The computery side of my observatory is going to be Linux-based, as far as possible, so I’ve had it in the back of my head that I’d be running something like Kstars and Ekos to control the mount and other kit. I discovered that Ekos has a polar alignment system based on plate-solving and thought it would be worth a try to see if I could do better than I’d managed with PHD2.
So, back to the observatory having installed Kstars, all the necessary bits of INDI (from the Launchpad PPA), astrometry.net and the astrometry index files on the laptop (the latter taking several hours over my lightning fast ADSL connection). The plate-solving can be done over the net, but the network connection to the observatory isn’t great and it’s much faster to do locally.
Ekos takes three images near the pole, calculates the centre of rotation of the mount axis and how it needs to move to coincide with the celestial pole and then overlays a line on the camera image (which can be moved to line up with a handy star) to show what movement is required.
As with PHD I found this a bit awkward when both ends of the line don’t fit on the image, but as long as I got one end visible and adjusted the mount to move a star as far as possible in the right direction, it was just a case of repeating the process to narrow down the error. It took me a while to get the hang of it, but once I’d got my head around which directions the adjusters moved the star in any given direction it became easier.
Using this process I got the polar alignment error measurement down to less than one minute of arc and thus encouraged moved the camera over to the TS Optics Photoline 72 on the second NEQ6, achieving a similar degree of error quite quickly.
My third mount (a HEQ5) however doesn’t have a view of the polar region so this method wasn’t going to be possible. I decided to leave it for the time being and consider what other options I might have.