Having Problems controlling a ship with the Cupola pod.

[AndrewBCrisp]

Hello all,

With the advent of .20 (and later .20.1) I figured I'd start over with a fresh save, and started to launch components for a space station. Inspired by Temstar and his family of launch vehicles, I took some time to design and test a couple of launch vehicles of my own and came up with the Sigma I, pictured below with its dummy payload:

The Sigma-I has a maximum lift capacity of 14.9 tons, and I've successfully used it to launch solar arrays for the station as well as Kerballed missions into orbit. The vessel is pretty easy to control, and I've even gotten my gravity turns to the point that I can almost reach orbit without further tweaking on ascent (total delta-V needed to attain stable orbit has ranged from 200 m/s to 500 m/s, a huge improvement over the 1000-1500 m/s delta-V I've needed with earlier, cruder designs).

Thus emboldened, I then decided to add a cupola pod to my station. The resulting vehicle is pictured below:

The launch vehicle remains the same, but the dummy payload has now been replaced with a 2-stage spacecraft; the lower stage is to get the module close to the station, while the module itself has been fitted with RCS jets and a monoprop tank for final docking.

Here's the problem: When I launch the craft, so long as it's headed straight up, it remains on course with ASAS activated. But when I begin my gravity turn (after 4 of the 6 boosters have been jettisoned), the spacecraft turns end-over-end and ASAS will not stabilize it. I'm forced to control it manually with SAS disengaged to stop the spin, wasting precious fuel. Ultimately, the spacecraft never reaches orbit, and I'm forced to either hit End Flight or kill a Kerbal.

I'm at a loss. The cupola pod is the only new element to the mix, and I don't think it should be doing this. I know others have successfully brought cupolas up to their stations or built stations around the cupola and have orbited them successfully, so I'm guessing it's something I'm doing - or not doing. Can anyone help?

[Procyon]

Only thing I can think of is , did you add it after removing the dummy payload, or build from scratch? Maybe a "control from here" issue? I think I have made all mine that use the new cupola from scratch and have never had an issue.

just throwing ideas at the wall

[Tokay Gris]

How about disabling some of the gimbals in the lifting engines?

Did wonders for my heavy lifters.

[AndrewBCrisp]

I swapped out the dummy payload and replaced it with the cupola + booster stage. I've done this for my more successful launches as well, so I don't think that's the problem. Thanks anyway, though.

EDIT: Hi Tokay, saw your reply after posting. I only have 1 gimbal engine on the spacecraft - a single LV-T45 in the core spacecraft. The other 8 engines are all LV-T30s.

Edited May 30, 2013 by AndrewBCrisp
edited to reply to Tokay.

[Tokay Gris]

Hm... how about the obvious stuff: Have you checked the fuel lines?

There seems to be some change with 0.20, some don't work the way they used to anymore. An imbalance caused by one tank emptying faster would explain the tumbling.

[Joe_Bender]

Well, I count six boosters with motors and possible one main engine but in the staging, it shows nine motors in the first stage.

Do you have a tri motor setup on the main tank? If so, why?

Also, you can combine stage five and stage six because you want the clamps to fire off when you crank the motors on. Besides, you dont need that many clamps, try three or two instead.

Is the final objective just to attach the copula only to the station? If so, I would flip the copula upside down so that when you dock it to the station, it faces up from the center of your base.

This is just some of my comments.

[Spaceisbeautifulul]

Nevermind forget this post.

Edited May 31, 2013 by Spaceisbeautifulul
Saw it was the same

[Tokay Gris]

I don't think this is the problem. Once he jettisons the boosters, he should be to high for the fins to have any effect.

Also, this problem would be there with the dummy load as well.

Have you tried clicking on the command capsule on the launchpad an "control from here"?

Or maybe the ASAS is upside down? Did you turn the design around during construction?

And do I see a "dummy" brain in there? Is there even a pilot in the cupola pod? If you put the "dummy"-brain in there first, it could actually be empty... and since I don't see any batteries, could the dummy-brain be out of energy?

Come to think of it.... is that probe (the "dummy" brain) upside down?

EDIT: correction... I see batteries... so thats not it.

[AndrewBCrisp]

Tokay: the fuel lines are fine. I watched the fuel flow as the rocket went up - no imbalance.

Joe_Bender: Answering last question first: The final objective is indeed to attach the cupola to the station. I did think of flipping the cupola at first and found that, when I launched my spacecraft two additional problems surfaced:

1. Control shifted to the Cupola module and I found my navball inverted (minor - I just had to remember which way the spacecraft was pointed to compensate).

