What do you think of the new SAS?

[DisarmingBaton5]

As most of the people on this thread have suggested, the new ASAS is a bit soft. I've found myself overshooting quite a lot on the navball or reloading SAS entirely. I hope it gets patched soon.

[Tiberion]

Its softness is directly related to how much control authority you have on your rocket. If you are drifting or taking too long to correct things, you don't have enough control. Add more (reaction wheels, wings, RCS, gimbals)

Another issue some of you may be having is power - reaction wheels take power. If you run out, you lose control and the SAS can't make changes no matter how badly it wants to.

I promise you, the computer itself is working as it should because I have long had designs that very clearly show SAS issues when they are present.

[Brotoro]

I like that the new SAS responds to rover steering control inputs when changing heading lock. But it often does not realize the rover is on the ground on a low gravity world like the Mün, so it only responds intermittently there.

[DMagic]

Hmmm - so now I am a bit confused.. What exactly do I need or not need for rockets? SAS or ASAS?? I still attach the ASAS module, do I even need to since I read that it was incorporated in the command modules?? Can someone give me the lowdown on that?

I launched a few rockets with the ASAS module and so far have had no drift at all, although the rockets were not too complicated.

See my SAS tutorial: http://forum.kerbalspaceprogram.com/threads/41941-New-SAS-functionality-and-You%21-0-22-Update. The new system is fundamentally pretty similar to that used in 0.21.

[Kasuha]

SAS was working wrong in 0.21 and is still working wrong in 0.22. It applies "corrective force" proportional to deviation from target angle, but does not take angular speed into account. So if you are at zero deviation it applies no correction even though your angular speed is still a degree per second, meaning in the next moment your angular deviation will start growing again.

This kind of regulation always leads to oscillations (which may not be that bad if the ship is rigid enough) and to depletion of medium used for corrections (monopropellant or, in career mode initially very precious, electricity).

What should it apply is force proportional to both angular difference and angular velocity. Instead of A = m.ad it should be A = m.ad - n.av

[Superfluous J]

I've been going slowly through career mode so the largest ship I've made only went to Minmus and back with a 1-man capsule. Maybe 70/80 parts max.

However, if they hadn't told me about it, I wouldn't know it had changed.

[RocketPilot573]

I'm not really sure about it yet... It seems to be quite sensitive and even twitchy sometimes. Other times it 's like it just "forgets" and lets the rocket drift. And sometimes it simply freaks out (On! Off! On! Off! On! Off! On! Off! On! Off! On! Off!). As far as I can tell, this thing really needs to see a doctor.

[Nao]

The new SAS is an improvement, but for me it's way too slow. Most of the time i keep it off during burns.

It would be awesome if we could get a way to set how strong it tries to control the craft. Or at least have 2-3 parts that have different settings with them.

[Apophis775]

I haven't seen my ships shake, but if you mean the heading, then yes. It's been driving me crazy that it's not as "tight" as it used to be. I could BARELY dock before, I can't even imagine it now...