Jump to content

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

r4m0n

By r4m0n
in KSP1 Mod Releases

 Share

Recommended Posts

drtedastro

drtedastro

Posted
Posted

Hello to all,

I am running dev build # 425. I am having 'problems' with "Reentry simulation timed out" happening now more than ever before.

I have built a small jet engine based hover pod, and it seems that the re-entry part of Landing Guidance does not like something to do with using the jet engine. My guess is that it does not like the spooling time required for the jet to come upto power...

This was not so much of an issue a while back. I don't remember what dev version I was last using where this was not happening so often...

is there something that I can do / set that will allow the reentry simulation to not time out so quickly???

many thanks.

Link to comment
Share on other sites
Gaalidas

Gaalidas

Posted
Posted

I'm no expert, but jet engines don't work very well outside of the atmosphere, so you'd likely only get any performance out of it after reentry. Doesn't much explain the error though.

Link to comment
Share on other sites
drtedastro

drtedastro

Posted
Posted

only using them in atmosphere...

after re-entry, and descent to lower altitudes, fire them up....

and try to use landing guidance same as with rocket engines.

all works 'fine' if adjust size and thrust so that all is almost perfect.. however, if any major course change is needed or such that needs a large change in thrust, the jets seem to take too long to spool up and mechjeb landing gets bored and the simulation times out and splat....

yup, jets would suck in a vacuum, IF they could suck in a vacuum.

Link to comment
Share on other sites
sarbian

sarbian

Posted
Posted

drtedastro : I don't think there was any recent change that would change the was the landing sim work with jet engines. The jet spooling is a long standing problem that The_Duck had a shoot at partially fixing last year and since then I don't thing anything changed.

The landing sim will most likely get quite a few code change and testing with 1.0 so you might have to wait until then.

Link to comment
Share on other sites
drtedastro

drtedastro

Posted
Posted

No problems...

Many thanks for this mod and all of your mods, sarbian....

I have found a work-around. I have replaced the single jet engine with 4 itty bitty scaled down versions and cut the power output way down. Now instead of all or nothing, the jet engines seem to stay at about 30% and then tweak up and down as needed... A red neck version of a power clutch..... But hey, its working.

Again, many thanks for all updates and additions over the ages..

very much appreciated...

Link to comment
Share on other sites
smjjames

smjjames

Posted
Posted (edited)

How do I get it to stop doing this:

screenshot54_zpszl1jsx3u.png

? I would like MJ to stop wasting RCS with abandon, and it doesn't seem like it knows how to use strong RCS.

I do have the folder ModuleRCSFX, but if it's supposed to be an RCS fix, it's not fixing it. Those are the combo SAS+RCS from RoverDudes FTT that MechJeb is trying to use.

Edit: Whoop, didn't mean to double post.

Edited by smjjames
Link to comment
Share on other sites
sarbian

sarbian

Posted
Posted (edited)

It's done.

And it will do that with the balancer. I can't look at it for now.

Edit : I pushed a new build without the RCS fx resizing of the balancer. Not tested so I have no idea of the consequences (should be purely visual)

Edited by sarbian
Link to comment
Share on other sites
Starwaster

Starwaster

Posted
Posted
It's done.

And it will do that with the balancer. I can't look at it for now.

Edit : I pushed a new build without the RCS fx resizing of the balancer. Not tested so I have no idea of the consequences (should be purely visual)

Should be purely visual. Or... it could cause all life as we know it to stop instantaneously and every molecule in our bodies to explode at the speed of light.

Link to comment
Share on other sites
smjjames

smjjames

Posted
Posted (edited)
It's done.

And it will do that with the balancer. I can't look at it for now.

Edit : I pushed a new build without the RCS fx resizing of the balancer. Not tested so I have no idea of the consequences (should be purely visual)

*Puts on testing helmet* :D lol

Edit: Well, it's definetly not doing the blinding kamehameha thing from every RCS port anymore. As for conserving RCS, it seems to be doing that, yes. Though this ship is using a lot anyway due to being massive and because it's fighting an imbalance problem.

