sarbian

[1.7.x] Anatid Robotics / MuMech - MechJeb - Autopilot - [2.8.4] [14 June 2019]

Recommended Posts

The problem exists only for the first stage (?).

Right after launch the MJ fails with reaching the adjusted inclination (i.e. 300 degrees = 60 degrees = -60 degrees, etc.).

When the first stage decouples, it is enough to enter correct value in the MJ inclination field, and it starts to works fine.

Share this post


Link to post
Share on other sites

Hi folks, I have a quick question for those of you using MechJeb as i am trying to find the best process to launch a ship from Kerbin to Duna using MechJeb.

What i am doing:

1. On the launchpad, enter in map mode, select Duna as Target

2. On the Ascent guidance module, click on "Launch at interplanetary window", click on "engage autopilot". Mechjeb warps to the launch window and liftoff.

3. Once in Orbit, I click on "Maneuver planner" and "Transfer to another planet"

My problem: The transfer maneuver sets a node more than 1year in the future!! The warp takes forever as the engine needs to use full simulation.

What am i doing wrong?

Share this post


Link to post
Share on other sites
11 minutes ago, SPD13 said:

Hi folks, I have a quick question for those of you using MechJeb as i am trying to find the best process to launch a ship from Kerbin to Duna using MechJeb.

What i am doing:

1. On the launchpad, enter in map mode, select Duna as Target

2. On the Ascent guidance module, click on "Launch at interplanetary window", click on "engage autopilot". Mechjeb warps to the launch window and liftoff.

3. Once in Orbit, I click on "Maneuver planner" and "Transfer to another planet"

My problem: The transfer maneuver sets a node more than 1year in the future!! The warp takes forever as the engine needs to use full simulation.

What am i doing wrong?

Turn off autowarp for a bit, time warp one year, turn it back on.

 

If that doesn't work, delete the node, time warp, and try again.

Share this post


Link to post
Share on other sites

Thank you for your reply, but my problem is: why the node has to be 1year in the future despite the fact i already wait for the interplanetary window before launch?

Share this post


Link to post
Share on other sites
Just now, SPD13 said:

Thank you for your reply, but my problem is: why the node has to be 1year in the future despite the fact i already wait for the interplanetary window before launch?

Getting there requires some time.

Share this post


Link to post
Share on other sites

edit2: Ok, so something weird is happening; even though the ascent guidance says -6 for inclination, it keeps launching into a +6 inclination orbit, thereby of course resulting in a large plane change burn for minmus. I have manually set the leading angle to 0, the inclination is properly set to -6, then for some inexplicable reason it keeps launching into a +6 inclination orbit. Perhaps it's an issue with directions on the device that's controlling the craft...

edit3: tried many different control points and it's still launching into the mirrored inclination. It appears to be happening at the circularization burn... It launches into the correct plane, then at the circularization burn, it inexplicably creates a node that mirrors the inclination into the wrong plane.

 

Edited by drhay53

Share this post


Link to post
Share on other sites
6 hours ago, drhay53 said:

edit2: Ok, so something weird is happening; even though the ascent guidance says -6 for inclination, it keeps launching into a +6 inclination orbit, thereby of course resulting in a large plane change burn for minmus. I have manually set the leading angle to 0, the inclination is properly set to -6, then for some inexplicable reason it keeps launching into a +6 inclination orbit. Perhaps it's an issue with directions on the device that's controlling the craft...

edit3: tried many different control points and it's still launching into the mirrored inclination. It appears to be happening at the circularization burn... It launches into the correct plane, then at the circularization burn, it inexplicably creates a node that mirrors the inclination into the wrong plane.

When you say Ascent Guidance launches into a -6 inclination orbit, does the rocket aim south of due east and increase in South Latitude?  That's what I see AG do.  -inclinations incline to the south and +inclinations incline to the north.  The orbit of the rocket will always be reported with a positive inclination from 0 to 180deg, as the difference between the orbit from AG having -inclination or +inclination is where the Ascending Node is.

Did some testing of the target features of AG.  Didn't bother with the phase angle (what I assume you meant as the leading angle, seen both used) as the test rocket was just an orbital one.

Once I set Minmus as the target, I'm offered two new buttons in AG: Launch to Rendezvous and Launch to Plane of Target.  LtR has a countdown clock that you can use autowarp to get to launch time and will launch the rocket.  But as far as I can tell, LtR only worries about phase and doesn't bother with launching near one of the Nodes to get a co-planar rocket orbit.  Alternately, LtP only worries about the plane and sets the orbit inclination appropriately (6deg for Minmus, which appears correct), assuming the sign -/+ is selected for the next node crossing.  But LtP has no countdown clock and you have to determine manually when the launch site is just about at the next Node, which is hard.

