Rover reversed controls on the Mun? (0.25 Linux)

[fommil]

Hi all,

I built my first rover and tested it out on Kerbin and everything worked great.

However, when I delivered it to the Mun[1], the steering controls were reversed! i.e. rolling left actually rolled right. The weird thing is that the animation showed the parts moving correctly in response to my joystick movements, so it felt like I'd encountered a bug. Has anybody else seen this? Also, the landed navball was the wrong way round, when I was pointing East it was saying West. Increasing the throttle moved the rover in the same direction as it did on Kerbin.

(The general control of the rover was much lower than on Kerbin, I suspect that is to be expected since the gravity is much lower.)

Another thing that bugged me was that when I built my rover for testing on Kerbin, I was able to place the autonomous command module *inside* the rover casing. But when I was building the rover as part of a larger structure, I was only able to place the command module outside (i.e. on top of) the casing. Is that normal? Is there any way to get it to go inside? It gives me more space to place instruments if the command module goes outside.

I'm happy to share parts files if it helps anybody answer the above.

It was a lot of fun driving the rover on the Mun :-D

Best regards,

Sam

[1]: I put it on the underside of a very small lander that had radial monopropellant engines, big legs and a junior stack decoupler, allowing a safe delivery on touchdown)

Edited November 11, 2014 by fommil

[Claw]

A .craft file or some pictures would help out.

However, I'm going to guess that what you're running into with the reversed controls is due to control part orientation (based on your later complaint about attaching parts). When you separate a smaller vehicle from a larger one, your steering orientation can get transferred to a pod or probe core that you didn't expect. You can right-click on the correct probe core/pod/docking port and select "control from here." That might fix your steering orientation problem. That's assuming the part was oriented the correct way when you built it.

As for when you're building things, many people will build their rovers as a sub-assembly. Then stick it onto your rocket when you're all ready to go (that makes it easier to test before hand). To do that, you have to build off of some other root part first such as a cubic strut. The sub assembly tab does not let you drop an entire craft on it. Unfortunately, because of the tree-like nature of KSP's craft, sometimes when building things this way you are limited in how you can attach things when building "backwards." This should be significantly easier in v0.26/0.90 (since they are reworking the editors), but that doesn't help you much now...

It'll be easier to give specific advice if you can supply at least a picture.

Cheers,

~Claw

[Pecan]

Control orientation seconded here. For rovers 'forward' should be towards the horizon so if the navball isn't pointing that way your control-pod or -probe needs to be turned around. Claw did all the rest anyway.

[fommil]

A .craft file or some pictures would help out.

I've uploaded my craft to here: http://www.filedropper.com/lunar-rover

I've never used filedropper before, but I was able to download and md5 matched my upload, so I have some trust in it.

I use quite a few mods, such as deadly reentry, ferram and various science component additions. On my setup, I'm able to fly to the Mun with this design no problem. I go to the radial monopropellant as late as I can to conserve fuel. It's probably way over specced because everything below the rover was designed to take an Apollo style Kerbal to the Mun, deploy, rendezvous, and back (and succeeded, which was utterly awesome).

However, I'm going to guess that what you're running into with the reversed controls is due to control part orientation (based on your later complaint about attaching parts). When you separate a smaller vehicle from a larger one, your steering orientation can get transferred to a pod or probe core that you didn't expect. You can right-click on the correct probe core/pod/docking port and select "control from here." That might fix your steering orientation problem. That's assuming the part was oriented the correct way when you built it.

I think I know what you mean --- I've encountered a similar problem before when attaching sub-vehicles via docking ports (especially when their control room is upside-down with respect to the main launch vehicle's controller). It would be a real pain to have to select the correct control room on the launch pad.

But in this instance you'll see I only have the one (autonomous) control room so that is not a problem. If you delete everything below the rover itself and deploy on the launchpad, the controls are as expected.

Re: placing the octo inside the rover, I look forward to 0.26 as that might allow me to do this!

Thanks for looking at this.

[Claw]

I've uploaded my craft to here: http://www.filedropper.com/lunar-rover

...

I use quite a few mods, such as deadly reentry, ferram and various science component additions.

Thanks for uploading. Unfortunately if we don't know what all doodads you have attached to that craft, it's not going to open. If you're using mods, your next best bet is to load up a few pictures (to some place like imgur.com, which doesn't require an account). Then post the link here.

If you're unfamiliar with dealing with screenshots in KSP, you can check this post: http://forum.kerbalspaceprogram.com/threads/92230-Stock-Support-Bug-Reporting-Guide?p=1380977&viewfull=1#post1380977

Specifically look for the spoiler that says "screenshots."

(Sorry for the back and forth...)

Cheers,

~Claw

[fommil]

Thanks for uploading. Unfortunately if we don't know what all doodads you have attached to that craft, it's not going to open. If you're using mods, your next best bet is to load up a few pictures (to some place like imgur.com, which doesn't require an account). Then post the link here.

If you're unfamiliar with dealing with screenshots in KSP, you can check this post: http://forum.kerbalspaceprogram.com/threads/92230-Stock-Support-Bug-Reporting-Guide?p=1380977&viewfull=1#post1380977

Specifically look for the spoiler that says "screenshots."

(Sorry for the back and forth...)

Cheers,

~Claw

Can you see this https://lh6.googleusercontent.com/-gEta6iMxIFE/VF6yEelvMoI/AAAAAAAAiWo/TcdzJas9el4/w1565-h879-no/rover-deployment.jpg

It's a bit of an early draft of the same design: the one I flew had actual side mounted monopropellant engines and a smaller decoupler.

[Brainlord Mesomorph]

Control orientation seconded here. For rovers 'forward' should be towards the horizon so if the navball isn't pointing that way your control-pod or -probe needs to be turned around. Claw did all the rest anyway.

I add a probe core to the front of the rover and use "Control from here" to fix the NavBall

and I just figured this out about my first rover:

In low-G the SAS torque was greater than the wheel torque . If the controls are in "staging" mode, pressing "W" to move forward also makes the SAS pitch down.

Make sure you are in Docking mode controls, and even then, not in rotation mode, but translation mode.

Was that it?

[Padishar]

In low-G the SAS torque was greater than the wheel torque . If the controls are in "staging" mode, pressing "W" to move forward also makes the SAS pitch down.

Make sure you are in Docking mode controls, and even then, not in rotation mode, but translation mode.

A better way (in my opinion) is to set up the rover driving controls separate to the rotation and RCS translation controls (I use the cursor keys for rover driving). Then you don't get any unwanted forces from your reaction wheels or RCS while driving and can still use them in an emergency, i.e. when you go flying over the top of a crater rim and need to rotate to ensure, or at least increase the likelihood of, a safe landing...

[DrMonte]

What I might suspect related to driving:

Forces of SAS rotation might be much higher than forces of the wheels steering. That is why your wheels rotates to the correct side but the rover to the other.

Local solution would be to use tweak-able and invert steering on wheels. Those having inverted steering everywhere.

Otherwise redesign in SHP is required. Simply placed command module points to the sky, those not very helpful for the rover.