2. The spacecraft began to roll almost immediately after launch, even with SAS engaged. This made control impossible, so I ditched the upended cupola for the design you see.

Answering the other questions last: First, I do have a cluster of 3 engines (2 LV-T30s for lift and 1 LV-T45 for control) at the bottom of the core, arranged in a line.

Second: I have kept the staging of the motors and clamps separate for my own reasons; the .5 to 1 second between ignition and liftoff does not use enough fuel to affect the flight profile. Further; unless I'm planning to launch a thousand-part monstrosity, saving 2 or 3 parts by ditching a few launch clamps hardly seems worth it.

Everyone: you seem to be looking at the launch vehicle as the source of my problem, despite what I've written in the original post. I am going to repeat myself here: the launch vehicle is not the problem. I have tested it repeatedly, both with the dummy payload and with actual spacecraft and space station components. The tumbling problem manifests with the cupola module only. May we please focus on that? Thank you.

[Tokay Gris]

Ah... that could be it: in the Sigma-I you started from top to bottom. So, the big probe-core is the controlling one.

But you then - that's what I guess - removed the dummy load with the big probe-core and the small one (underneath the big decoupler) became the new main one. And that one could be upside down... which would be no problem in the Sigma-I, because it was not controlling anyway. But in the new one it is.

P.S.: Is there a reason why there is a big decoupler? the pieces under and above are small ones...

[Tokay Gris]

Hm... How about linking the craft file?

What happens if you put some other command capsule on top?

Just change that one... and try it. If the problem persists, it is not the capsule. If it is gone, it is the capsule. Have you tried that?

[Tokay Gris]

Hm... so upside down parts are not the problem... It was an idea...

[AndrewBCrisp]

Tokay; I took your suggestion and swapped out the cupola for a Mk1-2 Command Pod and launched. Control to orbit was near perfect. So the problem remains with the cupola.

On the outside chance I missed something in the specs, I went over the cupola specs looking for anything unusual. SAS torque is the same as all the others, it's slightly more massive than the Mk1-2 pod but not so much as to unbalance the craft.

The only other thing I can think of are the 4 lights I put around the cupola; they clip with the pod itself. I can't see how that might be the problem, but I will try launching without the lights and see what happens.

[Colonel_Panic]

Hi, check that your navball isn't inverting after staging, and that you're controlling from the cupola, and not a probe core. also check for flex in the middle. If the cupola is wobbling, the navball will wobble, which can cause resonant oversteer problems with ASAS.

Consider swapping out your engine cluster at the bottom for a mainsail, and also check that when staging, the detaching radial tanks aren't knocking one of your engines off, since I think I see a lip sticking out past the bottom of the main tank.

All 3 of your cluster should be LV-T45s if you do keep the cluster, or thrust will be uneven. The radial engines should all be LV-T30s.

[AndrewBCrisp]

So much for the clipping lights causing loss of control. Even switching control to a probe core on the upper stage didn't help. More and more, I am convinced the problem lies in the nature of the cupola itself.

Unless anyone else has had this problem with a cupola as part of their spacecraft and solved it, I am declaring this problem UNSOLVABLE. Possibly a bug, possibly some trick I have not seen, but for now, no cupolas for me, it seems.

To Colonel_Panic: I will repeat what I have said to others earlier and which you apparently missed:

The launch vehicle is not the problem!

Swapping engines will not help. Staging is fine. There is no flex: I checked. Nothing is being knocked off: I checked. Every time.

And repeated use of the launch vehicle with its current engine setup worked flawlessly with other launches: both kerballed command pods and unkerballed space station parts with probe cores controlling them. It is only when the cupola enters the equation that flight profiles go pear-shaped.

If I sound a bit testy in my response, it is because I have had to repeat this statement over and over, and people STILL try to fix the launch vehicle. How many times do I have to state "The launch vehicle is not the problem" before people get that I mean what I say?

*deep breath*

Now that I've made myself clear, do we have any insights about the cupola itself?

[iamaphazael]

Tokay; I took your suggestion and swapped out the cupola for a Mk1-2 Command Pod and launched. Control to orbit was near perfect. So the problem remains with the cupola.

