Welcome to Keen Software House Forums! Log in or Sign up to interact with the KSH community.
  1. You are currently browsing our forum as a guest. Create your own forum account to access all forum functionality.

Lazy Gyro Roll

Discussion in 'General' started by TownWitchdoctor, Mar 10, 2019.

Thread Status:
This last post in this thread was made more than 31 days old.
  1. TownWitchdoctor Trainee Engineer

    Messages:
    9
    Using Q&E ships seem barely able to roll but pitch and yaw work fine, using both mouse and arrow keys. Anyone else experiencing this? A ship I'm working on at the minute has 16 gyros, I even placed them in different orientations, and still rolls lazily. I tried swapping 2 out for x25 modded ones for no real gain (just don't touch the mouse). Funny thing is if I use 60rpm override settings on one it spins at the same speed when set to any direction so it seems to just be Q&E not working properly.
     
    • Like Like x 1
  2. Spaceman Spiff Senior Engineer

    Messages:
    1,998
    Gyroscope orientation in the game is irrelevant. For maximum effectiveness, the gyroscopes should be placed as close to the center of mass as possible. Double-check that all your thrusters are working correctly. Also, exiting completely out of SE and reloaded usually fixes glitches like this.
     
  3. plaYer2k Master Engineer

    Messages:
    3,160
    About three years ago, the code had a multiplier of 0.2 to the roll. Thus any roll around the forward axis (Q/E) was 1/5th of any other rotation.
    This very likely was by design to prevent overresponsive rolls and couldnt be disabled.

    I am fairly sure that this still persists.
     
    • Agree Agree x 2
  4. ShadedMJ Apprentice Engineer

    Messages:
    265
    I've notice the lack of roll frequently for quite a while now. There is a 'zone' in adding gyros that it gets noticed, and about the only solution is to either decrease mass or increase gyros to get out of the bad zone.
     
  5. TownWitchdoctor Trainee Engineer

    Messages:
    9
    I'll try a rework around centre of mass, although I've got 8 of them and subgrids upon subgrids. I have noticed lazy roll with the vanilla lander though. This fighter has keybound control surfaces and thrusters and is meant to look agile but falls short in practice. I did initially have real wings that pivoted and it rolled superfast in atmosphere but was an absolute nightmare to control over 30m/s.
     
  6. vadersson Trainee Engineer

    Messages:
    78
    I ran into an issue that I could not roll my ship over with Q and E (or anything else). I would roll about 45 degrees and it would just stop and no go further. I was going for a full 180 but could never get there. I am not sure what was going on. I will have to try it again.
     
  7. TownWitchdoctor Trainee Engineer

    Messages:
    9
    I shifted some of them around and removed the ones in the wings, made it slightly better, I think the ones in the wings were just adding weight and making things worse. Putting one on the pivot point made the biggest difference, unfortunately the pivot point is in free air about 8 blocks below the bottom of the ship somehow. I've now got two on override that toggle on/off in the keybound timer blocks that control the wing and thruster pivots, they take a second to spool up but you hit 60rpm roll in no time.
     
    • Like Like x 1
  8. Ronin1973 Master Engineer

    Messages:
    4,955
    The size of your bounding box will influence how fast you can roll. The amount of mass and how far it is from the center of mass (along your roll axis) will also make a difference. Gyros are heavy, so their placement will make a difference... but so will other heavy blocks.

    But yes... roll seems a bit under-powered, especially when stationary.
     
    • Like Like x 1
  9. Stardriver907 Master Engineer

    Messages:
    3,368
    This is interesting because my experience has been that roll (Q/E) was the most responsive and yaw (A/D) was the least.
    This is true as long as you don't use manual override, in which case orientation is critical. Because I build large ships, it's just not fun to hold down a key to get them to move. I recently discovered that it takes my flagship a full twenty minutes of maximum effort to turn 180 degrees. My first lesson learned is that all gyros must face the same direction so they at least all induce motion in the same direction. My second lesson was that if you want the ship to yaw when you use the yaw override, your gyros should be facing forward. Modded gyros often do not provide visual cues as to "which way is up", so there was a lot of trial and error (mostly error) on my part. Now I orient my flagship using hotkeys linked to overrides.

    Also, it seems gyros are more responsive if they are run by the auto-pilot. My ships have done some amazing, almost terrifying things when they miss a waypoint. Maneuvers I don't think I could duplicate under manual control. I wish I had video.
     
