[1.9-1.10] Throttle Controlled Avionics

[Daveroski]

7 minutes ago, allista said:

Can you spare the save to reproduce this?

https://www.dropbox.com/s/drprgkoalkmn532/quicksave %235.loadmeta?dl=0

https://www.dropbox.com/s/7nfw11p2gtcp0er/quicksave %235.sfs?dl=0

Done.

It's on the Lauchpad.

D.

 

[allista]

1 minute ago, Daveroski said:

https://www.dropbox.com/s/drprgkoalkmn532/quicksave %235.loadmeta?dl=0

https://www.dropbox.com/s/7nfw11p2gtcp0er/quicksave %235.sfs?dl=0

Done.

It's on the Lauchpad.

D.

Thanks a lot! Got them too.

[chd]

This looks excellent, thank you! Just discovering it, obviously.

It doesn't look like this has a docking autopilot similar to MJs, but if they coexist peacefully (as long as they both don't try to run things at the same time) that's fine.

Thanks!

[allista]

9 minutes ago, chd said:

This looks excellent, thank you! Just discovering it, obviously.

It doesn't look like this has a docking autopilot similar to MJs, but if they coexist peacefully (as long as they both don't try to run things at the same time) that's fine.

Thanks!

Not yet, but I'm planning to make one.

Yes, TCA works with MJ alright.

[chd]

30 minutes ago, allista said:

Not yet, but I'm planning to make one.

Yes, TCA works with MJ alright.

Very cool. I don't see a way to automate Hohmann transfers either, so I won't be dropping MJ at this point, but it looks like they compliment each other pretty well.

Correct me if I'm wrong, but it also seems that because of the way this works, you'll generally have more control authority with rocket designs using multiple radially-placed engines rather than a singe on-axis engine, since you can then steer with differential thrust, yes?

Edited August 18, 2020 by chd

[Nando]

I'm still using KSP 1.8.1 because of mod compatibility is the current version downwards compatible? Thanks.

[allista]

11 hours ago, chd said:

Very cool. I don't see a way to automate Hohmann transfers either, so I won't be dropping MJ at this point, but it looks like they compliment each other pretty well.

Correct me if I'm wrong, but it also seems that because of the way this works, you'll generally have more control authority with rocket designs using multiple radially-placed engines rather than a singe on-axis engine, since you can then steer with differential thrust, yes?

Correct, but this is more relevant for VTOL crafts than for spaceships, since in space the ship would turn using RCS or reaction wheels with engines off most of the time. So in general a conventional 1-engine+RW design works well.

9 hours ago, Nando said:

I'm still using KSP 1.8.1 because of mod compatibility is the current version downwards compatible? Thanks.

Nope.

Visit the SpaceDock, ChangeLog section, to see all the versions and their compatibility info.

[BTAxis]

I'm trying to use this with KSP 1.9.1, but there seems to be a problem in that the game won't even load (gets stuck on the loading screen). I use the most recent version up on CKAN.

Here's my KSP.log.

[allista]

1 hour ago, BTAxis said:

I'm trying to use this with KSP 1.9.1, but there seems to be a problem in that the game won't even load (gets stuck on the loading screen). I use the most recent version up on CKAN.

Here's my KSP.log.

Use the latest version compiled for 1.9.1 from SpaceDock
https://spacedock.info/mod/198/Throttle Controlled Avionics/download/3.7.0

[BTAxis]

Alright, thank you.

[chd]

Are there any changes you're aware of between KSP 1.10.0 (which TCA lists as being currently for) and 1.10.1? I ask because I've tried it with 1.10.1 and seen some twitchy behavior.

For example (and I apologize for not having a save game where I can demonstrate this), I was using it to auto-land on minmus with a ship using 4 radially mounted Thud engines, and all seemed to be going well until just a moment before touchdown; then *blam* something goes totally wacky and the ship starts spinning as fast as possible and I can't get control back until running out of fuel.

My design was a bit sketchy in the first place, and it's possible that I created a situation where extending the landing gear destroyed one of the engines (happened too fast to be sure), but I'm guessing TCA would adjust to that and compensate? I definitely didn't lose more than one.

