Tiny unKerballed probe into LKO? Stability trouble over 25km.

[Mister Spock]

I'm trying the Unkerballed Start tech tree along with the Career Evolution contract pack, all of which emphasizes flying unmanned probes for a long time.  My mod list is pretty short otherwise: OPM, EVE, KAC, Final Frontier.  It's been a while since I played KSP, but I can't believe the trouble I'm having getting an unmanned satellite into orbit.  I've managed suborbital flight and recovery fine.  I've upgraded the launch pad once, so I can use up to 18t.  I'm in Career mode on normal default settings.  I've got Thumpers unlocked and the Reliant and Swivel, but those rockets tend to end up too heavy.  Plus I'm short on cash.

So I've been trying engines I've never used before -- Spiders and Twitches, stacked on tiny little Oscar fuel tanks.  I'm trying to launch the first Octo probe.  For just 4 or 5000 funds, three stages with these little guys gets me 3600 delta V or so, which is pretty cool.  The Twitches have plenty of TWR to get off the launchpad.  It's all very cute and small!  I love the idea of putting teeny little sats like this in orbit.

Except that I can't fly the darn thing.  The first stage is fine; I easily get to 10,000m and can start a gravity turn, trying to keep speed under 300.  Those little Twitches have nice gimballing.  Plus I have the small basic fin on the bottom of stages 1 and 2, which makes the thing look like something out of Ray Bradbury.  (Sorry, I forgot to take a screenshot. I can post one tomorrow if need be.) The second stage, with Spiders, starts okay too, but I gradually lose control authority until the thing starts spinning.  Maybe I should try Twitches here too?  Maybe get rid of fins on stage 2?  I've tried putting the smallest reaction wheel below the octo, above it; I've tried multiple reaction wheels.  Maybe I need no reaction wheels at all, or to turn down torque on it?  One thing I haven't done is fuss with how fuel drains, because I can't remember how to do it.  Eventually the thing just starts spinning wildly and I lose control.

Any advice? Thanks in advance.

[jimmymcgoochie]

Sounds like a weight balance issue to me, you should try setting the fuel tanks to drain from bottom to top which will keep the CoM as far forward as possible. If you have the stack-mounted Ant or Spark engines unlocked, they would be a better option than using multiple radially mounted engines to reduce drag and mass.

Have you tried using a 1.25m lower stage then a 0.625m second stage? It was a pretty common thing to do for real sounding and early orbital rockets and if you can squeeze inside the mass limits (and have some form of tapering part to bridge the size difference e.g. FL-A10 adapter or a fairing) that might make it easier to reach space than relying on the relatively inefficient smaller parts. You could also try using a small SRB as your first stage as they pack a lot of punch for their size.

[The Aziz]

You mean spinning or flipping?

If it's spinning (like, you know, a spinner) then it's probably something wrong with the fins, picture would be nice.

If it's flipping, then I would blame your approach to gravity turn (or more like gravity slap really, looking at how you describe it). Low speed, late turn is a thing of the past, over five years now. It may still work while the rocket is long, with first stage still attached but then it often goes sideways, literally. It's now more like a 1.2twr on the pad, 28° at 5km, 45° at 10k, 70° at 20km and almost horizontally at 40km. Don't go too far from prograde marker, turn on aero overlay and don't let the red arrow (drag) appear anywhere else than at the bottom of the rocket.

[paul_c]

Having a fairing has a big effect on aerodynamics. Don't forget, for a small rocket, its a significant weight though, so deploy it at the optimal time (which is, when the penalty in control/speed from aerodynamics is less than the weight penalty) - around 40-50km is a good guide there.

A rocket is controlled by 3 things: reaction wheel; RCS; engine gimballing. Its possible that engine gimballing has complete control over it - but only while engines are running. Reaction wheels convert electricity into control movements, and are needed once off the throttle. For orbital positioning, its very useful to be able to point it a direction with no throttle at all (for precision burns) so I'd always recommend at least some reaction wheel, although its interesting you got it into orbit without!

Full SAS controls is a bug in the latest version of KSP. It should be that OCTO has SAS, while the HECS gets prograde/retrograde controls too (I think....but we're going back a few versions now!)

