Jump to content

allista

Members
  • Posts

    2,208
  • Joined

  • Last visited

Everything posted by allista

  1. But as a small parting gift I would like to left you with another beta version. But any beta, mind you. Despite the numbers in a filename, it is a full fledged TCA-3 beta! I can't list all the new features from from the phone, but here's a few major ones: 1) support of multiple vessels simultaneously 2) advanced autopilot with navpoint navigation and automated landing 3) comprehensive engine profiles with automated switching (but that's the most untested one) You may download it here: https://www.dropbox.com/s/55scbipzo99z4gg/ThrottleControlledAvionics-v2.3.0.6.zip?dl=0 Please, do test it while I'm hunting for some hyperthermophilic archaea in hot acid swamps =]
  2. I'm terribly sorry I was unable to attend to the forum last weeks. Right now I'm waiting for a flight to Kamchatka where I'm going an expedition, so I doubt I'll have any means of communication for several weeks. Still, I'll try to answer the last few posts. Yes, now I see, sorry. To answer your question: I was going to factor out some utils, resized included, to a library, but that would be a decent but of work. Besides, why don't you just use the TweakScale? It is much more flexible, have all the options and is more stable and well supported. Find the TCA.glob file under ThrottleControlledAvionics subfolders and look at the top. There's an option just for that. Beware, though, that this file should be rewritten on update. Noted. I've tried to make a fix for that, but haven't tested it yet. It'll be included in the next beta.
  3. Glad for you Yes, the latest BETA version is 2.3.0.5, the latest STABLE is 2.3.0.1.
  4. Well, it should, because its main purpose is to balance engineS. One can only balance something with something else
  5. Currently only keyboard is supported. I'm thinking about adding joystick support in the future.
  6. Are you playing in the career mod?
  7. Yep, that should fix the issue. I'll check the problem with the CKAN. Again - - - Updated - - - ÒõрÑÂøѠúðúðÑÂ? ÃËœ úðúøõ ôòøóðтõûø? Òþþñщõ, хþрþшþ ñы ÑÂúрøýшþтþò. *ü/c -- ÑÂтþ ÑÂúþрþÑÂÑ‚ÑŒ
  8. "It"? Could you please be more descriptive?
  9. I'll check the thing with KAX. As for the first problem, what exact version do you use? v2.3.0 or v2.3.0.1? EDIT: KAX Jum Jets' configuration of ModuleEngines is not converted to KSP-1.0 format. Until they update their configs, I cannot provide any support; nor can TCA work with them properly.
  10. I'm afraid this is not possible: each ModuleEngine responds to the mainThrottle independently, so if it's set to 0, no matter the thrust limit, engine won't fire. That's why RCS is there. One plausible option for such beasts is to have custom-made (or TweakScale-d) RCS thrusters with big enough thrusting power. *BTW, in the next beta which I'm currently testing maneuver engines will respond to translation controls.
  11. So your maximum vertical acceleration is (TWR-1)g, which is ~2m/s2 in the first case. Lets say you are flying at 250m/s with zero vertical speed at 200m over the slope of 10deg. This is equivalent to the relative vertical speed of -250*tan(10)=-44m/s (i.e. the ground is approaching your craft at that speed). To counter that speed you have 200/44=4.55secs; but even if your engines react instantly you will only gain 2*4.55=9.1m/s.
  12. Hm... these pictures made me think: how much TWR your hover engines provide? Because if it's less than 1.5 it may be just too low to accelerate fast enough. Also, to balance your horizontal thrusters (its them that are trying to pitch down the craft; they're in Manual mode, right), TCA needs to decrease thrust of the aft hover engines which lowers the overall TWR.
  13. What kind of trouble? Slow response, or something else?
  14. What exact version are you using? And btw, could you test TCA with other engines? With stock, with others from KSPIE? Because I'm unable to reproduce it, so it should be either KSPIE, or some other peculiarities of your particular setting.
  15. Here's the next BETA version. It uses relative, rather than absolute vertical speed when the Altitude Above Terrain is enabled, thus having improved response time of Altitude Control It also has a completely rewritten main TCA architecture (modular instead of a single MonoBehaviour), so it may contain some new bugs Please, test it as much as possible.
  16. Then the whole "Altitude Above Terrain" thing should be provided with such part. I don't know if I want to add parts development to my current working schedule.
  17. Technically it is possible using raycasts (which is relatively cheap, as Unity and KSP do it). I already thought about it, but haven't got to it yet. - - - Updated - - - That was my thought, yea) But still it would be nice to have them balanced all the time. Even with RCS Build Aid it is a pain in the ass to manually balance thrusters. Rocket fuel depletes quite fast and the CoM moves...
