Satellites changing position in their orbit

[charlesjto]

Am I doing something wrong?

For some strange reason, without me taking control of the satellite, it has changed position in its geostationary orbit.

Before:

After:

Is this a bug of some sort? Its the second time it has happened and is rather annoying. Mods installed: KIS, KAS, remotetech, mechjeb, kerbal planetary bases.

[stibbons]

No matter how careful you are, it's impossible to get all of your satellites in precisely identical orbits. Because of that, it's inevitable that their position relative to each other will drift over time.

In the real world, this is handled with periodic orbital adjustments, speeding up or slowing down a little to maintain position. In KSP I prefer to cheat a tad. Once I've got my satellites as close to their final positions as I can reasonably achieve, I use Hyperedit to tweak each satellite's orbit so they all have the same semi-major axis. This means that they will indeed maintain their relative position without further adjustments. The downside to this approach is that the SMA will drift slightly whenever a satellite is made the active vessel, and will need to be reset, but that's a form of station keeping that I'm much happier with than constantly tending to my satellite network.

[charlesjto]

I do understand that satellites will eventually drift, the orbital period (according to mechjeb) was within ~30 secconds of 1 day on each of them. Only 23 days had passed kerbal-time.

Further more, I just re-loaded my save after being out for a few hours, and miraculously the orbits are back to forming a nice triangle.

Note the time in top left, no save editing was done (as the last time I tried it deleted 2 of my satellites):

[Stone Blue]

10 hours ago, charlesjto said:

I do understand that satellites will eventually drift, the orbital period (according to mechjeb) was within ~30 secconds of 1 day on each of them. Only 23 days had passed kerbal-time.

With KEO, even 30 secs error is A LOT... I recently saw a chart of different (small) amounts of orbital period error, and the resulting length of game time, before the constellation was unusable due to orbital drift... Wish I could find it...

Use RCS and or a thrust limited ANT engine to get < 1 sec error... Also, once you get it ~1sec, if you rotate the craft, even just using reaction wheels, you can throw off the period by as much as a couple of secs or more... So if you want a satelitte to orient in a specific direction (this is mainly for aesthetics, since physically pointing dishes does nothing, like in RL), for example to point a dish or scanner always at Kerbin, or to always point at the sun, its best to orient it in that attitude, THEN use RCS/ANT for fine period adjustments (this means taking this into consideration during satellite design & construction), and when you get < 1 sec, just leave attitude and orientation alone.
(also, the Persistent Rotation mod helps with that)...

EDIT: Actually, if you havent seen this, you should take a look (this isnt the "chart" I remember seeing, but it's listed on the RT tute, so it must be legit:

https://remotetechnologiesgroup.github.io/RemoteTech/tutorials/keo/#optional-steps

I dont know if its a linear equation for those calculations, but if it were, then 30 sec error would only last 40 Kerbin DAYS... And thats probably if all were in perfect placement to begin with, meaning even less usable time with compounded errors of actual placement of satelittes...

And yes, the satelittes will move back into alignment again and again... In your case, if you dont mind warping all your other craft/missions, till they align again, once they are, THEN you can possibly attempt to make adjustments to the periods if you still have a means of maneauvering the sats...

Edited March 12, 2016 by Stone Blue

[charlesjto]

The 2nd picture in my original post is only ~20 seconds game-time before the following picture, I find it highly unlikely it could have drifted that far in 20 seconds.

In regards to using the ANT engine, my satellites were equipped with Dawn electrical  engines, which as far as I know have the lowest thrust.

Thank you though Stone Blue for showing me that chart of orbit tolerances. Although 30 seconds was just an estimate as I could not remember off the top of my head, now having checked: sat 1 is 0.2 seconds off, sat 3 is 0.07 seconds off, sat 2 is 1.7 seconds off (guess ill fix that one now).

Edited March 12, 2016 by charlesjto