[paul_c]

1 hour ago, jimmymcgoochie said:

You don’t need a dedicated reaction wheel on a probe that small- each probe core (except the Stayputnik, OKTO2 and RoveMate) has one built in that is more than sufficient to control a smallish probe.

 

Apologies, I didn't realise some of the probe cores had a small reaction wheel included - if that's the case, then no, you won't need the additional one.

ETA In theory, you don't need 4 solar panels either. It would be a brave soul who took 1 fixed panel. 2 is marginal, 3 guarantees solar power so long as its not nose or tail to the sun (leave it in normal/antinormal - pink marker - to ensure that). If you get to the stage of unlocking the deployable panels then 2 of these will 100% guarantee sun (1 will be 99% and I've seen it obscured!).

If the worst happens and you end up in the chicken-egg scenario of out of power, and no power/control to command the controller to turn it, timewarp ahead about 1/4 year and the sun will have shifted round with respect to the craft and it should burst into life again.

[VoidSquid]

You could use an engine with gimbaling, e.g. the Spark, or the Caravel from the RLA Reborn part mod pack.

[king of nowhere]

yes, SAS won't help if you leave it turned off. or perhaps you have it on, but then the probe goes in shadow, the batteries run out, and sas deactivates.

Yes, you should add a battery. the basic one is a bit bad for aerodinamics, but there is the stackable circular one that you can put between the body and the nose cone for only 10 kg.

then the octo does not have the functionality to hold directions. it will keep your probe still, but it will not point to any specific direction. you have to direct it manually. you said your probe was drifting, right? i assume you were trying a prograde burn, started with the prograde indicator there, and then the probe gradually shifted away from the indicator? in this case it's not the probe drifting, it's the prograde direction drifting. as you circle the planet, where is prograde changes. so yeah, you will have to make the manuever by hand holding the direction and compensating for that.

Finally, i also see in the picture that the probe icon on top is in yellow. which means it has somewhat limited functionality. it's in the sun, so it has power. perhaps it is still hybernated?

[king of nowhere]

2 minutes ago, Mister Spock said:

I'm about to try a new contract, to orbit and return to earth, so this time I'll add a battery and a parachute.  Alas, I do not yet have that nice circular battery.  I'm stuck with the big clunky boxy thing.  Maybe I can put it on top of the Octo and cover it all with a nose cone.  I tried mounting 2 or 3 radially on the rocket, but the batteries dramatically increased mass.  For a tiny little probe like this, every bit of payload matters!  Even adding one battery and one parachute may require a change in the rocket.  I do see there are other tiny engines with more thrust, so I'll play around with those

A battery looks big, but it's just 5 kg.

a parachute, on the other hand, is no less than 100 kg - unless there is some modded lighter version.

on a probe so small, it may actually be more convenient to mount a more powerful engine and rocket brake.

[Spricigo]

5 minutes ago, Mister Spock said:

I wonder if my "slight adjustment" keys are working.  On Windows, by default that's alt plus the direction key (WASD), I think.

That by default  is control trim. And may be the reason why your navball is wandering around.  Alt+X will reset trim.

Precision control is now a mode, Caplocks toggle it on/off.

 

41 minutes ago, king of nowhere said:

on a probe so small, it may actually be more convenient to mount a more powerful engine and rocket brake.

Well, maybe  cheaper, cooler or something but strapping a parachute to recover the thing is as convenient as it get. 

[paul_c]

Its probably aerodynamics, at 25km trying to reach orbit you ought to be going pretty fast, and you're still somewhat in the atmosphere. That stages 1 and 2 are okay means you have the drag balanced okay on those (or that its just a bit slower).

You didn't include a pic but.....do you have a nosecone? The idea is, you want it slippery at the front and draggy at the back. If you don't have the available parts yet so that its a bit draggy at the front, sometimes you can deliberately make it "a bit draggy" at the back too, it will be less efficient but it will make it to orbit (given enough fuel).

Also sounds crazy......but if it starts flipping, turn off the throttle, wait (it will still have upwards acceleration), get it back to pointing the right way and put the throttle back on. 25km is kinda near the edge of where aerodynamics start to not matter any more...and you might just squeak it into orbit that way. Or make the 1st stage TWR higher, its inefficient too but it will "kick it" up higher for the 3rd stage.

