[1.9-1.10] Throttle Controlled Avionics

[allista]

37 minutes ago, Tokamak said:

Edit: Oh, I think I get it, at least partly. TCA has its own concept of "waypoints", that don't seem to correspond to the waypoints that Waypoint Manager uses. I think that's something to do with it

Correct -- too many differences between TCA's navigation system and stock waypoints.

But there's a way in TCA's main window (Navigation Tab) to load/use stock waypoints.

I still don't know why TCA doesn't detect targeted vessels though...

[Tokamak]

45 minutes ago, allista said:

Correct -- too many differences between TCA's navigation system and stock waypoints.

But there's a way in TCA's main window (Navigation Tab) to load/use stock waypoints.

I still don't know why TCA doesn't detect targeted vessels though...

I'm figuring out how to make do, combining functionality of MJ and TCA. It's a little fiddly, but hey, so it the whole game.

[aaaaaaaaaaaaaaaaaaaaaaaaaa]

4 hours ago, allista said:

7 hours ago, aaaaaaaaaaaaaaaaaaaaaaaaaa said:

TCA is really useful for all kinds of fancy aircrafts with jet engines! However, I didn't manage to make it work with rotors + propellers or heli blades from Breaking Ground. Is this a known issue of I'm doing something wrong? Could not find discussions of any issues about this.

I need to test it (never actually played with BG), but if they're implemented using anything other than the stock ModuleEngine, then no, TCA won't even see them as engines.

That's exactly what happens - TCA does not recognize the engines. Are there any propeller/rotor mods known to be working with TCA? I didn't find any info on this but for the mention of propellers in the first post here.

[B-STRK]

7 hours ago, aaaaaaaaaaaaaaaaaaaaaaaaaa said:

That's exactly what happens - TCA does not recognize the engines. Are there any propeller/rotor mods known to be working with TCA? I didn't find any info on this but for the mention of propellers in the first post here.

I recall in the past that Firespitter based propellers might work--then again I only used the electric prop from that set. But my suspicion is that the props and rotors that use Firespitter (is KAX still active and working?)  model them as thrust based engines and that's why they can be read by TCA. (Notable in that the FS electric prop for example, if I'm remembering the past right, is being read by RCS Build Aid in engine torque measuring mode.)

I am not sure that the BG rotors are ModuleEngine based, and even if they were, the implementation of them is rotational, not linear. And BG rotorcraft, like non-DLC stock "props", work on the basis of  (control surface) "lift",  not engine thrust. A whole new set of logic is required that accounts for RPM and prop pitch, plus cyclic/collective when enabled, against instantaneous prop lift rating and vector vs aircraft weight/gravity for an autopilot to get the most out of them. At least, that's how it feels to me when I'm testing rotorcraft performance. In that sense, thrust vector vs aircraft vector is way simpler. 

There are a number of other props (or proplikes) out there that I think are not Firespitter based but may also be using an implementation of ModuleEngine. Suggest try downloading them and see if they would work out--e.g., Buffalo/Heisenberg, not sure if the Mk2 expansion also uses FS or not...

Edited June 2, 2020 by B-STRK

[allista]

9 hours ago, Tokamak said:

I'm figuring out how to make do, combining functionality of MJ and TCA. It's a little fiddly, but hey, so it the whole game.

BTW, I'm planning to add a set of maneuver planners to TCA. And my own docking AP. Not to compete with MJ (that was never an intention), but to make TCA more useful and independent.

57 minutes ago, B-STRK said:

And BG rotorcraft, like non-DLC stock "props", work on the basis of  (control surface) "lift",  not engine thrust.

You mean they're actual propellers? 

How's that possible with 0.02s of physical frame? 

[Tokamak]

9 hours ago, allista said:

BTW, I'm planning to add a set of maneuver planners to TCA. And my own docking AP. Not to compete with MJ (that was never an intention), but to make TCA more useful and independent.

Excellent. The more ways to do things, the better.

Oh also, to be clear, I wasn't knocking TCA, which is excellent. It's just that MechJeb is part of my 'workflow' already, and when it comes to complicated mods, sometimes it's better to use what you have already figured out. 

And ho boy is TCA complicated. But you can either have something that is simple, or something that is versatile, and I like the latter.

[B-STRK]

10 hours ago, allista said:

You mean they're actual propellers? 