    • Like Like x 1
  10. Spaceman Spiff Senior Engineer

    Messages:
    1,998
    I just remembered that I recently built a ship that has a large subgrid of refineries that spin. Whereas the ship doesn’t roll, it will not maintain altitude when hovering within a planet’s gravity. When I disabled the subgrid, the unstable behavior disappeared. I suspect your problem is all your subgrids...it’s unstable.
     
  11. Calaban Junior Engineer

    Messages:
    994
    Do some tests with the gyro placement. This is a new Universe, so we are all learning he changes. This could be your specialty engineering research.

    In past iterations of the physics, which may still hold true today- is that grid gyroscope and deflection rate depended on their position on the grid relative to the grids center of mass (displayable). having gyros close to the center of mass gave one benefit, having them far from center of mass gave them another.
     
    • Like Like x 1
  12. plaYer2k Master Engineer

    Messages:
    3,160
    The reason why gyros, like any other block with high mass, should be placed close together (at the center of mass) is the moment of inertia.
    What is mass to a linear acceleration is the moment of inertia to angular acceleration.

    Here is a typical example for various bodies with same mass but different properties for their moment of inertia.

    [​IMG]

    As seen there, the solid ball accelerates best, the solid sphere follows. The red body is a hollow ball (sphere) and the green body a hollow cylinder (tube).
    https://en.wikipedia.org/wiki/Moment_of_inertia

    Thus if you spread the high mass blocks out over the whole ship, far away from the final center of mass, the moment of inertia is "pretty bad" (high).
    Also keep in mind that it increases with the squared radius, thus 2x as far from the CoM means 4x as bad.

    However as for the roll specifically, there still seem to be various elements in the code limiting the roll efficiency or "impact" as to not be overly sensitive, as mentioned earlier.

    The orientation of rotors only matters when you want to override them manually. As it is far easier to do it when they are all aligned the same way and you can just apply the same setting to every single one.
    If the orientation varies, then you had to check their orientattion and manually set them properly for each. Using scripts, you should get around that tho.
     
    • Informative Informative x 6
  13. Calaban Junior Engineer

    Messages:
    994
    Clangs' universe is a bit.. special... His laws of physics are equally... special. Oh, and by the way- He Hates You (mucking around in His universe).

    So reality may not be too relevant, there. Good info above, tho.

    Here is a practical example, tho:
    I use gyros to right my rovers when they flip over. I have learned through extensive tinkering on my favorite rovers that:
    - placing them in center of gravity helps stability, but does not leverage well for flippin that bug over.
    - placing them on the extreme end- under the back bumper in my case, works wonders- flips that rover right over. Bouncing past the wheels and all.

    Then, looking at the old yellow bus (RIP) and the new space pod, the gyro is also at the far rear of the mass, and boy that nose swings around, doesnt it?

    So, through my refusing to do any math, and sticking to Ash, from housewares wisdom of "shoot always, ask questions never", my gut feeling for getting insane roll rates on a ship is... put gyros on the wingtips.. and hold onto your lunch trying to fly that way.

    But let me watch THAT test flight:munch:
     
    • Agree Agree x 1
  14. Stardriver907 Master Engineer

    Messages:
    3,368
    Actually, back in the days before capital ship gyro mods, I "solved" my turn speed problems with my flagship by placing gyros at the corners. Keeping in mind that mass was handled a bit differently back then, and we're talking about vanilla gyros for the moment, the old ship took 45 actual minutes to do a 180 with gyros in the center, and that was cut by about fifteen minutes with gyros in the outermost corners. Still we're talking about hundreds of gyros.

    Again though, today with eight modded gyros I get 180 degree yaw in about twenty minutes. However, I can do a roll in less than a minute. I've never done a pitchover but I suspect it would take about the same time as a yaw turn. The fact that I have not considered putting gyros out on the corners again just means I'm satisfied with the turn rate since it is not a fighter.

    I believe Calaban may be right, though. If all the gyros are oriented the same way (just to be sure ;)) ant you have some on the outer corners or edges, you may see an increase in turn rate.
     
  15. Malware Master Engineer

    Messages:
    9,861
    The gyros don't matter at all when it comes to the actual turning. The game just collects the force available of all of them and uses that combined force to rotate on the center of mass. Location is not taken into consideration. Orientation is only taken into consideration for overrides as mentioned by others.

    This info comes from looking at the source code.
     