Regarding TWR, 1.2 sounds lovely and efficient etc but sometimes you want a bit more, it will trade drag with gravity losses though - all the time you are suborbital, you are wasting (or need for gravity losses...) fuel/energy. The higher the overall TWR, the less time is spent suborbital so the lower the "gravity losses" (but the higher the drag losses).

ETA you'll need a reaction wheel - the small one should be okay for a low-weight satellite/rocket.

[Mister Spock]

@jimmymcgoochieThanks for your reply.  I have indeed unlocked the Ant, though I don't think I have the Spark; I'll play around with the Ant.  I did indeed have a successful suborbital flight with a 1.25m lower stage, so maybe I should go back to that.  But I love the idea of getting to orbit with the smallest parts possible, just because.    How do I drain fuel tanks from bottom to top?  When I click on them in the VAB, I don't see an option for that.  Maybe I need fuel lines?

@The AzizYou're right, I'm flipping, not spinning.  And yeah, I haven't played KSP for 5 years, so I have been using the "old" gravity turn. I'll try your approach.  I like the idea of a lower launch TWR.  Er, I didn't even realize there was an "aerodynamic overlay"; it seems it's new since 1.0.1.  F12 key.  I'll try that.  After rebinding it so it doesn't conflict with screenshots.

@paul_cOK, I was wondering whether it could be aerodynamics at 25K altitude, but it sounds like it can.  I do indeed have a nosecone -- the rocket is pointy at top with big old fins on the bottom of stage 1 and 2.  I'll try the aero overlay to learn more about what's happening.

Thanks to all of you for your help.  I'll report back later today! Maybe this time with screenshots.

 

 

[paul_c]

1 hour ago, Mister Spock said:

    How do I drain fuel tanks from bottom to top?  When I click on them in the VAB, I don't see an option for that.  Maybe I need fuel lines?

 

 

 

Turn on "advanced tweakables" then set the priority on the right-click context menu, which will have a bunch more options available.

Edited March 18, 2021 by paul_c

[jimmymcgoochie]

@Mister Spock You should see a "fuel priority" value with a number and +/- buttons. Fuel will drain from the highest numbered tank to the lowest so if your stage has three tanks, set the bottom tank to e.g. 11, the second to 10 and the top tank to 9 so they'll drain from bottom to top. You might have to enable "advanced tweakables" in the main menu settings to see this option (spoiler: you really want it turned on all the time, there are several very useful extra features that get enabled).

From what I remember, the radially mounted engines e.g. Twitch and Spider are usually less efficient than the axially (stack) mounted engines e.g. Spark and Ant but have bigger gimbal angles.

Screenshots would definitely help though.

[Mister Spock]

Thanks for your further replies.  I'm thinking my issue isn't aerodynamics so much as reaction wheel.  I followed your suggestions and changed engine drain flow, used Ants inline, etc.  This time I got into space easily, but once there, the slightest touch on the keyboard sends the navball spinning.  I tried decreasing the reaction wheel's control authority to 1%, and that made things less bad, but still not controllable.  I tried disabling torque and that didn't help.  Using the alt key to make movements less drastic seemed to make no difference at all.  I was flying a simple two-stage rocket, first stage with Spiders that got me out of the atmosphere, and the second stage with an Ant.  Should I remove the reaction wheel?  Er, what exactly does it do, anyway?  I've never fully understood SAS or reaction wheels. 

Also, I worry the Ant is underpowered for circularizing the orbit, as it worked very slowly.  But putting spiders or other radial-mounted engines on an upper stage does complicate aerodynamics.  

Here are a couple screenshots.  First, a curiosity about the Ant -- how does one close the gap between the ant and the next stage?  Normally, a vertical decoupler adds a metal shroud around the engine.  But the Ant doesn't do that.  It looks like this:

So I tried covering it with these newfangled procedural fairings that weren't in the base game when I played.  A bit top heavy, but the drag indicators were red only at the bottom.  Here's an example in early flight.  Note, though, that even here I'm having trouble pointing the rocket in the correct (eastward) direction.

I'm going to try removing the reaction wheel altogether and going again. But I've had this problem with every unmanned flight I've done.  It's a bummer, because the whole point of the mod I'm playing is to visit the entire solar system with probes before flying manned.  Any further suggestions?

