I’ve hinted in previous blogs and articles that I thought plate-solving might replace encoders, at least in some installations. Encoders are great - they give a near instantaneous readout of position, but that position can be subject to many errors in practice.
PiFinder is a good example of a plater-solving finder scope. It works well, has its own library of objects and can feed SkySafari over wifi. It’s available as a kit and is gaining popularity with the ‘push-to’ telescope users. As I have described before, a breakthrough has been the Tetra and Cedar plate-solvers. Set up well, these can solve a 100ms exposure in around 10ms on a Raspberry Pi. The only limitation is ensuring the scope is still for the exposure. Celestron StarSense is another popular example.
Inspired by PiFinder success I decided to produce encoder-less variants of eFinder & ScopeDog. Having a Nexus DSC is still preferred, but these new variants offer a work around if your Nexus DSC is ‘having issues’, and a way of getting setup and observing ultra fast, or just not using a DSC at all.
‘eFinder Live’
My first application is eFinder Live. This is actually just an eFinder Lite with a gps module installed. If on power up if fails to find a Nexus DSC connected, it reverts to ‘Live’ mode. Here it continuously plate-solves (sensing if the scope is moving or not) and feeding SkySafari over wifi. Under typical skies it can cycle through an image capture and solve in less than a second. It could be faster but I am using a Pi Zero to keep the power consumption, size and cost down. The gps module is needed as normally I get location and datetime information from the Nexus DSC on startup. If you select a target on SkySafari, the eFinder display indicates how far to push the scope to reach the target.
Using it in practice, I initially really missed the responsiveness that encoders deliver, but after a while I got used to that and found the accuracy of pointing more than made up for it. Not having to install encoders and use a DSC is clearly an advantage for some, and a big cost saving.
‘ScopeDog Lite’
Spurred on by eFinder Live I decided to tackle ScopeDog. The principle here is to use the stepper motor drives to maintain a virtual encoder calculation of telescope position. The virtual encoders are ‘synced’ to real sky using plate-solving when needed. Like with eFinder Live, the sync is fast and is made a part of the goto sequence. Hence any conventional 2-star alignment process is not needed - instead of initially doing a 2-star alignment and trying to maintain that accuracy over a session, the alignment is done at the point of target observation. As with eFinder Live a wifi feed to SkySafari is maintained and this can be used to command goto’s and slew the scope. Comparing the various ScopeDog goto performances …
- ScopeDog GoTo: Scope slews to target position based on Nexus DSC readout. Will add any iterations necessary to get within the desired positional accuracy. These iterations are very fast, but ultimately depend on the accuracy of the Nexus DSC readout. In practice the error is often about 5-10 arc minutes.
- ScopeDog GoTo++: Scope slews to target position based on Nexus DSC readout. A plate-solve then accurately syncs the Nexus DSC to real sky and one or two more goto iterations are automatically performed. Overall this takes about 2 seconds longer than the standard ScopeDog GoTo, but the final positional accuracy is only really limited by the drive backlash. My own 18” gets to within about 15-30 arc seconds!
- ScopeDog Lite: Scope slews to target position based on virtual encoder estimate. A plate-solve then accurately syncs the virtual encoders to real sky and one or two more goto iterations are automatically performed. Time and accuracy are about the same as for ScopeDog GoTo+