What's the real difference between decouplers and separators?

[Tex]

Okay, this is a question that I've been wondering about for a while.

There are two forms of staging parts (technically 3 if you count docking ports, but that's a different mechanic entirely), decouplers and separators.

Now, as I understand it, decouplers leave only two crafts when the rocket is staged: The pieces above and below the decoupler, the decoupler permanently attached to one side of the piece. Then there are separators, which leave three: The pieces above and below, but also the separator itself.

My question is: is there any reason to use seperators rather than decouplers? By all standards, separators are inferior to decouplers: Higher cost and mass, with the same ejection force (the one exception apparently being the TR-XL Stack Separator, with a much higher ejection force than its Rockomax equivalent).

Edited February 23, 2015 by Maximus97

[cantab]

Sometimes you want both parts of the craft. If you use decouplers one of them has the decoupler left stuck to it and you might not want that.

Also, even with nominally equal ejection forces, I've found separators to have more "punch" than decouplers. Probably because they apply that force to both sides rather than just one.

[Tex]

Hmm, I suppose that makes sense. I myself only ever used the TR-XL because it shaved some length off of my Apollo-esque CSMs!

[GreeningGalaxy]

I did some (rather unnecessarily rigorous) tests involving decouplers, from which I can tell you the following:

1) the unit for "decoupler force" is kilopond-seconds (kilogram-weight-seconds, or dekanewton-seconds if you prefer). Technically impulse, not force, but whatever.

2) (the useful information) Separators actually supply twice as much impulse as decouplers, because the given number for force is per side, not total. So the 2.5-m separator provides a total of 1200 kp*s, since each side provides 600 kp*s, while the decoupler of the same size only provides 250 (if I remember correctly) kp*s total.

3) Separators are more reliable. Decouplers are prone to that bug where they fire but refuse to actually detach, while I have seen no such problem with separators. (also, even if they stick on one side, chances are the other side will still work.)

I personally almost always use separators, with the occasional exception of 1.25-meter ships, for which the separators are bulkier. I've had too many instances of decouplers getting stuck to things and making my life hard.

[OhioBob]

Sometimes you want both parts of the craft. If you use decouplers one of them has the decoupler left stuck to it and you might not want that.

I agree that this is really the only reason to use separators. And even then it really only makes sense if both parts are to perform propulsive maneuvers after separation. In that case you don't want either part carrying needless mass. On the other hand, if one of the parts will not be performing any further maneuvers, then there is no harm in it retaining the mass of a used decoupler.

[Renegrade]

3) Separators are more reliable. Decouplers are prone to that bug where they fire but refuse to actually detach, while I have seen no such problem with separators. (also, even if they stick on one side, chances are the other side will still work.)

I haven't seen that in forever - but I'm careful to ensure that my decouplers attach themselves to the upper stage by their upper node (I suspect the problem happens if you attach them by their bottom node, then horrible clipping happens and even if the other part falls away, you end up with a decoupler stuck to your engine as it's top node is the one that lets go)...

Also, separators look cool (I think they look neat, slowly floating away, doing a slow spin), but they leave extra debris if they're not sub-orbital :/

[Xannari Ferrows]

Just as Cantab said, when launch 2 ships at the same time.

Separators are more reliable. Decouplers are prone to that bug where they fire but refuse to actually detach, while I have seen no such problem with separators. (also, even if they stick on one side, chances are the other side will still work.)

That's not a bug. Decouplers have arrows on them pointing to which side they will detach from. They will hold onto the other.

[IllicitMedic]

I use separators because of their low profile, but only on trajectories that will result in the separator impacting or burning up (such as after Munar transfer or Kerbin deorbit burn etc.) to minimize junk.

I also use them when launching multiple payloads on a single vehicle.

[Accelerando]

That's not a bug. Decouplers have arrows on them pointing to which side they will detach from. They will hold onto the other.

So it's impossible to comprehend the idea that both sides of a decoupler might sometimes keep holding on after you try to stage it? It is a bug, and I've experienced it myself on numerous occasions. Try using your imagination next time.

[Xannari Ferrows]

So it's impossible to comprehend the idea that both sides of a decoupler might sometimes keep holding on after you try to stage it? It is a bug, and I've experienced it myself on numerous occasions. Try using your imagination next time.

Staging is not always reliable. Try a manual separation.

[IllicitMedic]

So it's impossible to comprehend the idea that both sides of a decoupler might sometimes keep holding on after you try to stage it? It is a bug, and I've experienced it myself on numerous occasions. Try using your imagination next time.

Simple misunderstanding, why be such a d***?