Share this post


Link to post
Share on other sites
1 minute ago, Jacke said:

When you say Ascent Guidance launches into a -6 inclination orbit, does the rocket aim south of due east and increase in South Latitude?  That's what I see AG do.  -inclinations incline to the south and +inclinations incline to the north.  The orbit of the rocket will always be reported with a positive inclination from 0 to 180deg, as the difference between the orbit from AG having -inclination or +inclination is where the Ascending Node is.

Did some testing of the target features of AG.  Didn't bother with the phase angle (what I assume you meant as the leading angle, seen both used) as the test rocket was just an orbital one.

Once I set Minmus as the target, I'm offered two new buttons in AG: Launch to Rendezvous and Launch to Plane of Target.  LtR has a countdown clock that you can use autowarp to get to launch time and will launch the rocket.  But as far as I can tell, LtR only worries about phase and doesn't bother with launching near one of the Nodes to get a co-planar rocket orbit.  Alternately, LtP only worries about the plane and sets the orbit inclination appropriately (6deg for Minmus, which appears correct), assuming the sign -/+ is selected for the next node crossing.  But LtP has no countdown clock and you have to determine manually when the launch site is just about at the next Node, which is hard.

I have now seen MJ do this inclination mirroring in other scenarios, but it doesn't happen all the time. For instance, I had a polar satellite that I tried to launch into plane; it correctly launched due south, but as soon as it created the circularization burn, it was for a complete flip of the orbit (burn retro at apoapsis to flip the inclination). However, I just did a launch to minmus where it worked fine. I can't find the rhyme or reason there. 

Launch into plane usually does give a timer, but as established from recent posts and my experiences dating back at least to May 2016, it does not properly calculate the angle (I do think it is a leading angle, and not an angle based on any absolute frame of reference). I have occasionally seen it not give any countdown timer, but I also don't know what's causing that. 

My diagnosis is that ascent guidance is quite broken when using launch into plane. 

Share this post


Link to post
Share on other sites
24 minutes ago, drhay53 said:

My diagnosis is that ascent guidance is quite broken when using launch into plane. 

I'd say It's not complete as it doesn't trigger the launch.  I was trying to use (Orbit) Longitude of the Ascending Node (LAN) for the rocket on the pad versus Target LAN to gauge when the pad passed through the relative Nodes with respect to the target, but from judging by eye when it passes through a Node versus the angular difference between LAN and Target LAN being 0 or 180deg, they aren't the same.  As a work around, judging by eye and using the Launch-to-Plane inclination on a regular Ascent Guidance will probably be good enough.

Launch to Rendezvous seems to be aimed at near co-planar multiple launches from the same pad, at least for a near-equator pad like KSC.  It judges the phase angle to get that first close pass soon.  And it does automatically launch.

Log files from a couple of games with several attempts at using both LtR and LtP, here and here.

Edited by Jacke

Share this post


Link to post
Share on other sites
1 minute ago, Jacke said:

I'd say It's not complete as it doesn't trigger the launch.  I was trying to use (Orbit) Longitude of the Ascending Node (LAN) for the rocket on the pad versus Target LAN to gauge when the pad passed through the relative Nodes with respect to the target, but from judging by eye when it passes through a Node versus the angular difference between LAN and Target LAN being 0 or 180deg, they aren't the same.  As a work around, judging by eye and using the Launch-to-Plane inclination on a regular Ascent Guidance will probably be good enough.

Launch to Rendezvous seems to be aimed at near co-planar multiple launches from the same pad, at least for a near-equator pad like KSC.  It judges the phase angle to get that first close pass soon.  And it does automatically launch.

Launch into plane automatically launches just fine most of the time; you still have to click 'engage autopilot'

Share this post


Link to post
Share on other sites
1 hour ago, drhay53 said:

Launch into plane automatically launches just fine most of the time; you still have to click 'engage autopilot'

I played around and Launch to Plane of Target finally always gave me a countdown.  And pressing Engage Autopilot got it to launch automatically after its countdown.

