Engineering a Desk-Mounted Flight Stick for Flight Simulation

[Klondiked]

TL;DR Summary

Trying to emulate real-life flight stick movement from a horizontal desk mount; i.e. lateral movement rather than angular movement.

Hi,

First off, full disclosure as mentioned in the title, I am not an engineer, but I'm working on an engineering project without the required base knowledge.

I'm a professional pilot that also likes to fly flight simulators for fun (mostly military planes, because I'm not allowed to fly upside down and blow things up IRL). I especially love the IL-2 series, which simulates WWII era planes. However, one thing that drives me nuts is that unless you build a custom sim pit, PC flight sticks are just joysticks with a rotation point at about the desk level, or occasionally slightly lower through an extension with a clamp. That, at least, allows the stick to be positioned in the right place, which is between your legs.

But in a real aircraft, with few exceptions, flight sticks are mounted at floor level, or possibly slightly underneath the floor level. So in practice, the stick moves laterally, with minimal angular change. I'm trying to mimic that motion by attaching the stick to a desk, with a horizontal extension, rather than from the floor; my goal is to have the stick move laterally in the X and Y planes, about 2-3 inches either side of center, in a square pattern. I don't need to simulate the subtle angular change. It would be easy enough to accomplish this with a floor mount, but I'm specifically trying to figure out how to do this from a desk mount.

The reason I'm posting here is because I don't want to wrack my brain trying to reinvent the wheel, if there's a Mechanical Engineering 101 solution to this that I just don't know about. If there isn't an obvious solution, I'd be happy to know that, too, and I'll go back to wracking my brain. After all, I have time now.

And lastly, I recognize that this is just half the problem; if I have problems with the other half, I'll bring it up at the Arduino forum. Thanks in advance!

*EDIT*

If this isn't the right type of forum for a question like this, please don't hesitate to tell me so.

 

[anorlunda]

I don't blame you. I'm a pilot too, and joysticks are nothing like the real thing.

Have you studied the stick mechanism on some small planes? You could reverse engineer that pretty easy.

Have you considered getting a real stick and floor mechanism from a junkyard, then adapting it as input to your simulator? I would be surprised if someone hasn't already done that before. Have you asked around on the flight simulator forums?

 

[sysprog]

Mostly the PC joysticks are basically XY cog set pairs with the angular orientations being rendered as averages.

 

[Lnewqban]

Welcome, Klondiked

A x-y plotter working in reverse comes to mind.
Something like this:

https://www.instructables.com/id/X-Y-Plotter/

 

[Klondiked]

@ anorlunda, there are actually quite a few options for the floor mount, including commercially available seats that allow you to bolt the joystick base underneath the seat itself, and a high-end joystick also has optional extensions which really do a pretty good job of imitating the actual aircraft. However, I don't really have the space available to create a custom setup like that, so I'm trying to make something that can bolt onto a desk and take up minimal space. I'm struggling to come up with a compact compromise.

@sysprog, I've actually successfully made a joystick with Arduino hardware and code, but really just as an exercise. The little thumbstick that I used is totally useless for this application, but I think I can apply that code to any potentiometer pair.

@Lnewqban, that's kind of along the track I was thinking of. All you'd have to do is extend one of the plotter arms and attach the stick to it. The only drawback is that if I want stick travel of 2-3 inches, the base assembly is going to be at least 4-6 inches square, but that's the best option I've thought of so far. I briefly saw a 2 axle linkage mentioned, and that was my first thought, as well. The problem with that is that full aft displacement with lateral center is going to be farther back than full aft with maximum lateral displacement. That's what I was really hoping to overcome, because it allows me to make a base that's much smaller and simpler. A 2 axle linkage with arms that could extend and retract is soooo close to what I want, but I don't see a way to make that move in a square pattern.

I even thought about using a gear like we used to use in the spirographs to make a (mostly) square, but I couldn't see a way to access any points inside the square. I'm probably going to try to come up with something like the x-y plotter and see how that works.

Thanks, all, for you input.

 

[Lnewqban]

Then, it would be nice to add simulated air resistance from the control surfaces.
Even nicer, to make that resistance proportional to airspeed, mimic stalls, etc.

Please, see:
https://en.wikipedia.org/wiki/Aircraft_flight_control_system

 

[sysprog]

@sysprog, I've actually successfully made a joystick with Arduino hardware and code, but really just as an exercise. The little thumbstick that I used is totally useless for this application, but I think I can apply that code to any potentiometer pair.

This may be a bit out of date, but I think something like this might suit your requirements -- it uses a standard game port: https://www.ebay.com/itm/Microsoft-Sidewinder-Precision-Pro-Joystick-X03-57540-/223044020414

 

[Tom.G]

Here is a sketch for the mechanics. Sensing is left as an exercise for the reader.

The 'Stick' is mounted on a plate. The plate has eight surface-mount ball bearings, four on top and four on bottom. An alternative could be Nylon slides.

The plate with Stick slides freely around in a thin box. The box dimensions and opening size it the top define the travel range.

Cheers,
Tom

 

[Lnewqban]

But that arrangement does not eliminate rotation of the stick.
How to sense the separate movements for elevator and ailerons and communicating those to the computer?
Some kind of anchoring to the desk will be needed, since you will be imposing moments and sliding forces on the base, things that articulated yokes don't suffer.

