Having developed and improved the eFinder, it had reached a good point to leave it as is. Quite a few people around the world have or are building their own, and it seems to be a stable design now. So what next!
I had previosuly tinkered with the Raspberry PI HQ camera (cheap and small) but left it in favour of the ASI 120 and its very good sensitivity and low noise. But others have had success (eg PiFinder) so time to have another look. Also, the Pi Zero 2 W was worth a look as an alternative to the Raspberry Pi4 or 5. It has a quad core processor, GPIO pins camera interface and can run Debian 12 which I use in eFinder. Drawbacks are slower clock speed, very low RAM size (512MB) and only one USB port.
Along with the change in camera and processor, I decided to look at Tetra3 as an alternative to astrometry.net as the plate-solver.
PI HQ Camera
With its high resolution and small pixels, the Pi HQ camera will never compare with a dedicated astronomy camera. Its advantage of being lower cost, small, and having a c-mount lens thread make it ideal for this project. I made a test jig that held the Pi HQ camera, and ASI120 and a Pi5 to drive them and save images. I tried 25mm f1.2 and 50mm f1.8 cctv lenses and captured about 100 night sky images. Conditions weren't great being mid summer, but that's a good test anyway!
The test images have been used to evaluate and improve the plate-solve performance of astrometry.net vs Tetra3 vs Cedar-Detect. My conclusions are that the Pi HQ camera,with a 25mm f1.2 lens and 4x4 binning, produces adequate images, often in just 0.1 second exposures.
Tetra3, Cedar-detect & Cedar-solve
With the camera test images, I was then able to try out Tetra3. A learning curve! At first I got no solves. But perseverence and experimentation paid off. Tetra3 uses a database that is matched to your camera field of view and sensitivity. Get that right and then you can start to optimise the solve process itself. My best results so far on a Pi5 are solves in about 10ms, which is amazing. I dont think it will be as bullet proof as astrometry.net, but probably good enough. It doesn't have all the features of astrometry.net and in particular my semi-automatic finder to main scope alignment routine needed major revision - but it still works.
I saw that PiFinder is using Cedar-Detect & Solve to extract the stars and solve from the camera image. These are extensions of Tetra3 by a new author. They are a lot faster than the original Tetra3 and works well on shorter exposures. You do have to make sure star images are reasonably well focussed and dont saturate in the camera though. I have now switched to Cedar-Detect for centroid extraction & solve.
eFinder ‘Live’
A recent update is to add a gps module to eFinder Lite. If the eFinder Lite is plugged into an ordinary 5V usb supply (ie not a Nexus DSC) then eFinder ‘Live’ will start instead. This uses continuous plate-solving to determine telescope position and relay it over wifi to SkySafari or similar.
Pi Zero 2W
This is basically a cut down Raspberry Pi. About 1/3 the size and price, and probably about 1/3 the max performance. Biggest issue is only 512Mb of RAM, which quickly gets used up and can lead to crashes and slow performance. I found a neat Waveshare USB and UART HAT that simplifies the build, and then the OLED display can just be plugged directly on top. The whole eFinder Lite draws about 200mA which the Nexus DSC is happy to supply down the same USB cable used to communicate with the eFinder Lite. Slimming down the OS, the eFinder code and using Tetra3 & Cedar-Detect the performance of the Pi Zero turns out to be OK. Most solves are less than 0.3 seconds!
Build
The Pi Zero, USB/UART HAT and OLED make a neat ‘stack’. 6 wires are soldered from the Pi Zero GPIO pin underside to the 5 way navigation switch. A adxl345 accelerometer is wired to the Pi Zero i2c pins (4 wires). The adxl345 is used to see which way the eFinder Lite is mounted and if needed flip the display and buttons ‘upside down’. So the eFinder Lite can be mounted on either side of a scope. A custom case was designed and 3D printed.
Click here for build instructions