Edited by smjjames
Link to comment
Share on other sites
ctbram

ctbram

Posted
Posted

Is there a place to make bug reports for the dev versions of mechjeb? The last couple versions have been doing some very strange things.

I installed 425 yesterday and I would instruct it to execute a node and the ship would turn to the node then not autowarp and then just fly through the node and not fire the engine??? Then I updated to 426 today and I am getting the same behavior. For the last 5 or more versions the auto ascent module has been very bad. It launches and pitches to the west then corrects back to the east then porpoises wildly as it's pitching over by over rotating then trying to correct and over correcting and it goes back and forth. I end up with crazy orbits no where close to what I set it to.

I do clean installs be completely deleting old versions before installing new ones. I am going to roll back to the release version and if that one acts strange I am going back to like the 300's because it was pretty stable then has gotten more and more unstable in the 400 builds.

If there is any information I can provide to document the issues videos, config files, etc. please let me know and I will provide them.

Link to comment
Share on other sites
smjjames

smjjames

Posted
Posted
Is there a place to make bug reports for the dev versions of mechjeb? The last couple versions have been doing some very strange things.

I installed 425 yesterday and I would instruct it to execute a node and the ship would turn to the node then not autowarp and then just fly through the node and not fire the engine??? Then I updated to 426 today and I am getting the same behavior. For the last 5 or more versions the auto ascent module has been very bad. It launches and pitches to the west then corrects back to the east then porpoises wildly as it's pitching over by over rotating then trying to correct and over correcting and it goes back and forth. I end up with crazy orbits no where close to what I set it to.

I do clean installs be completely deleting old versions before installing new ones. I am going to roll back to the release version and if that one acts strange I am going back to like the 300's because it was pretty stable then has gotten more and more unstable in the 400 builds.

If there is any information I can provide to document the issues videos, config files, etc. please let me know and I will provide them.

I've been having the same problems with the ascent AP with the moving left and right horizontially, at first I thought maybe it was the design , and then FAR got updated and I had to wait for the FARext to get updated to know for sure, and now that it's updated, it's still having that problem.

Link to comment
Share on other sites
sarbian

sarbian

Posted
Posted (edited)

ctbram : here is the right place.

I'll have a look. IF you could share your save and ship it would help. You could also try #411 since #412 has a change that may be related to what you see.

Edited by sarbian
Link to comment
Share on other sites
Climberfx

Climberfx

Posted
Posted (edited)

Will MechJeb have support for Persistent Rotation? I ask because if i put on Mj Attitude, and it goes timeworn in between maneuver nodes, it keep stoping to fix node angle, it only work when i put on Stock Attitude. But Stock attitude with Mj won't control the rotation to the nodes very well, so i keep needing to change in between of the two settings.

If Mj recognize the Persistent Rotation, in some way, that can be solved.

I already ask for Persistent Rotation author the same, and will try to see how can he solve that. I recommend you two to talk each other, if already didn't. :)

So, can i count on you?

Cheers and thank you for your attention.

Edited by Climberfx
Link to comment
Share on other sites
veber

veber

Posted
Posted
It launches and pitches to the west then corrects back to the east then porpoises wildly as it's pitching over by over rotating then trying to correct and over correcting and it goes back and forth. I end up with crazy orbits no where close to what I set it to.

I have the same problem with KSP v0.90.0 & MechJeb starting with 413, 414 (including latest builds). Solved the problem switch off the option "Corrective steering" on the panel "Ascent Guidance". Previously, everything worked perfectly with the option "Corrective steering" (3## builds).

