Jump to content

[PART, 1.0.2] Anatid Robotics / MuMech - MechJeb - Autopilot - Historical thread

r4m0n

By r4m0n
in KSP1 Mod Releases

 Share

Recommended Posts

Meithan

Meithan

Posted
Posted

Hi. Is this the place to report bugs? Last night I was getting incorrect delta-v values. Here's a picture of the test rocket:

UqE4MwOl.jpg

As you can see, the staging detected by MJ is a bit weird. And the total delta-v reported is off by a factor of ~two. Computing the delta-v manually, I get somewhere between 2100 (atmo) and 2500 m/s (vacuum). And indeed, when launching that rocket, after I run out of fuel MJ reports a "delta-v expended" of 2310 m/s. What's weirder is that the burntime seems to be correct; it's just the delta-v figure that's off.

I redownloaded MJ's last version just to be sure, and it still calculates it incorrectly.

Thanks.

Link to comment
Share on other sites
SHaFT7

SHaFT7

Posted
Posted

I've had a recent development problem. I built a fresh rocket, and upon ascent guidance launch, it tries to turn one loop before getting back underway correctly. My rocket had enough thrust to get out of it, but still it was a new thing. I tried several things and from what I've found, when I launch now, my north (red line on navball) is facing down instead of up. switching various "control from here" doesn't work. I have a pod, a probe core, and docking port and switching between them does nothing. I find the real issue is that mechjeb is trying to reorient the rocket to north-facing on the navball, and instead of just rotating the vessel, it loops it to get it right, wasting tons of fuel in the process.

I did find that if I go back into the VAB and rotate the entire rocket 180 deg and then relaunch, then everything is correct.

Just to make sure I didn't spontaneously forget how to build rockets, I loaded up the kerbal X and just put a mechjeb ar202 on it and launched it. Same thing. North facing down and it tries to loop on launch (except the Kerbal X doesn't have the TWR to pull out and it crashes)

Any ideas on what is going on?

Link to comment
Share on other sites
sarbian

sarbian

Posted
Posted

Iasus : open the custom window editor for the Delta-V window, go to the thrust category and add the last one "dV include cosine losses".

This option change the way MJ compute dV when engine don't point back.

Meithan : your root part pod is the Remote Guidance Unit under the decoupler. If you change the root part to on of the top pod you'll get the correct result. You can do it with the "SelectRoot" mod.

With this part as the root part the MJ dV simulation want to decouple the top at the start. The displayed stats are what you would get with only the tank, engine + RGU.

SHaFT7: is this with v2.1 or the dev version ? Try the dev version and tell me if it's any better.

Link to comment
Share on other sites
Meithan

Meithan

Posted
Posted
Meithan : your root part pod is the Remote Guidance Unit under the decoupler. If you change the root part to on of the top pod you'll get the correct result. You can do it with the "SelectRoot" mod.

With this part as the root part the MJ dV simulation want to decouple the top at the start. The displayed stats are what you would get with only the tank, engine + RGU.

Thanks a bunch for the reply, definitely makes sense. This was just a test rocket; the top pods were to simulate 15 tons of weight. But it's good to know what conditions can set MJ off.

Link to comment
Share on other sites
SHaFT7

SHaFT7

Posted
Posted
SHaFT7: is this with v2.1 or the dev version ? Try the dev version and tell me if it's any better.

I was running 2.1. Installed dev #122 and that seems to have fixed it. thanks!

one other thing, the MJ settings window in the VAB is right over the parts list on the upper left. While I can make it go away by clicking the MJ button on the right (which incidentally is completely empty, there is no menu) I wonder if there is a way to move that settings window?

Link to comment
Share on other sites
sarbian

sarbian

Posted
Posted

Sorry I forgot to commit part of my last patch. I just did it and it should fix things.

With this version window will use different saved position for the editor and in flight.

Link to comment
Share on other sites
plausse

plausse

Posted
Posted

Maybe this is a dumb question.. I am aware MJ causes a fine "stutter"... which it always did for me. But suddenly in 0.22 it started causing major lag to the point of going from an fps of ~40 to an fps ~15. Even on small vessels and with few or no other mods. Normally I run loads of mods.. and its ok. Any thoughts? Thanks

(Now I have just resorted to piloting manually.. which is ok.. but my precision just isnt quite MJ standard)

Link to comment
Share on other sites
sarbian

sarbian

Posted
Posted

A few user reported that but I do not have enough info to look into it since it works fine for me.

