My objective for this session was to do a baseline guiding assessment of the G11. I have some work to do, but tonight was a break through night! I was able to get pretty good guiding and understand the data that I was getting and translate it into some actionable items.
I went outside to power up and polar align at 1820. It was about 51 degrees and there was no threat of dew.
With one exception, I was able to work through the basics to get everything connected and working with ease. This is the result of knocking some rust off over the last two times out.
My basic work flow in PHD tonight was to connect mount and camera, start looping, and then force a calibration. After the calibration completed, I noted the RMS, and stepped into a three-minute Guiding Assistant run. After accepting the results, I repeated the this preliminary cycle. After the third cycle, and seeing that my backlash wasn’t improving, I moved on to the final 30 minute run to produce the log file that is the object of this exercise. Here’s the result of my preliminary cycles:
Calibration: Advisory that high Dec backlash might result in degraded guiding
Guide RMS: Ra = .95”, Dec = .61” Tot = 1.13”
Guiding Assistant: Polar alignment error (PAE) fluctuated from 0 to .8’. Dec backlash was 1130ms (accepted/compensated in PHD). Even with this, I got the best backlash measurement graph that I have ever seen.
Calibration: Same advisory
Guide: RMS: Ra = .84”, Dec = 1.46, Tot = 1.66 (Dec and total got worse)
Guiding Assistant: PAE still very low, Dec backlash = 948 ms
Same calibration advisory
Guide RMS: Ra = .98, Dec = .59, Tot = 1.09
Guiding Assistant: Dec backlash = 935
Same calibration advisory
Similar Dec backlash
Thirty-minute Guiding Run
RMS: Ra 1.01, Dec 44, Tot 1.11 (consistent)
Ra dispersion was notably wider in the scatter graph
Ra pulse spike appeared. Each positive was followed by a negative. No consistency in timing.
After completing the 30-minute run, I attempted a shot of Flame/Horsehead Nebula.
Here the things that I have decided to do before the next baseline run:
Check the Ra Axis for stiction. It is working pretty well, but I am concerned about the spikes. I think that I can get the blocks aligned and the backlash set properly again.
Tear apart the Dec worm/blocks. Check for stiction while it is apart. Set align blocks and set backlash.
I examined the logs and I closely examined the image that I shot after the PHD baseline run. I can see that there is more work to be done on the mount, but since the test image is pretty darned good, I am going to set that work aside for a while to regain my imaging skills.
In the logs, I see two issues that need to be addressed. Backlash in the Dec axis, and stiction in both axes.
My recent work on the Ra axis has cleared up the backlash in that axis, but I haven’t worked on the Dec backlash since I have been closely studying mount issues over the past several weeks and now that I have a better understanding of how to set the worm blocks. Backlash, which PHD measures in milliseconds, is about 1000 ms, or 1 second in my Dec axis. I have read where PHD has been able to compensate for 2.5s, but that seems to invite guiding trouble. I would like for mine to be lower, and I think that I can achieve that now.
Stiction is where the mount resists torque from the drive system and doesn’t want to move with an ordinary guide pulse. The next pulse that comes along will be a big one, and the big one pushes the axis through the sticky point. I see that going on in both axes. It is a matter of getting in there and making sure that everything that is supposed to be lubed is lubed, everything that is supposed to be clean is clean and everything is assembled with washers, spacers and bearings in the correct order. Regarding correct assembly, there are some washers, spring washers, spacers and thrust bearings in each axis that have come from the factory in the wrong order. I now know the correct order, so I can check that and fix if necessary.
The only catch to checking stiction is that I have to remove the worm to check it. In the case of the Ra axis where I seem to have set the worm/blocks very well, I am reluctant to perturb that goodness just to chase better metrics.
At the end of the day, the baselining guide document says that what matters more than guiding metrics is the photo quality. That’s where the optics and sensor trump mechanics. I posted a single 180s monochrome image of Flame and Horsehead on flickr. I have round stars that do not appear to be bloated.
In my case, my sensor scale (a factor of focal length and pixel size) is 1.44”/px, and my total RMS error is about 1.1”. I see round stars in the image that I took last night, so I think that I am going to accept the Dec backlash and the stiction for the time being. This would never work with the Meade, but I can save further work on the G11 until it gets more comfortable outside.
Instead of further heroics with battling cold, windy, and cloudy weather just to get a wee amount of time to tune the mount, I am moving on to imaging. I’ll start with a star cluster while the moon is still around, moving on to dimmer stuff later.
As for an astronomy man conference, how about we set two conditions: a high above 50 degrees and a clear evening sky? I am thinking that a mid-afternoon start would give us time work some things out, and then we could set up and work with your mount afterwards. It doesn’t get dark enough to start working at the scope until 6:30 now, so that would give us time to grab a pizza, which we could eat outside. Or, we can do two separate sessions if you prefer.
I have been rethinking the 8-bit vs. 16 bit issue. It seems that a pixel value of 255 in 8-bits ought to be as bright as a pixel value of 65,535 in 16 bits. The only advantage of 16 bits is that you get a smoother transition as a star brightness decreases. I don’t know about it for sure, but hands on testing is required.