograph. I finished the night with a 4 hour run on M31 and V1. This time, PHD Guide threw an exception near the end of the sequence. No idea the cause, but as long as it is this rare, I won't worry too much about it.

The party photos are here.

And the Iris is below:



NGC 7023, the Iris Nebla
15x900s Luminance, 1x900s RGB
Astro-Tech 10-inch RC
Astro-Physics Mach1GTO
SBIG ST2000XM w/Baader LRGB filters


I installed an 18-inch USB extender cable on the Meade DSI. It's sold as a way to move the USB socket on a printer (for example) to a panel mounting plate. In my case, I can attach this cable as solidly and stubbornly to the DSI guide camera's loose socket as I like (shims, cable ties...) and then simply plug and unplug the wire to the computer from the other end. The big deal is that this lets me throw a cover over the telescope and mount but bring the CCD inside during inclement weather. Without some such arrangement, there was no way to remove the imaging package from the telescope without messing up the persnicketty mechanical and electrical connections to the guide camera. (Thiis might also address the mysterious exception from last night since momentary loss of contact with the DSI has produced some flakey behavior before.)



:: top ::



                   © 2011, David Cortner