But it appears Ascent Guidance is miscalculating when best to launch to put the rocket into the plane of the orbit of the target, at least when aiming at Minmus.  Appears to be often 45 to 60 degrees early. 4 runs of Launch to Plane autopilot gave me relative inclinations of final orbit with respect to Minmus's orbit (the target) of 4.56, 6.93, 7.75, and 2.55 degrees.   Using map view and the equatorial orbit of Mun to manually judge when at the Descending Node to launch, with an inclination of -6deg, using the regular autopilot gave me a relative inclination of 0.34 degrees.  Logs of the 2 games here and here.

Share this post


Link to post
Share on other sites

Thus far, every time the circularization inclination-mirror bug has happened, it's occuring under the following scenario:

- Set launch into plane angle to 0 (probably not relevant). Click launch into plane. If the inclination is negative, the circularization burn will be incorrect, and will burn in such a way as to mirror the desired inclination. 

Share this post


Link to post
Share on other sites

While I was doing my testing, I set up my flight windows for MechJeb for KSP 1.3.0 and I'm happy to see improvements.  Old issues solved and many new features I'll have to check out.  Thanks to @sarbian and all who brought these about.

 

4 hours ago, drhay53 said:

Thus far, every time the circularization inclination-mirror bug has happened....

I time-warped several times and saw the inclination sign update by itself without any input from the player, which is good.  But it flips sign for the upcoming Node between about 60 to 80 degrees before the true closest alignment with the Node.  Seeing as it also launches between about 45 to 60 or 70 degrees before the true alignment and I'm not surprised it can sometimes launch with the wrong sign.

I don't think any more comment is needed here about this part of Ascent Guidance.  Until @sarbian or someone else can dig into the code and address this, there's a manual workaround.

1.  First use Launch to the Plane to get the size of the launch inclination.
2.  Use map view and Mun's orbit as a projection of the equator, see the proper sign for that inclination: positive for an Ascending Node, negative for a Descending Node.
3.  Manually judge the launch site alignment with the Node.
4.  Launch with the regular autopilot of Ascent Guidance.

 

Share this post


Link to post
Share on other sites
12 minutes ago, Jacke said:

While I was doing my testing, I set up my flight windows for MechJeb for KSP 1.3.0 and I'm happy to see improvements.  Old issues solved and many new features I'll have to check out.  Thanks to @sarbian and all who brought these about.

 

I time-warped several times and saw the inclination sign update by itself without any input from the player, which is good.  But it flips sign for the upcoming Node between about 60 to 80 degrees before the true closest alignment with the Node.  Seeing as it also launches between about 45 to 60 or 70 degrees before the true alignment and I'm not surprised it can sometimes launch with the wrong sign.

I don't think any more comment is needed here about this part of Ascent Guidance.  Until @sarbian or someone else can dig into the code and address this, there's a manual workaround.

1.  First use Launch to the Plane to get the size of the launch inclination.
2.  Use map view and Mun's orbit as a projection of the equator, see the proper sign for that inclination: positive for an Ascending Node, negative for a Descending Node.
3.  Manually judge the launch site alignment with the Node.
4.  Launch with the regular autopilot of Ascent Guidance.

 

I'm not sure if you understand the particular bug that I'm talking about, because this method is not a workaround. The issue is:

1) MJ launches with the correct inclination, at the correct time

2) After reaching the target apoapsis and coasting to space, MJ creates a circularization burn that mirrors the inclination. i.e., if the inclination at launch was -6, instead of circularizing at -6, MJ circularizes to +6, doing a 12 degree inclination change along with circularization at the same time. 

My experience is that setting the launch into plane angle to 0 always works to launch at the right time, meaning that the bug is in the calculation of how long to lead the line of nodes, but this is a totally separate bug from the "circularization inclination flip" that I'm seeing when the initial launch inclination is negative. 

Edited by drhay53

Share this post


Link to post
Share on other sites
42 minutes ago, HaydenTheKing said:

Mech Jeb not installing

 

Mod list: https://imgur.com/a/XgYnj

Output Log: https://www.dropbox.com/s/vdy13u70137aptf/output_log.txt?dl=0

PC Hardware: I7 4790k, 32gb DDR3, GTX 1080TI

Make sure you are loading 2.6.1.0 on KSP 1.3.0 install... Looks like the hashes match.  Do you have mechjeb for everyone or something else that adds the mechjeb module to the pods and cores you are using?

Edited by Bornholio

Share this post


Link to post
Share on other sites
1 hour ago, HaydenTheKing said:

It is a 1.3 instal

ok duplicated what you are seeing in that install, i'll add an edit when i get a solution to it. But i'm guessing its just a simple module manager MJ module adder problem. Double check if you can see it in Sandbox VAB.

If you want to force MJ to unlock early in career.