[Mister Spock]

Update: I got that little two-stage rocket into orbit -- with lots of fuel to spare!  Kind of a thrill!  I've achieved orbit countless times before, but never with a tiny unmanned rocket like this.  And I don't think I've ever used the Spider or Ant before. In the VAB, total dV was 3,450 m/s.  

That said, I'm still having big problems with control.  For this flight, I just removed the reaction wheel altogether.  I guess the Octo has its own SAS system, because I can still cycle among SAS, Pilot, and some other setting.  Everything  was fine until I got to space; then the navball drifted constantly.  At least it didn't jump around madly, and I kept the craft from tumbling end-over-end.  It helped marginally to keep control authority pretty high.  SAS has to be on too.  I was able to point it in a wobbly prograde direction and achieve a quite high orbit, as you can see from this shot.  But in this shot, the probe was tumbling slowly end-over-end. 

The probe itself is rather useless.  It has almost no payload other than 4 basic solar panels.  I took out the antenna, batteries and science experiments so I could concentrate on the basics of flight.  But for a cost of 3,156 funds, I completed two big contracts worth many times that.  So "Tiny 1" has done its job.

Anyway, eventually I will mark this thread as "answered," but before I do, I invite anyone to chime in to help me fix my problems controlling probes.  I had this problem with the Stayputnik, too.  Am I missing some basic setting that prevents them from spinning around?

Edited March 19, 2021 by Mister Spock
Minor syntax edits. Adjusted funds figure.

[jimmymcgoochie]

You don’t need a dedicated reaction wheel on a probe that small- each probe core (except the Stayputnik, OKTO2 and RoveMate) has one built in that is more than sufficient to control a smallish probe.

You’re probably running out of electricity which is why the probe is tumbling; probe cores have puny batteries and the smaller engines don’t tend to have any power generation capacity when they’re running so you could easily run out of power especially with a big, power hungry reaction wheel to use it up. If you have the probe inside a fairing you can ditch the nosecone and put two solar panels on the top of your probe, then point that at the Sun in orbit to recharge the batteries; you can also enable ‘hibernate in warp’ on the probe core to make it automatically go into hibernate mode and save a lot of power that way.

[jimmymcgoochie]

16 minutes ago, VoidSquid said:

You could use an engine with gimbaling, e.g. the Spark, or the Caravel from the RLA Reborn part mod pack.

I second this- RLA has a lot of small parts, from fuel tanks to mini NERVs to electric monopropellant thrusters with 4-figure ISP, so if you’re doing a lot of small probes and such then it’s a good addition; not sure if it’s configured for the alternate tech tree you’re using though. Restock+ also has some additional fuel tanks in the 0.625m size range and Luciole adds a good collection of great-looking parts in this area too.

[Mister Spock]

Thanks for your replies. That RLA mod does have a lot of cool-looking probe cores and engines.  I'll probably give it a try.

Just to be clear, I do have some control over the probe throughout the flight and right now, as it orbits with the engines off.  (I have toggled on hibernation mode for now.)  It responds to commands to tilt this way or that.  In fact, I can spin it quite fast if I want.  But I just can't get the thing to hold a particular attitude, which is the most important thing -- if you can't hold a steady attitude, you can't burn in a particular direction.  Maybe I have enough electricity to rotate it but not enough power to make it hold still?  

On electricity: I have 4 solar panels evenly spaced around the Octo, as my screenshot indicates.  So it's getting some power.  But I did NOT pack batteries at all.  Is this perhaps the problem?  Are solar panels useless without batteries?  

Finally, I'm still curious why the Ant isn't covered by the stack decoupler, the way most big liquid-fuel engines are.  The resulting "gap" is depicted in my first photo above.  Thanks!

Edited March 19, 2021 by Mister Spock
Added question about the Ant decoupler.

[The Aziz]

First off, probe cores carry a very tiny amount of electric charge, so they can go dead after few minutes of no sunlight. Or even faster after moving them around too much. Single radial battery is more than enough, does not add a lot of mass, fits with the stack and easily protects from running out of juice.

Second, your pictures show that you don't use SAS. It should be always on. (The only core that does not have that feature is stayputnik, but I see you're already past that) that could also explain unplanned movements.

Edited March 20, 2021 by The Aziz

[Mister Spock]

Thanks for your replies.  I'll check whether I have SAS on.  Yes, @king of nowhere, it's possible the craft is in hibernation mode in that last screenshot.  I was toggling hibernation on and off to see how it works.  I've never used it before.

