Over the years I’ve tried various routes to move away from having a Windows computer sitting hooked up to my rig. Long USB cables, x86 Single Board Controllers, voodoo, but the setup I have now for my main scope seems to be optimal.
I have a NUC sitting in a fairly waterproof plastic crate next to the rig running Windows 10 Pro so I can use remote desktop from the house and run EQMOD and APT locally. This then runs from the mains.
Trouble is if I want to go somewhere (which to be honest doesn’t happen that much) I have to duplicate the whole setup on my laptop and inevitable something gets missed.
What I want to be able to use is a Raspberry Pi. Main advantages are that it’s cheap, small, light and will run off of 5V (so I can run it from a 12V battery if I want with a 12V to 5V USB charger).
I’ve tried in the past with things like INDIGO Sky and in theory it should all be doable. In practice I’ve found it to slow and I’ve had trouble with my AltairAstro cameras.
But now we have the Raspberry Pi 4 so I thought I’d have another go.
The plan is simple. Take a Raspberry PI 4, mount it on the tube to reduce the stupid number of cables I have running everywhere, plug cameras and mount into PI, turn it on and it just works 😉
I’ve downloaded AstroBerry and tried it with my GPCAM2 and GPCAM3 on my secondary rig (EQ3-2 + ES ED80APO) and it seems to work fairly well in doors. Looks like the SDK pushed out by AltairAstro has caught up enough to make these cameras stable.
So I bolted it all to my rig on Saturday night (with a bit of 3D printing) and tried it out. The only aim of the night was to see if I could align the mount with kStars (using my Mono GPCAM2 + guide scope) which took a while but that was mainly me not wanting to read the instructions and also it being fairly cloudy (I don’t like to waste good nights on learning new kit or software).
I also tried to image from the main GPCAM3 but had trouble getting an image I could focus on. After the event I worked out that I’d set the camera to 8 bit mode (not a huge problem) but somehow I hadn’t set the exposure time right so very short exposures, hence no light :-s
Some things I learnt from the first night:
- I need a better 3D printed bracket. The one I’d done rather quickly didn’t do the job and I had to tape the thing on towards the end.
- My GPCAM3 doesn’t like USB3 much. I’ve had this problem before. AltairAstro cameras can be very iffy if the USB power isn’t just right. (Note to self: try a USB3 powered hub at some point). Solution plug it in with a USB2 cable.
- With kstars you seem to need a fairly good alignment to start with (or you just keep restarting the polar alignment until you get it right). Not to much hassle.
So on the whole we can chalk night one up as a success.
Objective: Polar Align in a reasonable time, focus and image some actual data.
So polar alignment went a bit quicker this time since I had some idea of what I was doing.
Then onto the main event: Focusing. So this time I managed to get a reasonably bright image of a star for my Bahtinov mask and that went well. Centreing the star was easy enough, I just plate solved and synced then used the kstars map to point at Caph and it ended up in the centre of the FOV. Bit of a tweak of the focus (manual for now) and we were there.
OK, so technically I did manage to image some data. What I’d failed to do was set the image type up correctly, so instead of 16 bit RAW data I got 8 bit RGB which looks terrible. Anyway, 20 x 30s frames of terrible data captured.
Then disaster struck! My raspberry Pi 4 decided to literately blow up. Looks like the power regulators have gone (probably due to my dodgy power supply).
Oh well, and it was going so well :-s