It may just be that the cupola doesn't have as much torque as the regular pods. It may be intentional

edit:

In fact, the cupola's parameters are:

rotPower = 30
linPower = 10
Kp = 1.0
Kd = 1.0

And the 3 man pod's are:

rotPower = 20
linPower = 20
Kp = 1.0
Kd = 1.0

I'm not entirely sure what rotPower and linPower control, but I'd suspect that if you edited the cupola's linPower to 20, you'd find that it behaves the same as the other pods

Edited May 31, 2013 by iamaphazael

[Colonel_Panic]

It may just be that the cupola doesn't have as much torque as the regular pods. It may be intentional

edit:

In fact, the cupola's parameters are:

rotPower = 30
linPower = 10
Kp = 1.0
Kd = 1.0

And the 3 man pod's are:

rotPower = 20
linPower = 20
Kp = 1.0
Kd = 1.0

I'm not entirely sure what rotPower and linPower control, but I'd suspect that if you edited the cupola's linPower to 20, you'd find that it behaves the same as the other pods

It's also heavier so you may want to take that into consideration as well. More weight, less torque.

I've never had a problem with one, but I never have used -just- a cupola for a command pod... if I want to go light, it's hardly efficient, so I only really use them as station parts or observation decks on really massive ships.

[rodion_herrera]

I am betting this is a torque issue. Compare torque values of the Cupola and your previous capsule, and any probes you might have attached. See if the torque differences are radical enough to affect ascent with SAS/ASAS.

EDIT: I didn't realize that iamaphazael is talking exactly about the same thing, my bad.

[Joe_Bender]

Well, all I can say is that I have launched several vehicles with the Copula as the main and I did not have any issues with flipping.

I take it that you are just swapping the dummy with the copula on a built rocket.

Have you tried building a brand new rocket with the copula using the same parts and see if you still have this issue?

When I experience flipping, it is typically because my rockets are flying faster than what they should be. The point of thrust is not inline with the rockets retrograde point and the rocket has >3.0TWR and the motor just pushes the rocket over......aka oversteer.

[Tokay Gris]

When I experience flipping, it is typically because my rockets are flying faster than what they should be. The point of thrust is not inline with the rockets retrograde point and the rocket has >3.0TWR and the motor just pushes the rocket over......aka oversteer.

Hm... can't be the case here.

I will try tomorrow to rebuild that lifter (maybe you could link to the craft file?) and see what happens.

[AndrewBCrisp]

Checking in on my lunch break so I'll make this brief.

From what iamaphazael posted, it looks like the actual torque on the Cupola is stronger than other modules. I'm not sure what "linPower" is supposed to refer to (perhaps torque for translation rather than rotation?) but the rotPower is a lot bigger than expected.

Combine this with Colonel_Panic's statement that he's never launched just a cupola to a station (as I'm trying to do), and we might have a solution: I need to weigh down the cupola.

A larger module with a cupola on top might provide enough mass to compensate for the cupola's crazy-strong torque and allow for more control, again like Colonel_Panic said.

Since I'm at work, I can't try this now, but I'll whip up a heavier module tonight. I'll probably also need to switch launch vehicles to the bigger Sigma-II, as I'll probably exceed the Sigma-I's lift capacity.

Tokay: you'd asked me to link the craft files earlier and I missed that; my apologies. I can upload my craft files to Google Drive; I'll do that when I get home.

Thanks everyone; not sure if this is the answer, but it seems to fit the evidence. I'll let you know what my tests come up with.

[triffid_hunter]

Have you recently installed FAR? that causes most of my rockets to flip wildly. Have you tried 0.20.2 which apparently includes fixes for the cupola?

[AndrewBCrisp]

So much for uploading to Google Drive; I can share a file or folder but only with specific people, it seems. Tokay, could you PM me an email address? I can email you the files instead.

triffid_hunter: I have not used FAR; I tend to use as few mods as I can get away with (Crew Manifest, Subassembly, and Chatterer. Currently using none; wanted to wait until any issues with compatibility had been solved before re-installing). And I am using 0.20.2, though the issue still appears in that version (I understood the cupola fixes were due to cupola-equipped spacecraft and stations "spontaneously disassembling"). Sorry.

Edited May 31, 2013 by AndrewBCrisp

[Nao]

Maybe open the craft file, copy all the contents, and paste it into http://pastebin.com/ it has 512kb max file size so it should be ok.

Im getting more and more curious about that phantom coupola

[AndrewBCrisp]

Hello again all:

SUCCESS!

Weighing down the cupola was the key. I whipped up a suitable module and a more robust upper stage, and because the mass of the new payload exceeded 15 tons, put it on my Sigma-II launch vehicle. You can see the completed unit in the screenshot below:

Liftoff and staging went as well as before, and I started my gravity turn at 10 km. This time, control was near-flawless. I was able to turn the craft and stop the turn as easily as if I had any other command module on board. Once I got the craft into orbit and deorbited the last stage of the launch vehicle, I conducted a few tests of the now much-lighter craft and control remains good. I think we nailed this one. Thanks again to everyone who helped!

Now to dock the module to the station, which will be a challenge in and of itself...