How's that possible with 0.02s of physical frame? 

Yep! Actual propellers (and rotors).

And with 1.9, the blades have independent pitch variance for differential collective and cyclic control (meaning they can torque the aircraft on their own, even without relying on SAS torque). 

(Yeah, I kept the SAS on because I'm lazy that way ) 

Whatever black magic the devs did in getting all of the physics to work in such compressed physical and temporal space, it works  (To be fair though, stock bearing-based rotorcraft have been on the market, I think even before Breaking Ground, so even then the game could handle rotating lift surfaces, maybe that's where Squad got the idea, all they had to do was make a part to rotate said blades. And I was running my game on default physics timeframe settings, that didn't get touched at all in Settings) And it can get crazy in there: each rotor can have up to eight nodes for blade attachment.

As to whether an autopilot could be developed to handle BG rotorcraft or VTOLs... okay, I promise I'm not pitching for the feature from TCA. Rotorcraft like this versus how TCA handles thrust-based VTOL craft are an entirely different kind of flying altogether, and we're quite thankful with the suite we have with TCA as it is. As a thought exercise, though, one shortcut I think could help for programming such an autopilot is to just leave RPM at a set rate when on, then just vary the blade pitch to vary lift. Then it's just a question of adjusting pitch to change lift force to achieve the desired vertical speed for a given vector of lift, similar to how TCA in VTOL/surface flight mode would vary thrust to achieve target speed whether in hover or moving to target. Although in my experience, I get faster responses mixing RPM and blade pitch adjustments, but then again that might be down to me personally feel-gutting the rotorcraft maneuvers. This is just a thought exercise though, I really really promise.  And that goes for everyone else reading this too, please, no pressures at all!

[allista]

44 minutes ago, B-STRK said:

This is just a thought exercise though, I really really promise.  And that goes for everyone else reading this too, please, no pressures at all!

Ok, I'll take a look. I still think they simulate+animate the rotors rather than use stock aerodynamic lift+drag model to make them work. Considering blades' total area on your screenshots, they can't provide enough lift until rpm is much, much higher than physical frame rate can possibly handle.

[allista]

52 minutes ago, B-STRK said:

differential collective and cyclic control (meaning they can torque the aircraft on their own, even without relying on SAS torque). 

There's another long-standing plan for TCA that I'm yet to find the presence of mind to start to implement...

Gimbal-based collective thrust-vectoring. Imagine a multi-engine vtol with several pairs of gimballed nozzles that can point it's thrust quickly wherever needed (subject to general constraints, of course) without ever pitching or rolling.

But that, among other things, would require disabling actual ModuleGimballs, highjacking their too-complicated configs 

[rukhafi]

hello everyone

anyone can advise me please? in flight mode, I only have a "Main Window" without floating windows. What could be the reason for this? Thank you.

[allista]

3 minutes ago, rukhafi said:

hello everyone

anyone can advise me please? in flight mode, I only have a "Main Window" without floating windows. What could be the reason for this? Thank you.

Early career game? If so, it's possible you just don't have corresponding TCA modules purchased yet.

[rukhafi]

11 minutes ago, allista said:

Early career game? If so, it's possible you just don't have corresponding TCA modules purchased yet.

No, in sandbox game

To be honest, I have another problem with engine H-VR "J Edgar" malfunctioning. It's VTOL Engine from Mk2Expansion mod. TCA starts thrust control only after the engine rotates 90+ degrees after start. The problem is that the vessel is turning over too:) 

 

[allista]

1 hour ago, rukhafi said:

in sandbox game

In any case, unless there are some relevant exceptions on the log, I'd say the reason is the absence of TCA modules that provide corresponding UI panels.

1 hour ago, rukhafi said:

engine H-VR "J Edgar" malfunctioning. It's VTOL Engine from Mk2Expansion mod. TCA starts thrust control only after the engine rotates 90+ degrees after start. The problem is that the vessel is turning over too:) 

Need to test this, thanks.

[GummiRevolution]

11 hours ago, rukhafi said:

hello everyone

anyone can advise me please? in flight mode, I only have a "Main Window" without floating windows. What could be the reason for this? Thank you.

I'm having the same problem, running on KSP 1.9.1. I'll try a clean install if it's any mods I have, but I don't think it's mod interaction. The modules just disappeared.