[Renegrade]

So it's impossible to comprehend the idea that both sides of a decoupler might sometimes keep holding on after you try to stage it? It is a bug, and I've experienced it myself on numerous occasions. Try using your imagination next time.

Are you using part clipping? I haven't seen this issue since 0.22...and back then, things going wrong with my builds were more likely user error (on my part) than anything else.

Not to say that KSP isn't LOADED with bugs, of course. But I've done literally tens of thousands of launches and test flights and not seen that in over a year, so I'm kinda questioning the validity of this 'bug'.

If this IS a bug, and that can be proven, then we need:

- Logs

- Screenshots

- Version information

- Steps to reproduce / craft /etc.

and most importantly:

- a bug report.

(doesn't have to be from a specific person. If this is really a bug, it needs to be fixed ASAP - "1.0" is coming, afterall)

[Axelord FTW]

So it's impossible to comprehend the idea that both sides of a decoupler might sometimes keep holding on after you try to stage it? It is a bug, and I've experienced it myself on numerous occasions. Try using your imagination next time.

Because Ockham's razor. The simplest explanation to the problem was that, as a lot of people before you, didn't put the damn thing the right way.

Simple misunderstanding, why be such a d***?

Indeed, why, I ask you.

[GreeningGalaxy]

Jeez, everyone calm down.

I thought this bug was relatively well-known; it's been getting me since several versions ago.

To be clear, what happens is that the decoupler (attached properly, without any weird clipping) doesn't detach either side when fired through action groups, staging, or with the context menu. The ship stays fully connected, parts can still be selected on either side of the decoupler, and the decoupler is still clickable, but its context menu lacks the 'decouple' button and it can't be fired again. I see it most often with the 1.25m decoupler, and sometimes (but less often) with larger ones. In .24, the large NASAMission one seemed to get it very frequently, but I haven't seen it in a while- most of my troubles of that nature lately have been with docking ports.

Still, regardless of whether or not it's still around, I prefer to use separators. Better safe than sorry, right?

[Renegrade]

Still, regardless of whether or not it's still around, I prefer to use separators. Better safe than sorry, right?

Well, safe is definitely better than sorry

But.. I believe the separators are heavier and more expensive than their equivalent decoupler.. plus there's that space debris (depending on situation of course) issue.

I have heard about bugs with docking ports being derpy (I think Claw has a fix for that), and there definitely is a radial decoupler bug* (and Claw definitely has a fix for that). I generally don't make heavy use of the 3.75m parts (only occasionally for a bottom stage of something very heavy), so I can't comment on that decoupler (it has #lolmassless issues to begin with anyhow), but I've seriously not run into any problems with two stages sticking together in the 2.5m or 1.25m (or even 0.625m) sizes since 0.22.

I do make sure to carefully attach the TOP node to the bottom of the upper stage though, as it's so easy for it to click over to attaching it's bottom node to the stage above, as that 100% for sure results in a decoupler stuck to the stage above, which is almost always bad. (I don't consider that a bug at all though, more like a UI design .. issue) NB: the arrow will still be pointed in the correct direction if this happens, just the decoupler will be slightly higher than normal.

* - if this is still present in 1.0, I will become very angry with Squad, and I will write them a letter telling them how angry I am.

[Accelerando]

Because Ockham's razor. The simplest explanation to the problem was that, as a lot of people before you, didn't put the damn thing the right way.

Plainly denying that issue exists, which is what Xannari did, is not the same as saying that there might be a solution that someone hasn't tried before or that the issue may not be what they had thought. Assuming idiocy does nothing to help. I think this should be patently obvious.

Simple misunderstanding, why be such a d***?

Talking at someone as though they were a fool is not a neutral action or a simple misunderstanding, it is just unnecessary denial.

Are you using part clipping? I haven't seen this issue since 0.22...and back then, things going wrong with my builds were more likely user error (on my part) than anything else.

I'd say this was around .22 in my own experience, around the last time I seriously played KSP - I haven't been on much since then. Granted, several times it was when I used MechJeb on my builds, but other times it was not. I'm not claiming that I've rigorously tested this myself, however - just that I've encountered the problem myself before, and most importantly that it's a fairly simple concept to entertain that a part in a video game might remain stuck to another part; and that entertaining this concept need not be an obstacle to recognizing and/or solving the bug, as some here seem to believe.

At any rate, I agree with Greening's assessment - separators are definitely what you want if you need a little more kick when you toss off a stage. And the record agrees, too - without the TR-XL, we wouldn't have had our good friend Plaid.

Edited February 23, 2015 by Accelerando
pruning

[Red Iron Crown]

Locked at request of the OP.