Anyway, I don't expect you to spend much time on this since I can't give you much hard data to work with at this point, but mostly just FYI and to ask about any known issues running it under 1.10.1. I'll try to save more info in the future if I see more oddness.

Thanks!

Edited August 26, 2020 by chd

[allista]

On 8/14/2020 at 11:28 AM, Daveroski said:

https://www.dropbox.com/s/drprgkoalkmn532/quicksave %235.loadmeta?dl=0

https://www.dropbox.com/s/7nfw11p2gtcp0er/quicksave %235.sfs?dl=0

Done.

It's on the Lauchpad.

D.

Ok, so the Luciole craft is not a problem; limiting AoA depending on dynamic pressure brings it safe and sound to orbit.

But the Little Squawk gives me nightmares!

Even on the steepest trajectory with thrust at 10 degrees to the prograde almost all the time it only manages to achieve 71km ApA ~40s ahead, then falls back, because the last stage has no thrust to circularize in the 80s span.

Adding relative ISP to equation doesn't help that, even though it saves some more fuel.

So I'm asking to make a video of your manually controlled ascent to learn from.

On 8/26/2020 at 6:45 AM, chd said:

Are there any changes you're aware of between KSP 1.10.0 (which TCA lists as being currently for) and 1.10.1? I ask because I've tried it with 1.10.1 and seen some twitchy behavior.

For example (and I apologize for not having a save game where I can demonstrate this), I was using it to auto-land on minmus with a ship using 4 radially mounted Thud engines, and all seemed to be going well until just a moment before touchdown; then *blam* something goes totally wacky and the ship starts spinning as fast as possible and I can't get control back until running out of fuel.

My design was a bit sketchy in the first place, and it's possible that I created a situation where extending the landing gear destroyed one of the engines (happened too fast to be sure), but I'm guessing TCA would adjust to that and compensate? I definitely didn't lose more than one.

Anyway, I don't expect you to spend much time on this since I can't give you much hard data to work with at this point, but mostly just FYI and to ask about any known issues running it under 1.10.1. I'll try to save more info in the future if I see more oddness.

Thanks!

I heard of incompatibility with 1.10.1 of some mods (mine including) to the extent that the game crashes on load screen.

But I unfortunately haven't had much time to spent on KSP past few weeks due to the workload and family stuff, so I haven't even install 1.10.1 yet 

[allista]

@Daveroski ok... After some trial and error I've managed to achieve a 306km LKO with 166m/s of dV remaining. But the target ApA was 71km 

Well, TCA managed, not me per se.

The reason is, obviously, the acceleration that had to be continued far past initial ApA. I'll try to figure out how to do it without moving ApA past the target value...

[Baslifico]

I was having the same issue as a poster above (Stuck on loading screen at Squad/parts/Command/cupola/cupola/cupola)

I'd installed TCA via CKAN.

After a bit of a digging, CKAN currently only lists AT Utils 1.9.4, however the SpaceDock download is bundled with v 1.9.6

Taking the download from SpaceDock, extracting the AT_Urils folder and copying it over my existing one resolved the issue.

No doubt ckan will get twitchy about files it didn't modify itself in future and I may need to manually uninstall AT_Utils before upgrading it next, but it works for now.

[toric5]

So I can see the idea behind the module system, where modules are unlocked via tech, and once a craft launches, you cant update things in flight.

However, every time I unlock an upgrade, I have to open all my ships in the VAB, select every avalible module, (or all the ones that would be useful for this craft) and resave them. Then, when I am building a new craft, I have to remember to enable all TCA modules every time.

It would be much more user friendly if choosing modules was not part of it at all. What if, when a craft is launched, all avalible modules get added to it automatically, instead of adding them in the VAB?

Of course, I would not greatly complain if modules were only unlocked through research, and when you researched a new mode, all craft now have acess to that new mode. RL spacecraft do get software updates, after all.

EDIT/PS

Also, I think I saw TCA changing the throttle limiter on an SRB. (I use pure SRB lower stages for small launchers)

Ill see if I can get a confirmation screenshot tomorrow.

