That's a very nice piece of work, especially for the methods used.

If you want to make a front panel without going out and buying a large format laser printer, there's Front Panel Express. All they do is make metal plates with holes and lettering. The lettering is engraved and powder coated. I've used their service, and they do a nice job.

In Europe that would be Schaeffer in Germany. They do the complete housing as well. Very nice.

In fact it seems they were involved in setting up Front Panel Express: http://www.schaeffer-ag.de/en/company/company_history/

Yes, if you try to visit the Front Panel Express site from Europe, you are redirected to Schaeffer's site with the same service.

I've used Front Panel Express's services as well - +1 - they really do a nice job.

That's a really cool project.

The choice of joystick is disappointing though - an arcade stick? Those are digital only, usually either 4- or 8-way. Kerbal would be better controlled with a real analog joystick. But I guess the game probably doesn't support analog input.

It probably does, but to the best of my knowledge, no one makes anything analog that you could use as a three-axis translation controller without some serious rework. A digital device is much easier to work with.

Also of note: while I've never played Kerbal, years ago I was very seriously into Orbiter, which is a somewhat more rigorous simulation. Especially when maneuvering in close proximity to another vehicle, you actually wanted to avoid analog control of RCS thrust, because that's hard to control precisely and it's very easy to find small errors compounding and your ship getting away from you. So you'd put the RCS in what I want to say I remember being called "vernier" mode, where each control input regardless of magnitude or duration produces a specific, calibrated quantity of thrust, which can be precisely counteracted by the opposite control input. So at this point, you're effectively treating an analog control device as digital - which works, but something like the three-axis controller detailed here would actually be a lot more comfortable to use for vernier control. If I still played Orbiter at all, I'd be asking OP for more details on the parts he used and how he put them together, so I could build something similar for my own use.

> It probably does, but to the best of my knowledge, no one makes anything analog that you could use as a three-axis translation controller

Huh? There are plenty of twist-handle joysticks that give you 3 analog axes. Some will even give you 5 or 7 (One or two mini-sticks)

No one is disputing that there are analog joysticks with three or more axes; the matter under discussion is how to get the right axes.

The idea behind a controller like the one under discussion here is that its three axes are, first, all translational, and second, all in direct correspondence with the motion they impart to the spacecraft. With such a device, no mental mapping between control axes and motion is required. You just push the controller in the direction in which you want to add thrust.

If it's still unclear what that means, look at this diagram [1] of the Apollo translation controller. Disregard the limited rotational capability for which I have been unable to find documentation. Observe the translation axis diagram alongside the handle. Note the three degrees of freedom it depicts. Consider their correspondence with the three degrees of freedom in which a spacecraft may experience translational motion.

To provide a matching capability, the joysticks you're talking about would need to allow for pushing the handle into, or lifting it out of, the base. I doubt anyone makes a joystick like that, because it would not be a very useful capability to add, and the engineering requirements and effect on retail price would be significant.

[1] http://nassp.sourceforge.net/wiki/Image:THC_diagram.jpg

It does support analog input, but using analog input only marginally improves gameplay.

Analog input usually isn't that important because the inertia of the thing you're rotating typically overpowers your system's ability to generate torque by a fair amount. So fine-grained control of orientation/rotation isn't as critical as being able to turn at all.

It's listed as a 3-axis translation joystick, so (I assume) it's being used for RCS control in space, of which digital switches would be better, for accurate control.

No, it supports analog input.

I have to say, though, if you're building a custom digital controller, it's usually easier to just crack open a cheap USB keyboard and repurpose the PCB. They are small, support lots of buttons, and already have drivers on your PC.

Or use this device https://www.adafruit.com/product/1535 It lets you put 12 switches on it and then acts like a keyboard. I've used it for a couple of custom input projects and it works well, nothing to do on the PC.

Using a digital joystick would then be the way to go. I like his choice of that input device.

But, like Quequau, I've never been able to get into Kerbal. I guess I could go whip a panel up, but at the end of the day not sure that it would be enough to catch my interest.

Yeah, that works if you've planned your project out and junk like that :) I'm usually just slapping junk together at the last minute and cannibalising whatever I have laying around (and for some reason, I have a lot of keyboards laying around).