  18. Thanks! Well, that is expected: ducted fans should have some considerable response time, right? Besides, 200m above and 260m/s horizontal speed... let's say you're facing a 20deg slope; then time to crash at level flight is 2.11 secs. Even if you're going up at 10m/s (which is a current limit), it's still seconds... But the problem noted. I'll think about how to prevent such situations. Either limit the horizontal speed, or increase the vertical speed above the limit.
  19. Does anyone want to test several new features? Here's a BETA version with the following changes: Horizontal Speed autopilot now considers reaction wheels when it calculates its 'attack angle'. Single-engine crafts are 'optimized' differently: if the engine generates no torque, its limiter is set to 100%, otherwise it is set to 0%. This allows to use the Vertical Speed, Horizontal Speed and Altitude autopilots with single-engine balanced crafts. Improved balancing of engines that are almost aligned with the CoM. Such engines are mostly left alone by optimization and thus should have ~100% thrust. Improved Vertical Speed Control for slow engines. No more bouncing. Added Maintain Altitude option with Above Terrain suboption. This is an autopilot that manipulates vertical speed to maintain current ASL altitude. The suboption makes the autopilot use the height from the current terrain (hills, mountains, buildings (yep, they are also 'terrain' in KSP terms)). Using main throttle control you may gradually change the altitude. Using 'full throttle' (default Z) and 'cut off throttle' (default X) you may change the altitude by 10m increments.
  20. Gimbals are acting independently on control input and there's no way to influence them, except to limit the angle. But TCA re-reads thrust vectors of engines each p-frame, so when ModuleGimbal rotates these, optimization algorithm takes it into account automatically. The autopilot that kills h-speed also calculates total thrust vector of a craft each frame (so it knows about gimballing), then tries to direct it against h-speed vector to some degree. It uses several heuristics to calculate that degree, namely: current speed, acceleration, available torque from engines, current TWR and moments of inertia of main axes. In order to understand the problem, I'll have to test your craft and crunch some figures for a while. RCS have the same problem: you can't control them programmatically, they act on control input independently and don't even have limiters. The only thing you can do is to enable/disable them. So in principle TCA could balance them using PWM technique, but that will require a whole lot more computations then is performed now. A simpler option is to balance RCS during construction and then just use translate controls from the autopilot. I'm exploring both paths right now.
  21. In previous versions there were some issues with using TCA and MJ's smart-ASS because of the indirect feedback loop the formed: in many craft designs this loop was positive and caused ever-increasing oscillations, until the Steering Gain was decreased considerably. But I haven't tested the issue for a long time, so you're welcome to try. As for RT, I haven't done anything to integrate it, but I check the vessel.IsControllable stock property, so if RT sets it to false somehow, TCA shouldn't work when a vessel is uncontrollable due to RT mechanics. Yet again, you're welcome to test it. Naturally, I would kindly ask you to post back if you test these things. Not yet. You may use RCS thrusters as usual, but TCA will know nothing about them, so you will need to balance them during construction.
  22. TCA v2.3.0.1 with several fixes for the latest bugs is released. All download links are in the main post of the thread. Changelog: Fixed issues with engines' balancing when Vertical Speed Control is disabled. Fixed handling of flameouted engines. Fixed issue with changing TCA key on some OSes.
  23. It should if that engine is in the Main Engine or Balanced Thrust role. Why? EDIT: I see what you mean. It's not the VSC, it's balancing that doesn't work with a single engine, because, well, you there's nothing to balance it with. I'll try to work around this somehow, but generally TCA works better with many engines; otherwise many steering vectors will lead to an inconsistent system of equations (that's what the "Engines Unoptimized" status means).
  24. Actually, you've just found a bug which I confirm. Without vertical speed control in some circumstances balancing does not work. I'll release the fix ASAP.
  25. Well, it should. I haven't tested it with Buran, but others did and reported that it worked well. Could you describe/show the design of the craft you're using?
×
×
  • Create New...