Edited August 31, 2020 by toric5

[Daveroski]

11 hours ago, allista said:

ok... After some trial and error I've managed to achieve a 306km LKO with 166m/s of dV remaining. But the target ApA was 71km 

 

Ah No... 71km.. I never go to orbit at less than 100km unless I am doing a launch-to-intercept.

I took Little Squawk to 100km. I said so in my post, except that I missed a 'k' so it looked like a hundred meters.

I can't understand the obsession with people wanting to put things on the edge of Kerbin's atmosphere.
It means battling atmosphere ALL the way to Ap. Those little thruters are designed for vacuum so it makes sense to me to get them where they need to be as soon as possible.
The higher Ap gives time to burn to bring the Pe up to get the orbit. You don't need a video and I don't pretend I can  teach you a thing... except maybe patience.

D.

 

 

[allista]

19 hours ago, Daveroski said:

Ah No... 71km.. I never go to orbit at less than 100km unless I am doing a launch-to-intercept.

I took Little Squawk to 100km. I said so in my post, except that I missed a 'k' so it looked like a hundred meters.

I can't understand the obsession with people wanting to put things on the edge of Kerbin's atmosphere.
It means battling atmosphere ALL the way to Ap. Those little thruters are designed for vacuum so it makes sense to me to get them where they need to be as soon as possible.
The higher Ap gives time to burn to bring the Pe up to get the orbit. You don't need a video and I don't pretend I can  teach you a thing... except maybe patience.

D.

Ok, so with the help of @JonnyOThan I've managed to do at least this:

don't look at the Max. Angle of Attack -- it was 10 deg all the way, but I changed it afterward by accident

But for higher orbits in this case conventional autopilot's logic is still faulty and needs improvement...

[Daveroski]

3 hours ago, allista said:

don't look at the Max. Angle of Attack -- it was 10 deg all the way, but I changed it afterward by accident

Seriously?
10 deg flight through atmosphere?

My ghast is flabbered. Very flabbered indeed.

Shear pushing the nose down.
It made me feel that so much of the energy was wasted fighting the atmosphere.
With such a low ISP in atmosphere making it drain fuel faster and over a longer period of time.

When I tried it at around 5 deg I ran out of fuel at around 23K.

I felt that the atmosphere didn't provide any lift, indeed quite the opposite with the nose below prograde.

Having the nose even 1 degree below prograde starts to produce drag because of shear. The air speed on the belly being so different from that on the dorsal surface. I thought it was always in danger of producing a cascade effect that could only end up pushing the nose beyond recovery, presenting too much of the dorsal surface to the atmosphere causing it to 'flip'.

I was rather pleased with myself when I eventually got it to 100k orbit at around 2-3 deg.

I was still sort of impressed when MJ managed to get there at 2 deg. It certainly couldn't manage it at 5 deg.

My hat, were, I currently wearing one, would most certainly be off to you. I am humbled.

D.

Oh I almost forgot.
There was something else I wanted to discuss with you.

I don't like not being able to add your autopilot to existing craft.
Installing your mod in a mid-late game where many other ships are on the play-field, I tend to forget which ship had the mod and which doesn't. Far easier if I could just update the software of any ship on the fly.

I am not too happy that it saves the craft with the modules installed.
It gives errors to people who might want one of my craft. Telling some that these errors are harmless does not seem to pacify as it should.

I have tried the 'profile' part of your autopilot but I can't seem to get it to save a profile so that I can 'one-click' it for craft that I build. Which, if it worked , would negate having to have them saved as part of the file.

Regards,

D.

Edited September 1, 2020 by Daveroski
Senior Moment

[allista]

51 minutes ago, Daveroski said:

Seriously?
10 deg flight through atmosphere?

My ghast is flabbered. Very flabbered indeed.

It was modified based on dynamic pressure down to about 5 deg in lower atmo, but then the limit was 10. Though the actual AoA was rarely larger 5.

UPD I can upload the video of the whole thing.

53 minutes ago, Daveroski said:

I don't like not being able to add your autopilot to existing craft.
Installing your mod in a mid-late game where many other ships are on the play-field, I tend to forget which ship had the mod and which doesn't. Far easier if I could just update the software of any ship on the fly.

