GEMs are the worst choice for tracking satellites.? It's not the equatorial mounting, per se, but rather, the need for a meridian flip with GEMs.? A fork mounted EQ mount on a wedge, for example, has no need for a meridian flip.? When polar aligned, as most EQ mounts are, a meridian flip occurs ... near meridian.? Meridian has nothing to do with a general satellite track, particularly not those that show details in our scopes.? They culminate where they culminate, and that's rarely meridian.? It's not the better build either, but rather, the ability to get a good alignment because of that better build, much more easily.
?
For OP, use the "Simple Track Improvements Test" that Gary recommended, and round down rather than rounding up.? That way, it will be slower and lagging, so all your corrections will be to "speed up", while maintaining the proper direction.? The alternative, rounding up, will have you sending command to "slow down".? Rarely, except around culmination, will simply "slowing down" switch directions (it's moving too fast for that, except in alt around culmination), but there are practical benefits of keeping the satellite slightly behind the center of the fov.? That benefit is reacquisition should it slip out of the fov.? When it's slightly behind center, there is typically more room in the eyepiece when it does drift out, so you are better able to know where it went.
?
Back to the "better alignment" issue, AN + SC has some odd features that make that more cumbersome that it need to be.? Typically, satellites that show details are low, so they have passes shortly after sunset, or shortly before sunrise.? If the pass is before sunrise, there is plenty of time to gather a large run of data to run TPAS, resolve issues, get comfortable, ... .? When the pass is shortly after sunset, there is hardly any time at all to run a new TPAS model.? Unfortunately, ArgoNavis will JUNK your alignment any time you update a TLE, so you have to do another TPAS run to get those non-constant terms.? Because ArgoNavis JUNKS the alignment, it is best to avoid having to rely on TPAS at all.? That is unfortunate, and no fix to this bug is likely coming soon.? Because TLEs need to be updated often, it is even better to not rely on ArgoNavis at all.
?
SkyTrack is seemingly an alternative, but ServoCAT is not ready for that yet.? SkyTrack loads TLEs into its own software, and has the ability to plate solve along the projected path to make its own model.? It works incredibly well with my other commercial mounts from Celestron and SkyWatcher.? Those are not incredibly well built mounts, but the alignment is what overcomes that.? One can build an excellent alignment using those mounts respective control software, through plate solving, without user intervention.? One can then further refine those models using SkyTrack itself, through plate solving, without user intervention.? SkyTrack then has the ability to fine tune the tracking rates with a gamepad or the like, adjusting the corrections based on the present tracking speeds.
?
What keeps SkyTrack from working with ServoCAT at the moment ServoCAT's inability to accept custom rates through ServoCAT's USB port.? One is limited to the preprogrammed rates of the handcontroller.? Because the satellites' rates change along the pass, that is now a limitation of ServoCAT itself.? ArgoNavis need not be involved anywhere, and any other DSC (or none at all) would do otherwise fine, preferably one that does not junk your alignment when one you update a TLE.? As soon as ServoCAT's USB port is "up to snuff", SkyTrack will be my preferred solution for satellite tracking with my dob.
?
For OP, for now, ArgoNavis is the only workable solution.? But again, it JUNKS your alignment whenever you update a TLE, which is all the time.? In practice, that means getting by with as few stars as your are able to do so, when the pass is shortly after sunset, to get the best alignment you can.? Adjust your mount itself to make those non-perpendicularity terms as close to zero you can get them, use classical bi-directional offset so that your mirror points where your mount points, and level your mount as best as you can.? Structural rigidity will keep flexure terms at bay, and they are the hardest to deal with.? Once you can get as best an alignment as you can, with as few stars as are available, then, and only then, will AutoLock work properly.? Getting the ratios adjusted in ServoCAT as best as you can at that point simply means that the tracking rate adjustments because of AutoLock are now minimal.
?
Bear in mind that you will still need to make tracking corrections.? TLEs are never exact, and their utility decays as one gets farther from their epoch.? Your time and location need to be as precise and accurate as you can get them.? Your alignment needs to be as precise and accurate as you can get it.? Even then, you cannot account for variables such as refraction, without actually modelling them first.? All of your efforts will lead you to having to make smaller corrections to the tracking speed.
?
Even with all of these challenges with the present ArgoNavis + ServoCAT, I can still keep ISS and CSS in the FOV of my 15" with a 6 Ethos, yielding about 300x in a 100 degree AFOV.? The TFOV is about twenty arcminutes from one side to other, so the "available room" for error is about half of that, at only about ten arcminutes.? That's rather good!? But, there is always room for improvement, and a better protocol of ServoCAT's USB port is the way forward, for things such as updating TLEs, doing alignments through plate solving (which can align closer to sunset than my eyeball), getting the most precise and accurate times and locations as one can, and having the ability to make better models on the fly.
?
With respect to the zenith problem of an alt/az mount, I can simply tip my dob to change the mount's zenith with respect to the sky's zenith.? That avoids a pass straight through the pole of the mount.? With a computer, making the adjustments for having "changed" one's latitude and longitude because of that tipping is rather straightforward.? It's much harder with a fixed single purpose computer that expects you to enter those values yourself.? Clearly, with any sized telescope, there is a limit to how far your structure can handle such tipping (the az bearing is the weakest point), but that limit increases rather quickly with larger sized scopes.? The solution that specifically designed satellite trackers employ is going to alt-alt-az, rather than simply alt-az, or a full gimbal.? That's not an option for us, so tipping in the poor man's solution to getting alt-alt-az, then simply telling the alignment software our "new" latitude and longitude (preferably through plate solving at the mount's pole).
