Important for anybody who is using a space station

[ArmoredLipid]

If you're new to space stations, there is one incredibly important thing to remember (wow, that sounded like clickbait). If you have a docking port on your station, that you intend to dock other ships to, you must select EVERY reaction wheel system on BOTH the docking ship AND the station, and set them to "pilot only". On the downside, this will make your station rather slow to maneuver and change attitude. The upside is that you do not trigger an infernal feedback loop/kraken attack that will result in your ship being rammed through your station at a 180-degree angle at a sizable fraction of c. 

The reason this feedback loop happens is something along the lines of the following:

1. Docked ship wobbles - just a tiny bit.

2. Reaction wheel notes wobble, attempts to correct it with counterbalancing force.

3. Reaction wheel inevitably goes a millinewton or so of torque too far and pushes the ship in the other direction.

4. Reaction wheel attempts to correct this new wobble, and goes too far.

5. Repeat until the docked ship has been turned into a relativistic projectile and/or battering ram.

[Superfluous J]

29 minutes ago, ArmoredLipid said:

wow, that sounded like clickbait

5 things <city you live in> residents didn't know about space stations, and the adults couldn't stop staring.

I personally recommend turning off SAS entirely and leaving it off for good. Luckily whenever things dock together, SAS (and RCS?) turns off automatically so it'll never accidentally turn back on.

[ArmoredLipid]

2 hours ago, 5thHorseman said:

I personally recommend turning off SAS entirely and leaving it off for good. Luckily whenever things dock together, SAS (and RCS?) turns off automatically so it'll never accidentally turn back on.

Turning SAS off completely is generally a bad idea. It leaves your station unable to orient in a required direction (i.e. docking alignment) without using bulky monoprop. Additionally, since monoprop is mass leaving the ship, it will destroy precise orbits over time. My advice is to only use monoprop for altitude control thrusters (the Place-Anywhere 7 Linear RCS Port is your best bet here). SAS does not expel mass and can have the (also massless, although batteries are a bit heavy) resource used to activate it (electricity) supplied by a number of means other than shipping it to the station, such as solar panels, radisotope thermoelectric generators, your altitude correction thrusters firing, and fuel cells (listed in order of practicality).

In mass terms, electricity is a one-time cost (if an expensive one), while everything else is lighter, but the gift that keeps on ungiving.

Edited June 21, 2018 by ArmoredLipid

[Ultimate Steve]

1 hour ago, 5thHorseman said:

5 things <city you live in> residents didn't know about space stations, and the adults couldn't stop staring.

# 4 will shock you!

[Superfluous J]

1 hour ago, ArmoredLipid said:

Turning SAS off completely is generally a bad idea. It leaves your station unable to orient in a required direction

Nothing a combination of time warp and a docking alignment indicator mod can't fix.

And don't tell me time warp is cheating. I consider it a bug fix for having to leave SAS off

[ArmoredLipid]

9 minutes ago, 5thHorseman said:

Nothing a combination of time warp and a docking alignment indicator mod can't fix.

And don't tell me time warp is cheating. I consider it a bug fix for having to leave SAS off

SAS is faster unless you're using the built-in command pod reaction wheels to do so. I personally take the Advanced Reaction Wheel Module, Large, stick 6 on a Rockomax HubMax Multi-Point Connector (the big one, not the industrial espionage-produced Dinkelstein knockoff), set them to Pilot Only, and then stick another one on the end of each module I attach. That's 60 kilonewtons of force a second on each module, which is enough to rotate most stations rather quickly.

Edited June 21, 2018 by ArmoredLipid

[LordFerret]

You will find, as others have found who have been playing KSP for years (myself included), that it is best to leave SAS OFF on space stations... which is why we're telling you this.

There really isn't any great need or requirement (that I know of) that a space station be reoriented, so perhaps that should be looked at as more of a personal play preference. Typically, the docking vessel is far smaller than a completed station and can easily fly about it and orient for docking... I do it all the time. Most of my stations are built of modules which have no SAS units incorporated into them at all, or if they do they're turned off once docked (as in disabled, so no 'accidents'). And the mono-propellant my stations carry in abundance is only for use with tugs and replenishment of visiting vessels. Being I dock visiting vessels at 0.1m/s, there's no resulting change in the station's orbit. And proper planning / assembly of solar panel modules insures power generation without regard to station orientation. I think if you check out the stations built by others, you'll see how most approach this build aspect. There's also several (if not many) posts by people seeking help with their exploding stations; The result of nearly all being SAS having been left on.

Check it out. Good luck.

Edited June 21, 2018 by LordFerret

[qzgy]

4 hours ago, ArmoredLipid said:

Turning SAS off completely is generally a bad idea. It leaves your station unable to orient in a required direction (i.e. docking alignment) without using bulky monoprop

..... what? I'm not sure why you say that. Even if you have SAS off you can still rotate the station about since the reaction wheels (SAS Module part thingies that you so helpfully tell us to set to pilot only) still work. The self stabilizing autopilot is off though thats all.

[The Aziz]

Never had any problems. My stations have set the MJ SmartA.S.S to  or , whatever is more suitable for the design (so it acts more or less like the ISS) and autostruts on. Perfectly stable.

[Hotaru]

I've never had trouble with SAS on a station, at least not in the age of autostruts. I generally keep mine oriented / and then set SAS on my arriving ship the opposite way, that way they automatically stay aligned. Even with a docking port alignment mod that makes life a lot easier.

And I'm pretty sure SAS automatically goes into attitude hold mode when two ships dock; I've never seen a ship try to reorient itself after docking that I can recall. As far as spontaneous kraken-induced disassembly goes, don't know if it's due to my habit of autostrut abuse or tendency to build relatively compact stations, but I've yet to have that happen, SAS or no.

[katateochi]

4 hours ago, The Aziz said:

Never had any problems. My stations have set the MJ SmartA.S.S to  or , whatever is more suitable for the design (so it acts more or less like the ISS) and autostruts on. Perfectly stable.

I use the MJ SmartASS on stations too (on everything, I basically fly by SmartASS, best. tool. ever.). The really nice thing about it is that as you approach the station in another ship, when you get in range the station will automatically reorients itself to the desired heading without you having to switch to it. (the horrible thing about it is when you undock and a previous SmartASS setting on the undocking craft kicks back in and it suddenly spins round and wallops the station).

Another way of reducing SAS induced station death-wiggle is to auto-strut parts across the docking port.

I think also the design of the station's SAS is important. I always put my station SAS close to the centre (COM) of the station and make sure all the SAS modules are oriented the same way. When you bring new modules up to the station, don't have probecores (with their own SAS) and SAS modules on each station module, have each module as inert (uncontrollable) sections and use a docking tug to push them into place. That way you can keep a single central station control module and not find you've accidentally added more SAS to the extremities as you add more modules.

[ArmoredLipid]

I guess my issue is that I generally put stations in stationary orbits, to provide a known launch window at certain times. I just don't want to use RCS to accidentally modify it.