When it comes to detecting movement/positional data I found a pretty easy method that works a treat - light. Sure it is defendant on having a sealed enclosure for the sensor itself, but let me explain a little more..

A work colleague recently had a dilemma with his flight sim flight control grip that he is building (an F16 cockpit) - they are a high G model with limited movement, about 10mm if that full scale deflection for either axis. Here's the replica;



The real life grip adapter utilities a near inflexible force transducer to provide X/Y positional data. It also costs a lot of $$$. A cheaper DIY approach was made and it worked a charm!

Here's the underside of the DIY adapter;


A high grade aeronautical silicon applied to the upper side of the adaptor (not shown) ensures incredibly tough movement resistance and high tension springs (shown above) are loaded just before the tension point allow quick/accurate returns to zero upon release. From there, a high brightness LED is mounted within the square tube, providing a bright light shadow that directs down. The sensor consisted of two LDR's that were heavily masked to provide a means for X and Y data capture;


You might see how this is working already, if not then read on.

The light pattern is directed down, and biased so that it lay directly in the center of either sensor. From there, the limited movement of the control grip allows the light pattern to move from one side of either LDR to the other without physically reaching the edges.

The square sensor and light pattern ensures that if the Y axis is moved directly up/down, the X axis will maintain the same amount of "light". The same concept applies for X axis movement. Here's the light pattern and sensors at work with the side panel off (it is completely enclosed during normal operation);


The program for this application goes into quite a few good little features such as detecting the min/max voltages, setting Vrefs to ensure max resolution and bubble sorted sampling along with various other controls being handled at the same time. I find it takes away from the concept at hand which I thought was share worthy!

Sure the results are non linear, but that too is accounted for within the program. On top of that, most HOTAS applications will allow joystick calibration etc to suit the characteristics of different aircraft. The completed grip assembly;

__________________
digital-diy.com - Hobby microcontroller projects and tutorials. Assembly, PICBasic and C examples.

 

What the heck are you building, a backyard jet fighter? That thing looks like it's bomb resistant =)

__________________

 

Hi Graham,

Saw this on your site. WILD! - no that's not an acronym

I'd love access to his sim.

Perhaps McDonnell Douglass would be interested?

EDIT: Whoops, that would be Lockheed Martin.

It was the F15 that McDonnell Douglass did.

 

It is a little sturdy - may as well do it right the first time around!

I'm in awe of some of the other avionics sim packages he has self-developed. Complete, functional CNI panels & displays. We're both avionics technicians though, which does help out

__________________
digital-diy.com - Hobby microcontroller projects and tutorials. Assembly, PICBasic and C examples.

 

Post some links to what you guys are talking about, I'm curious.
You'll get no complaints from me gramo, I'm a big fan of overbuilding.

__________________

 

Check out some of these:
Google