Link to comment
Share on other sites
SaintWacko

SaintWacko

Posted
Posted (edited)

Is anyone else running into issues with MechJeb and Karbonite? I'm running the latest dev version, and as soon as I place a Karbonite engine and a tank containing Karbonite on a vessel, the delta-v window goes blank, and Mechjeb begins spitting out NullReferenceExceptions in the debug panel, along with this error:

[Log]: MechJeb2 -> An element with the same key already exists in the dictionary.
[Log]: MechJeb2 ->   at System.Collections.Generic.Dictionary`2[System.Int32,System.Collections.Generic.HashSet`1[KerbalEngineer.VesselSimulator.PartSim]].Add (Int32 key, System.Collections.Generic.HashSet`1 value) [0x00000] in <filename unknown>:0   at KerbalEngineer.VesselSimulator.EngineSim.SetResourceDrains (System.Collections.Generic.List`1 allParts, System.Collections.Generic.List`1 allFuelLines, System.Collections.Generic.HashSet`1 drainingParts) [0x00000] in <filename unknown>:0 
  at KerbalEngineer.VesselSimulator.Simulation.UpdateResourceDrains () [0x00000] in <filename unknown>:0 
  at KerbalEngineer.VesselSimulator.Simulation.RunSimulation () [0x00000] in <filename unknown>:0 
  at KerbalEngineer.VesselSimulator.SimManager.RunSimulation (System.Object simObject) [0x00000] in <filename unknown>:0

Edit: I don't have this problem with the official release.

Edited by SaintWacko
Link to comment
Share on other sites
futrtrubl

futrtrubl

Posted
Posted (edited)

Found an issue with warp and timed ascent. When it gets close to the countdown time it comes out of warp, sees that it's a little early, goes into 2x physics warp but then never comes out of it. Needless to say even 2x physics warp can be pretty fatal with large launchers in the first moments of ascent.

Also posted on github.

Edited by futrtrubl
Link to comment
Share on other sites
smjjames

smjjames

Posted
Posted (edited)

It seems to be ignoring the 'use RCS for rotation' toggle during the landing guidance, and the 'use RCS for small maneuvers', like they're both on by default. I'm using a pretty big and heavy ship atm, so I haven't tested with stock RCS on a smaller lander.

Edit: I'm using the Honeybadger RCS+SAS+battery combo part on this ship.

Edit2: Anybody having severe problems landing huge and really heavy ships?

Edited by smjjames
Link to comment
Share on other sites
jmbailey2000

jmbailey2000

Posted
Posted

Edit2: Anybody having severe problems landing huge and really heavy ships?

Pretty much any ship......I went to build a base on the Mun and put a lander in a spot as the target location. No matter what, every lander acts the same. Set for target and let it rip. MJ orients properly, sometimes does a plane adjustment, warps to de-orbit burn point, comes out of warp, makes a minor correction and starts its descent. Up until it comes out of warp to make a correction, it is smack dab on target (within a few meters). But once out of warp it makes some type of adjustment and then start landing. At this point (post-warp, post-adjustment), MJ puts the lander down some 15-50km away from the target. Really strange. I've seen this behavior once in a while but it usually appeared to be something related to a very high crater wall that to me looked like it was interfering with proper readings. That was fine and understandable. This, however, has just really been weird. No idea when it started since up until now, most of my landings had been single craft.

Link to comment
Share on other sites
sarbian

sarbian

Posted
Posted

BARCLONE : side effect of using KER code. You get different numbers in the Editor but the same in flight. I won't fix for not (not sure not having the atmo number while vacum is that important) and I ll see what I do when I work on the 1.0 changes.

Link to comment
Share on other sites
smjjames

smjjames

Posted
Posted (edited)

Theres definetly something funky going on with the handling of RCS, while those are a different model from stock, they are the same strength as stock RCS and MJ has used them before without issues.

Edit: It's also really screwing itself up with whatever it's doing with the RCS, it's like theres two different autopilots going which are at crosspurposes.

screenshot8_zps3kkmzghj.png

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