    • Agree Agree x 1
  16. TownWitchdoctor Trainee Engineer

    Messages:
    9
    Not locations per se but mass distribution and pivot definitely is. Here's a picture of the mass and grid pivots on this ship: https://steamcommunity.com/profiles/76561198293071986/screenshot/998016940157941154

    You can see the main grid pivot is below the landing gear by about 8 small blocks, so the gyros are basically trying to swing 70 tons around at the end of a long stick. Probably why the override gyros need to spin-up first, they're mounted to the front of the cockpit either side close to the centre of mass and close-ish the gird pivot. I'm getting good maneuverabilty now, but before it was pretty sluggish.

    When it still had modded wings it was impossible to roll in atmosphere with them locked, it could only do it by pivoting the control surfaces (makes sense really), but the wing rotors were easily overpowered in maneuvers so I took that functionality out. Now it just pivots the control surfaces coz it looks cool. I know subgrids are "unstable" but I ususally have pretty good luck with them in offline play, the occasional terrifying moment or wonkiness is a fair tradeoff for all the extra stuff you can do.
     
  17. plaYer2k Master Engineer

    Messages:
    3,160
    Pivot point doesnt matter, or shouldnt.

    You can change the pivot (which is essentially just the block coordinate at 0,0,0 iirc) by pasting that ship onto another grid. So place a single new block floating in space, paste that ship onto it, remove the one block and compare both ships. It shouldnt make a difference. It certainly didnt back then and if it does now then it's worth a bug report.
     
  18. Malware Master Engineer

    Messages:
    9,861
    Indeed. Center of mass, not pivot point.
     
  19. TownWitchdoctor Trainee Engineer

    Messages:
    9
    Can confirm I just moved it to the landing gear, I've never actually looked at those aside from sensor field range. I'm guessing that gyros don't work throught subgrids then? There used to be a load in the wings til I took them out.
     
  20. Malware Master Engineer

    Messages:
    9,861
    No control mechanisms work through subgrids.
     
  21. TownWitchdoctor Trainee Engineer

    Messages:
    9
    Fair enough. Thanks for the help everyone.
     
  22. vadersson Trainee Engineer

    Messages:
    78
    Is there something that shows your pivot and the center of mass? I knew Kerbal has it, but I did not think SE had that ability.

    Thanks,
    Duncan
     
  23. Calaban Junior Engineer

    Messages:
    994
    Too bad I cannot play a video game at work... and get away with it, so I cannot test it. But here is the conjecture:

    The average placement of gyroscopes in relation to the center of mass (CoM)determine their power (in terms of degrees per sec) the starter space pod for example:

    For Pitch, its very responsive, because the pitch axis plane, viewed as a vertical slice down the centerline, has the gyro far from the CoM on that plane, from middle to back end of the ship, a good 15m or so (from memory)

    For Yaw, viewed as a slice horizontally through the middle of the ship, the gyro is also far from CoM equally by 15m so equally responsive.

    For roll.. now that "plane" is viewed as a rod right through the centerline nose to tail... and if we peek our imaginary eye down that roll "plane".. huh! the gyro and CoM are on top of each other! thats why there is so little response on roll rate; theres no "leverage separation" or whatever laws of conservation of momentum Clang has dictated.

    At least, thats the conjecture. Now get testing! I've got work to do.:munch:
     
  24. Ronin1973 Master Engineer

    Messages:
    4,955
    The game caps the turn rate of large and small grids. The maximum turn rate is affected by the distance from your center of mass to the furthest block rotating in that axis. In other words the speed of the rotation is governed by the maximum speed of that block turning on the outside. You can add an infinite amount of gyros but that speed limit will always be observed. More gyros means faster acceleration to the maximum rotational speed. But once you get there, you're done.

    Basically look at the size of your bounding box and the center of mass. That will give you an idea of how slow your ship will turn. Extremely long ships will turn slowly, simply because the outermost block is hauling-a as it pivots around the center of mass.
     
    • Agree Agree x 2
  25. TownWitchdoctor Trainee Engineer

    Messages:
    9
    • Like Like x 1
  26. Ronin1973 Master Engineer

    Messages:
    4,955

    I did something like this. Fighter with atmospheric and ion thrusters capable of exiting atmospheres. On my build I used six rotors (locked with landing gear for safety) with converging gatlings that could be set to converge up to 800 meters... when you really need to concentrate some fire at a large ship or when dogfighting.
     
Thread Status:
This last post in this thread was made more than 31 days old.