 

Edited by Bornholio

Share this post


Link to post
Share on other sites

@HaydenTheKing I think this might be the problem:

Quote

MechJeb moduleRegistry creation threw an exception in LoadComputerModules loading IRSurfaceSampler, Version=1.0.1.5, Culture=neutral, PublicKeyToken=null: System.Reflection.ReflectionTypeLoadException: The classes in the module cannot be loaded.
  at (wrapper managed-to-native) System.Reflection.Assembly:GetTypes (bool)
  at System.Reflection.Assembly.GetTypes () [0x00000] in <filename unknown>:0 
  at MuMech.MechJebCore.LoadComputerModules () [0x00000] in <filename unknown>:0 
  

Get the latest version of surface sampler from the link in this post:

https://forum.kerbalspaceprogram.com/index.php?/topic/104535-112-magic-smoke-industries-infernal-robotics-202/&do=findComment&comment=3140574

 

Share this post


Link to post
Share on other sites
1 hour ago, HaydenTheKing said:

Mech Jeb not installing

 

Mod list: https://imgur.com/a/XgYnj

Output Log: https://www.dropbox.com/s/vdy13u70137aptf/output_log.txt?dl=0

PC Hardware: I7 4790k, 32gb DDR3, GTX 1080TI

In "MagicSmokeIndustries/Parts/Rework_Utility/Plugins/" you have a IRSurfaceSampler plugin that seems to be outdated. It creates a bunch of exception in your log and is the most likely source of your MJ problem.

Share this post


Link to post
Share on other sites
5 hours ago, sarbian said:

In "MagicSmokeIndustries/Parts/Rework_Utility/Plugins/" you have a IRSurfaceSampler plugin that seems to be outdated. It creates a bunch of exception in your log and is the most likely source of your MJ problem.

 

5 hours ago, Aelfhe1m said:

@HaydenTheKing I think this might be the problem:

Get the latest version of surface sampler from the link in this post:

https://forum.kerbalspaceprogram.com/index.php?/topic/104535-112-magic-smoke-industries-infernal-robotics-202/&do=findComment&comment=3140574

 

 

updated and got new errors. https://www.dropbox.com/s/xtflij8blm0luy2/output_log.txt?dl=0

uninstalled infernal robotics and still getting problems https://www.dropbox.com/s/cq777p3rancmetn/output_log2.txt?dl=0

 

 

Edited by HaydenTheKing

Share this post


Link to post
Share on other sites
1 hour ago, HaydenTheKing said:

updated and got new errors.

I'm not seeing any MJ related errors or exceptions in that new log . What are the symptoms you're getting?

Share this post


Link to post
Share on other sites
7 hours ago, Bornholio said:

ok duplicated what you are seeing in that install, i'll add an edit when i get a solution to it. But i'm guessing its just a simple module manager MJ module adder problem. Double check if you can see it in Sandbox VAB.

If you want to force MJ to unlock early in career.

 

been testing in sand box

4 minutes ago, Aelfhe1m said:

I'm not seeing any MJ related errors or exceptions in that new log . What are the symptoms you're getting?

Mech jeb not showing up in the game, i see nothing on tool bar, nothing in cockpit.  shouldn't there be something?

Edited by HaydenTheKing

Share this post


Link to post
Share on other sites
50 minutes ago, HaydenTheKing said:

Mech jeb not showing up in the game, i see nothing on tool bar, nothing in cockpit.  shouldn't there be something?

The MechJeb file paths in the log all look correct and the AR202 is being compiled OK so that shouldn't be a problem.

Just to get the obvious out of the way first - are you seeing the "MechJeb 2 (AR202 case)" in the command and control category in the VAB/SPH and are you adding one to your vessel?

I'm not seeing any ModuleManager patches from a mod like MechJeb for Everyone without which none of the MechJeb buttons/windows will be visible without an AR202 attached.

Edited by Aelfhe1m

Share this post


Link to post
Share on other sites
1 hour ago, Aelfhe1m said:

The MechJeb file paths in the log all look correct and the AR202 is being compiled OK so that shouldn't be a problem.

Just to get the obvious out of the way first - are you seeing the "MechJeb 2 (AR202 case)" in the command and control category in the VAB/SPH and are you adding one to your vessel?

I'm not seeing any ModuleManager patches from a mod like MechJeb for Everyone without which none of the MechJeb buttons/windows will be visible without an AR202 attached.

its there now!

The RPM window has only one page of options, is that how its supposed to be?

Edited by HaydenTheKing

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.