I am not too happy that it saves the craft with the modules installed.
It gives errors to people who might want one of my craft. Telling some that these errors are harmless does not seem to pacify as it should.

I have tried the 'profile' part of your autopilot but I can't seem to get it to save a profile so that I can 'one-click' it for craft that I build. Which, if it worked , would negate having to have them saved as part of the file.

Hm... need to think about it. The whole framework of TCA configuration is based on the idea that it's part of the ship. To the extent, that docking/undocking preserves configuration and grouping of TCA instances.

So it is possible to, say, attach a new core with newer TCA configuration to a station and get all the benefits.

Edited September 1, 2020 by allista

[Daveroski]

8 minutes ago, allista said:

It was modified based on dynamic pressure down to about 5 deg in lower atmo, but then the limit was 10. Though the actual AoA was rarely larger 5.

I just took her up again manually.

She flies herself.

https://imgur.com/gYLecFA

At launch tap east a couple of times just to move prograde in the right direction.

At 150mps set SAS to prograde.

Activate staging as required.

Wait until Pe is over 70K

Done.

And look at how much fuel remains.

D.

 

[Cap Arcona]

This mod is frickin' incredible. Unfortunately no matter what I do, TCA wants to point my ship into the ground unless I have an actual engine fitted. Some tiny craft are better with just RCS, so it renders those useless since the pilot is sitting on his head.

[allista]

9 hours ago, Daveroski said:

At launch tap east a couple of times just to move prograde in the right direction.

At 150mps set SAS to prograde.

Activate staging as required.

Wait until Pe is over 70K

Done.

And look at how much fuel remains.

Ok, l didn't try this particular approach, but I should be able to reproduce it with the new settings I've already introduced into ToOrbit.

Looking at the remaining dV I realize that the drag indeed eats much fuel.

I wonder what would following prograde do for heavy and powerful rocket with respect to fuel efficiency...

But there's one thing I didn't mention about the orbit I achieved: it has 5.1deg inclination, witch also costs some fuel, both as is and for continuous correction.

Even more, inclination correction requires more than 2-3 deg deviation from prograde to be efficient.

[allista]

3 hours ago, Cap Arcona said:

This mod is frickin' incredible. Unfortunately no matter what I do, TCA wants to point my ship into the ground unless I have an actual engine fitted. Some tiny craft are better with just RCS, so it renders those useless since the pilot is sitting on his head.

Well, RCS are not engines, they are controlled differently, so, indeed, TCA cannot handle them as such. But there is a monopropellant-fueled engine (Vernier it's called, if I'm not mistaken) with the same Isp as conventual RCS thrusters.

[Daveroski]

1 hour ago, allista said:

But there's one thing I didn't mention about the orbit I achieved: it has 5.1deg inclination, witch also costs some fuel, both as is and for continuous correction.

Even more, inclination correction requires more than 2-3 deg deviation from prograde to be efficient.

I just did a test for you. The method is sound. The variable is the couple of taps eastward at the start.

I made a 3.5m single hull with nothing but a nose cone, Large probe core, Large Reaction Wheel (2.5) Large battery (A circle of little solar panels around the Reaction wheel Couldn't help myself.)

Two large 3.5 stock tanks. (Told the bottom tank to drain first) and the Big 3.5 Thrusters on the bottom.

No Wings, tail fins or any other gubbins.

But with this big lad, a couple of taps still takes it almost vertical all the way. So not enough taps methinks.

So I turned it east  about 7 degrees right after launch. While it's speed was still too slow for shear to matter.

At 155mps (same as with Little Squark) I told the SAS to go prograde.

When It got higher than 70K I cut throttle and plotted the orbit. Had plenty of time to do it. When I was done with the map I still had 40s left before the burn.

Got it to orbit with an inclination error of 0.3 degrees.

If you turn it so that it can still ascend after the turn even though there is still a little shear, once it hits 155mps It is going for prograde and very little drag all the way to orbit.

D.

 

[JadeOfMaar]

2 hours ago, allista said:

But there is a monopropellant-fueled engine (Vernier it's called, if I'm not mistaken)

The 'Puff' engine, actually. The vernier engine is the RCS thruster that uses LFO.