If you were writing the consuming app, you could get analog support with an ADC into a decoder into individual or combinations of keys. I say you'd have to be writing the app because I don't know many existing apps that support that sort of key binding (though I do seem to recall MS Flight Simulator doing it, at least for throttle).

Yea, how I fail the "True Maker test". I tend to plan stuff out, dig through a pretty well stocked set of parts. Then China markets for parts for projects that I can wait on, Digikey for a project I want to build tomorrow.

I do refactor parts (I've done the keyboard thing mentioned above) but now since I'm older I realize I don't have the time to dork around with a dodgy keyboard (ummm is that the reason I tossed it into the junk bin?) and want things to work after I fix my design errors, not after I fix a broken part.

I'm more a pragmatic Maker. Can I run to Home Despot and buy wood, fasters, plan, cut, assemble and install shelves? Yes. Can I go to Ikea, buy something close and make it fit that then looks really nice? Yes. With 400 linear feet of shelving in the robot lab, sometimes Ikea wins.

Likewise, some nifty Adafruit board is the quicker way to win.

Now that's a clever idea, never thought of that.

The pin-out on every PCB is different. I've even seen two apparently identical keyboards have completely different pin-outs. But it's not very hard to figure out the mapping. Just open up a text editor and test out combining two different pins. Usually, there are three "long" pins on the ends of the PCB and about 30 short pins in the middle. You combine a long and a short pin to make a key code. This is where N-key rollover problems come from, so if you want your controller to support multiple button presses, you'll have to be creative in your selection, which means you will only be able to use it in apps where you have arbitrary control over key bindings.

Ironically, the controller itself looks more sturdy than the wobbly, jello ships in KSP proper (doesn't help that Unity's physics engine chokes on KSP's complexity).

To be fair, I'm not sure there is a physics engine that can handle KSP in real time.

Here's another mirror, as Google's cache doesn't display the images. https://archive.fo/bRaGJ

A guy at our makerspace in Sydney built a couple times larger system with working seven segment speed & fuel indicators, two joysticks, etc. That was last year, I think the project kept growing since.

Here's a demo video from last year's Makerfaire: https://www.youtube.com/watch?v=uCuNyiseJ5g&feature=youtu.be...

edit: grammar

I love that solid hardware. It reminds me of this unique Xbox game that came with an extremely large controller: https://en.wikipedia.org/wiki/Steel_Battalion -- 2 joysticks and 40 buttons. $200 for the game and controller.

I never could get into KSP. I really love this project, though I have the feeling I'd spend the time, energy, and effort building something similar and get right back to the point where I just can't get into the game.

Their osx support killed it for me. I don't have a gaming computer anymore, and trying to run it on my mbp ran into nothing but terrible performance and constant crashes. I would switch to the playstation version, but I don't want to buy the game again, when I already purchased it and it doesn't run well.

When was the last time you tried it?

The OS X version is still a bit slow to load, but it's far better than it was a year or two ago. I get basically no crashes and fairly reasonable performance these days, so it might be worth another look.

I get performance that's largely indistinguishable from my desktop - although, I will admit that I only played on my mbp maybe 5% of my total play-time.

About a year ago. I'll give it another try.

What about the game did you find lacking?

It's not about something being lacking. I adore KSP as well, but i can't get myself to play it. Just not my slice of cake.

well answered!

Google cache, since the site seems to be buckling under load: https://webcache.googleusercontent.com/search?q=cache:JEozcO...

Unfortunately that doesn't pick up the images which are half of the content here.

The images appear to load for me.

What I mean is that the images are still being served from the original server. They load eventually, but very slowly for me.

I think I saw something similar in one of the Linux magazines in Barnes and Noble like a couple of years ago. I wonder if they're the same thing. Still a cool project.

Thats amazing! Love the design, programming can be improved I think.

on one hand, i would say 'some people have too much free time'

on the other hand, i DO say 'i know what i am gonna spend my weekend on'

> on one hand, i would say 'some people have too much free time'

I'd say, that's why free time is awesome, and we need much more of it.