First did you try the dev build ? I made a performance fix 1/2 weeks ago. I don't think it's related to your problem but it may help.

Second is there any errors in the debug log ( the one in KSP_DATA/output.log ) ? Large lag is often related to error spam in the log.

Link to comment
Share on other sites
Iasus

Iasus

Posted
Posted
Iasus : open the custom window editor for the Delta-V window, go to the thrust category and add the last one "dV include cosine losses".

This option change the way MJ compute dV when engine don't point back.

Sorted. Thanks.

Ias

Link to comment
Share on other sites
sarbian

sarbian

Posted
Posted (edited)

XanderTek & DerekL1963 : done. It was a bit more complex than anticipated. If anyone get some error with the warphelper in this version please report.

Edit : And I added a "Warp to Phase Angle with Target"

Edit bis : and added a new "Resonant Orbit" option in the Manoeuver Planner. It puts you in a new orbit having an orbital period equal to fraction of your original orbit. So you can launch a ship with 3 satellites attached, go to a synchronous orbit and ask MJ to put you in a 2/3 resonant orbit. Then if you launch a sat each time you are at the AP they'll be at a 120° angle from each other.

It would be more efficient to start from the lower orbit but it's all you get for now. Those who want to it properly can make a copy of my google doc spreadsheet with all my satellite math

Edited by sarbian
Link to comment
Share on other sites
softweir

softweir

Posted
Posted

Yes, it does work in career start. The AR202 module (with basic functionality) appears in an early node, then various more advanced functions provided are unlocked in further nodes. If you have already unlocked those nodes then you need to view the tech tree, click on nodes, and in the panel on the right select the Mechjeb modules/functions and research them.

Link to comment
Share on other sites
Hatsune_Neko

Hatsune_Neko

Posted
Posted
Yes, it does work in career start. The AR202 module (with basic functionality) appears in an early node, then various more advanced functions provided are unlocked in further nodes. If you have already unlocked those nodes then you need to view the tech tree, click on nodes, and in the panel on the right select the Mechjeb modules/functions and research them.

Thank you

=(^.^)=

Link to comment
Share on other sites
DerekL1963

DerekL1963

Posted
Posted

OK, upgraded to the 123 build this AM and found a minor oddity in Smart A.S.S.. When I select ORBT/Normal +, it orients the vehicle to the proper attitude, but unlike 102 (which I was upgrading from) it leaves the vehicle with a roll rate.

(And thanks for warp-to-time!)

Link to comment
Share on other sites
XanderTek

XanderTek

Posted
Posted
XanderTek & DerekL1963 : done. It was a bit more complex than anticipated. If anyone get some error with the warphelper in this version please report.

Edit : And I added a "Warp to Phase Angle with Target"

Wow, that's a fast response! Thanks!

Link to comment
Share on other sites
skips

skips

Posted
Posted

I would like to point out that the pointing specific directions in Smart ASS only constrains the pitch and yaw orientations of the vehicle. The roll is completely unconstrained by the pointing requirement. It is formally correct that forcing one of those pointings should have no affect on the roll orientation or stability. That said it might be useful if Smart ASS also implemented a set of roll constraints. (i.e., inerital, sun-pointing, main-body pointing, ...) real spacecraft often have constraints such as these implemented in the safe-hold programs.

skips

Link to comment
Share on other sites
rosenkranz

rosenkranz

Posted
Posted

Having trouble with landings. Normally without FAR I can land on a dime just using the landing marker from the landing AP. But with FAR, I absolutlely suck at hitting a target. Is there any way for MJ to talk to FAR to at least ball park a landing trajectory? Barring that, any tips or tricks to landing with FAR? (wrong thread for that I know)

Link to comment
Share on other sites
rosenkranz

rosenkranz

Posted
Posted

yyeaaaaah, i had a feeling. I just need to figure out a methodology that let me hit a target with some consistency.

I remember hating docking until I finally learned how to use the nav ball properly. I see this is going to be much the same.

Link to comment
Share on other sites
Awesome_Guy

Awesome_Guy

Posted
Posted

so ive just come across this problem

i have been using mechjeb fine up until now to set up my first sat relay system

the first 5 were fine, now the problems start

when i engage the auto pilot the throttle jumps to about 50% and vibrates there and is un-responsive

this is happens if i manually put the craft into space and try use the maneuver node

any info would be great.

PS i have reinstalled mechjeb 2 times

Link to comment
Share on other sites
Guest
This topic is now closed to further replies.
 Share
Go to topic listing
×
×
  • Create New...