 

[Klondiked]

So right now, I'm using the Virpil MongoosT-50 with their VPC desk mount V2. It's a great stick, great base, great mount, and it's even available with a longer extension to simulate a floor mount, but I was hoping to build something much more compact. I may have to look for another project that's a bit more realistic for me. Thanks for all your time, folks.

 

[hutchphd]

It is not clear to me what you want. If you wish to move a small XY table with the joystick then get some linear potentiometers (sliders) (often you want 5k) and build what you want. Wire them into an old style pot based joystick like the sidewinder shown above .
Of course the newer joysticks use optical encoders either analog (reflectance from a graduated gray target) or digital (pulses through an optical gate) so they will be harder to kluge.

 

[Tom.G]

But that arrangement does not eliminate rotation of the stick.

Yea, I thought of that while waking up this morning. Came up with a fix, I think, but it got a bit involved.

How to sense the separate movements for elevator and ailerons and communicating those to the computer?

Knowing nothing about flying airplanes, I have no idea. Please enlighten me.

It looks like the OP has decided that he is better off following a different route. Whew!

Cheers,
Tom

 

[Lnewqban]

...
Knowing nothing about flying airplanes, I have no idea. Please enlighten me.

Sorry, Tom, I was not paying proper attention and though I was responding to an OP's post, who is familiar with the terms.
Regular flight simulators and radio control transmitters have two separate angular sensors (rotating potentiometers) that are simultaneously manipulated by a common lever.

One senses pilot's inputs for rolling the airplane sise to side via ailerons, which are control surfaces located by the trailing outer edge of each wing (A surfaces in animation of post #6).

The other senses pilot's inputs for pitching the airplane nose up and nose down via elevators, which are control surfaces located by the trailing edge of the horizontal stabilator (C surfaces).

 

[sysprog]

My favorite thing to do with the MS Flight Simulator (this was way before 9/11) was to take off in a Beechcraft from Meigs Field, and fly around downtown Chicago and slam it sideways or at whichever angle into the Hancock or the Sears Tower -- the heck with learning how to (reliably) land it properly -- back then that idea was just a game thing.

 

[Klondiked]

sysprog, I remember those days. I got my first taste of aviation as a teenager flying the MicroProse Flight Simulator on a Commodore 64 and I used to do exactly that. I haven't the faintest idea if that was a Beechcraft or not, but I remember the wireframe buildings, and the frame rate was pretty dang close to 1 frame per second. But at the time, it was amazing. I even went so far as to learn how to land it and navigate with ground based radio aids.

Even though I fly airplanes for real now, I still really enjoy the simulators, and I'm frequently frustrated that the people with the resources to make accurate, dedicated controllers keep doing it so wrong. I halfway wonder if it's worth getting a second degree in engineering just to have a crack at doing it right, but that seems like a lot of time, work, and money just to play video games.

 

[Klondiked]

Also, it appears that a picture is worth a thousand words in this case, so I got this random picture off the internet of a Russian WWII fighter cockpit. You can see that the stick is attached to the floor, and at full deflection off to one side, the handle is barely angled, and it will travel several inches before it reaches the other side. I was looking for a desk-mounted solution that would imitate that movement.

 

[anorlunda]

My favorite thing to do with the MS Flight Simulator (this was way before 9/11) was to take off in a Beechcraft from Meigs Field, and fly around downtown Chicago and slam it sideways or at whichever angle into the Hancock or the Sears Tower

Man those days were fun. I loved the Cessna 172 from Meigs Field and heading straight for Sears Tower. If I did everything right, I could gain just enough altitude to clear the roof by 100'.

Another fun challenge was to land the Lear Jet on the USS Nimitz. Remember, the Lear has no tail hook. I tried that for a whole year; hundreds or thousands of tries. Eventually I did it once.

My son and I used to compete trying to fly the 747 inverted through the hangar. Same thing, after hundreds of tries, only failure. But one day, my son showed me a replay where he succeded. I was amazed and impressed. But I later learned he cheated. He flew the Sopwith Camel inverted, recorded it, then he manipulated the recording to substitute 747 for Camel during the replay. Smart kid.

Thanks for the reminder of those fun days. No computer game today is as fun as that was. But I hear great things about the coming MS Flight Simulator 2020. Hopefully, that will make it fun again. I want to use a sim to jump off The Matterhorn in a wing suit.Edit: @Klondiked , Apologies, we are getting a bit off topic. If you object, I can fix it. Click on my user name and start a conversation to let me know.

 

[NTL2009]

Also, it appears that a picture is worth a thousand words in this case, so I got this random picture off the internet of a Russian WWII fighter cockpit. You can see that the stick is attached to the floor, and at full deflection off to one side, the handle is barely angled, and it will travel several inches before it reaches the other side. I was looking for a desk-mounted solution that would imitate that movement.

I'm on my tablet, so not easy to draw any thing, but it seems to me that you could use two inverted domes, with a radius matching the floor to stick length. The larger dome would be mounted on the table, a smaller one, with stick attached would ride on top. I guess you would need a third one with a hole in the center to restrain the one with the stick.