[allista]

7 hours ago, GummiRevolution said:

I'm having the same problem, running on KSP 1.9.1. I'll try a clean install if it's any mods I have, but I don't think it's mod interaction. The modules just disappeared.

Share the Player.log, please. Let's see what's going on.

[Tokamak]

This mod is both massively complicated to use, and massively powerful. The former is an inevitable result of the latter, obviously. This must be a huge project. Thank you for making it! It allows me to make some very fun and useful craft, and be able to use them with my meager piloting skills.

[allista]

On 6/2/2020 at 9:37 PM, allista said:

Ok, I'll take a look. I still think they simulate+animate the rotors rather than use stock aerodynamic lift+drag model to make them work. Considering blades' total area on your screenshots, they can't provide enough lift until rpm is much, much higher than physical frame rate can possibly handle.

@B-STRK just took this look at the rotors. They are actual thing. Rotors are just that: a set of rotating fixed joints; and the blades are lifting surfaces with tuned custom curves for lift/drag and all. And the lift is so large that you don't need them to rotate that fast: engine rotor is limited to 420rpm. That's 50 degrees per normal (without time warp) physical frame.

Anyways, not a chance I could handle them in TCA. To do so I need to approximate the thrust they would produce; that is, to simulate stock aerodynamic model. Each physical frame. Because TCA relies on prior knowledge of each engine's characteristic to optimize thrust limiters before the thrust would be applied in the current frame. And it focuses on systems of multiple engine's.

The best way to approach this, in my opinion, is to make a dedicated mod like this:

And treat each rotor as a control system problem, transfer function and all that.

[JajajTec]

Thanks so much for the mod! You are a life saver! btw. is it possible tu use engines as a large-one direction rcs? because i am planning to make a SPACE thug...

[GummiRevolution]

I fixed it, turns out I extracted into ToolbarController. That happens to me a lot.

[allista]

11 hours ago, JajajTec said:

Thanks so much for the mod! You are a life saver! btw. is it possible tu use engines as a large-one direction rcs? because i am planning to make a SPACE thug...

It's possible, but currently there's a bug in optimisation algorithm that makes limiters of the maneuver-only engine systems to low to be useful. I'm working on it since Friday.

[Pikapolar]

Stupid question probably, but does this work with MJ?

[allista]

1 hour ago, Pikapolar said:

Stupid question probably, but does this work with MJ?

Read from here on:

 

[udayan M]

Hi! In the past couple of day I have turned into a fan for this mod. Thank you for making this mod. But I have a question.. I have created this heavy lifter for lifting the rovers from one place to another place in the kerbin, and it works fine. But, I would like to know if I can go back during (ETA 1hr) to the space center or tracking station when this mission is continuing in autopilot. Once the heavy lifter reaches the target kerbal alarm clock pops to let me know so i can continue the mission!!!???... 

[DStaal]

7 hours ago, udayan M said:

Hi! In the past couple of day I have turned into a fan for this mod. Thank you for making this mod. But I have a question.. I have created this heavy lifter for lifting the rovers from one place to another place in the kerbin, and it works fine. But, I would like to know if I can go back during (ETA 1hr) to the space center or tracking station when this mission is continuing in autopilot. Once the heavy lifter reaches the target kerbal alarm clock pops to let me know so i can continue the mission!!!???... 

No.  KSP doesn't work that way - ships not in the current physics bubble (of a radius about 2km from the current ship) are unloaded and take no actions.  If they're orbiting something, they'll continue in the same orbit.  If they are landed they'll get reloaded in place.  (Likely moving a small amount when the terrain and their location regenerates.)  If it's flying in atmosphere, it will crash.

There are some mods that modify that, but no mod allows the ship to do multiple actions when not in the physics bubble.  They either move the ship (along a mod-determined path that really doesn't exist - they're just moving where it will re-create next time you visit it) or they hold the ship in place.  In both cases all the mods really do is modify the ship's location, they still aren't taking actions.

[Tokamak]

Quick question, because I m ust be doing something wrong:

I'm trying to use the "land" function from orbit. I'm using the one in the "orbital autopilots" of course. I target the craft that I want to land next to, click "land", and click "start", and TCA gives the error "Target should be a vessel or waypoint". But my target _is_ a vessel. What am I missing?