2009/07/22 Total Solar Eclipse

Despite the poor weather prospects, this eclipse with its ~6 minutes of totality was too tempting to ignore. If it just happened to be clear (unikely - it's in monsoon season), then 6 minutes would be a significant boost to my personal tally of totality time - plus I'd always wanted to visit China.

I booked into a tour organised by my old club, the Sutherland Astronomical Society (SASi). I'd been on a tour they'd organised for the 2002 eclipse in Ceduna but despite that I decided to risk it again. This tour turned out to be with an even larger group of people than Ceduna, some 86 in total in 2 busses, and provided many more possibilities for disaster. However, most of us survived and had a good time.


I spent many months prior to the eclipse deciding on what equipment to take. It had to be light and portable, yet good enough to provide reasonable results. In the end I went with a completely new rig.

The telescope was to be my William Optics SD66 refractor (I wanted to use my William Optics Megrez 90 but decided it would be too heavy to lug around China after the event). The SD66 has very good optics and with a focal length of 400mm and operating at f/6 it seemed a pretty good match to my Canon 350D DSLR. I also wanted to look at the eclipse so included a flip mirror in front of the camera. This would allow me to flip the mirror up and look with a 20mm Erfle (20×, 3° field) and flip the mirror down and image (2·1° × 3·2° field). With a little work I got the eyepiece parfocal and concentric with the camera. Perfect.

The mount was going to be a bit of a risk, but lots of testing convinced me that it would work. My friend Colin had already bought one and showed me its potential. It was a Tasco Starguide single-arm fork with computer-control, complete with 60mm refractor and bought for the princely sum of $150. Intended as a beginners telescope and made rather cheaply, but it actually seemed to work. The adjustable light-weight aluminium tripod was actually quite sturdy (if not raised too high) and the computer control even seemed to do the right thing.

I managed to buy one, also for $150, and started testing. I first had to modify the housing to allow my SD66 to attach in place of the Tasco 60mm f/11 telescope and remain in balance. That done, on-sky testing showed a number of bugs in the software (it was an early version I had) which were also document on the web when I looked. They weren't a major problem, but newer mounts with improved software would have made the job easier. The mount would indeed point and track on the Sun (and anything else you called up) but you had to get the alignment right. Not a problem at night, but during the day it was as you couldn't use the Sun as an alignment star (one of the things fixed in later versions). The mount would run for ~4 hours on a set of 8 AA-size NiMH batteries (more than enough) and would track the Sun adequately for that time. However, its short-term tracking was dreadful (as expected from the crude gearing used). It would take 1 second exposures with no trailing, but any longer produced trailed images. This, however, was good enough and better than a non-tracking tripod, plus it had slow motions for centering. Definitely a step up from a tripod.

The mount and telescope during testing. The slip-on solar filter is made from an old pill bottle (courtesy of my brother) and its lid (sitting down on the tripod) has been made into a Hartmann mask to aid in focusing.

The final part to make it all work is some free software called Eclipse Orchestrator (EO). This fantastic piece of software is designed to control cameras during a solar eclipse, thus freeing you up to enjoy the eclipse instead of spending your time looking through a camera. (There's now a "pro" version of the software that costs actual money, but I'd already been using the free version and didn't see the point in upgrading as I didn't need the additional features offered.) I ran this on a new netbook (a Toshiba NB100) that came "free" when I bought my new Toshiba HD TV last year.

Eclipse Orchestrator is driven by a script which you have to write. I spent ages writing and testing it until I was happy that it was going to do what I wanted, which included voice reminders to flip the mirror back to the camera before it started taking pictures again. It copes with clouds by having an emergency script which can be started instantly by one keypress. I was feeling like this one was going to work - if the weather cooperated...


...which it didn't. Waking up on eclipse morning showed that the weather had performed exactly as predicted - heavy cloud and rain. I was not looking forward to this - even though I came on the trip expecting to see nothing. We proceded to our planned eclipse viewing site - the Hua Zhuang Agriculture Farm, about 15 kilometers from Tongxiang (120°31'16.85"E, 30°43'36.46"N). We stood around under cover, listening to the rain come and go.

The only safe place for a telescope is under cover. At least it was cool when it was raining. I didn't even bother to set up my 'scope, although I did have the netbook running showing how the eclipse was progressing above the clouds. Eventually I did set it up - just in case - but as soon as I did, it rained again and I had to put it back under cover.

And then a miracle about 10 minutes before second contact. In the reverse of what happened to me in 1991 in Hawaii, the clouds thinned around the Sun. The rain clouds parted a little leaving the sun visible through a slightly thinner layer of clouds. Amid joyous cheers from those around, I quickly got the telescope going. After a quick "alignment" I told it to point at the Sun. It did. Things were looking good.

There was no time to bring the netbook outside (and sit it in a puddle) so I went fully manual, all designs on automation gone. I could only focus through the viewfinder - not very reliable, but all I could do with only a few minutes to go. And then just start snapping away. (I did bring the cable release just for this possibility.)

Below are a selection from the ~80 images I took. Most images are the full frame resampled for web viewing, but 2 around 2nd contact are cropped and shown at half-scale. The times shown are from the camera and are about 20 seconds slow. According to EO, 2nd contact should have occurred at 01:34:35.6 UT, and 3rd contact at 01:40:26.3 UT. Due to the cloud, my guesses for the exposure times around 2nd and 3rd contact are a bit short and so the images are a bit underexposed. Oh well.

Image # Time (UT) Shutter ISO Comments
7279 01:28:35 1/4000 400 partial phase - no filter (just lots of clouds)
7282 01:30:35 1/4000 400
7288 01:33:31 1/4000 400
7294 01:34:14 1/1000 400
7295 01:34:15 1/500 400 2nd contact-ish - larger scale than previous images
7296-7300 01:34:19 1/250 400 combine of 5 images taken 11:34:16 to 11:34:21
7305 01:34:40 1/8 400
7306 01:34:43 1/4 400
7309-11 01:35:13 1 400 Composite, plus radial filter to enhace coronal detail
7312 01:35:31 1 800 There's a lot of cloud there!
7326 01:38:26 1/13 800 less clouds than the last image
7327 01:38:32 1/4 800
7330 01:38:55 1/15 800
7342 01:40:04 1/2000 800 near 3rd contact, chromosphere
7344 01:40:08 1/1000 800
7345 01:40:10 1/1000 800 diamond ring
7346 01:40:11 1/1000 800
7352 01:41:14 1/500 800 filter ON
7355 01:43:18 1/160 800 and then the clouds came back


Given the poor conditions I am amazed that the images turned out as well as they did. The mount proved to work well; I only touched up the pointing once during totailty. Lugging a full equatorial mount around would not have been possible, so this seems to be the next best thing. Ditto the SD66. Bigger would have been better, but it was good enough. Having the flip mirror meant that I (and a few other people) got to see the eclipse through a good telescope, plus I got the images I wanted. In all, a reasonably successful expedition.

Christine, our guide, looking at the partial phase through my telescope, just after third contact and before the next burst of rain.


Page last updated 2009/08/15
Steven Lee