By "drifting," I mean the navball wanders around even with the engines off, in space.  It doesn't want to stay still.  Again, I may have inadvertently disabled SAS.   

I wonder if my "slight adjustment" keys are working.  On Windows, by default that's alt plus the direction key (WASD), I think.  My alt key can be a little unreliable, so maybe I should bind these keys to something else.  In previous iterations of KSP I had no problem orienting probes, burning in various directions, etc -- I had dozens of satellites.  So I'm starting to wonder if there's just something wonky with my keyboard.

I'm about to try a new contract, to orbit and return to earth, so this time I'll add a battery and a parachute.  Alas, I do not yet have that nice circular battery.  I'm stuck with the big clunky boxy thing.  Maybe I can put it on top of the Octo and cover it all with a nose cone.  I tried mounting 2 or 3 radially on the rocket, but the batteries dramatically increased mass.  For a tiny little probe like this, every bit of payload matters!  Even adding one battery and one parachute may require a change in the rocket.  I do see there are other tiny engines with more thrust, so I'll play around with those.

Thanks again for all your help.  Once I figure out how to operate my probes correctly, which I hope is soon, I'll mark this thread "answered."  Not quite there yet, though!  I appreciate all of your input a lot.

Edited March 20, 2021 by Mister Spock

[Mister Spock]

1 minute ago, king of nowhere said:

on a probe so small, it may actually be more convenient to mount a more powerful engine and rocket brake.

Wow, I hadn't though of that at all!  That would be really fun to try.

I have unlocked a mini heat shield, I think.  I suppose I'd need one of those too, even if rocket braking.  I don't know how heavy they are.  

Anyway, great idea!  I will play around with it in KSP later today.

[paul_c]

You can reduce the amount of ablator on a heatshield to reduce its weight.

[Mister Spock]

Another idea I hadn't thought of!  Thanks, @paul_c!

[king of nowhere]

if you are coming back from low orbit, you can forego the heat shield if you don't mind losing the solar panels.

from interplanetary destinations you need a heat shield.

the good thing about those small probes is that they are very light and with a large surface, so they aerobrake a lot. they fall down at around 50 m/s without parachute, so they only need a little braking.

i have experimented with even more extreme avenues of weight saving. i tried to make a lithobraking probe, no parachute and no rocket, just some structural parts that could take the impact. couldn't make it work without adding too much weight.

and i tried to add two wings to land like an airplane. it worked, sometimes. problem was, the only "wings" light enough to make it worthwhile are the basic fins, and they have abysmally low thermal resistance, so they tend to burn during reentry. but they were fun attempts.

 

[Mister Spock]

3 hours ago, Spricigo said:

That by default  is control trim. And may be the reason why your navball is wandering around.  Alt+X will reset trim.

Precision control is now a mode, Caplocks toggle it on/off.

OMG, this may explain my woes.  I've been stabbing at alt-WASD thinking I was making small movements, when in fact I was setting trim, lol!  That means I may have told my craft to "pitch up" constantly, heh. 

What's more,   I accidentally hit Caps Lock all the time, when I'm trying to hit Shift to increase throttle, or Tab to do something else.  I will try rebinding the Precision Control toggle to something I'm less likely to hit by accident.  I had no idea precision control was now a toggle.  Which is a welcome thing, I might add.  I hate using the alt key for anything.

For hotkeys, I've been consulting an outdated wiki post (on my phone, so I don't have to alt-tab out).  Sounds like I need to find a different hotkey website!

I'll try flying again tonight and report back.  I look forward to marking this thread as "answered" soon, I swear!

[Mister Spock]

OK, success!  I did a contract to get this little rocket into a polar orbit and nailed it on the first try.

Really it was a combination of things, and a combination of answers from this thread.  First it was aerodynamics, lol.  And fuel management.  And yes, I had my hotkeys mixed up, so I was hitting "trim" instead of "precision" controls.  Plus I was inadvertently toggling SAS on and off frantically because I was confused about that hotkey, too!  Plus the new gravity turn is a bit different from back in the day.  So I've marked a half-dozen posts as the "answer," as you were all right, lol.

I'm having a bit more trouble with the mission to recover the probe, but I think I can figure it out.  Thanks for all the help!