SimCube build Part 1

February 15, 2018 / Comments (0)

Building Tips

/ By Gabriel Accascina

I started building this simulator (‘SimCube’, as it looks like a cube on the outside), in March 2017. I am convinced that, to be effective, simulators used in actual training need to have cockpit dimensions very close to the aircraft they represent and should be fully enclosed. So my design criteria included side doors with windows, a windshield, all in a cabin that was about 46″ wide and about 52″ tall at the center. Picture the size of a Bonanza or a Seneca. I also wanted to be able to replicate either a single or a twin and no specific aircraft per se. While I am working on another aircraft-specific simulator, this particular project was more directed to create an environment realistic enough to be able to learn to fly the avionics correctly.

The previous year I had purchased various aviation parts from eBay and I wanted to put them to work. I was also training on the G1000, so being the winter what it is in upstate New York, I thought a simulator would give me a great platform to twist knobs, and dial in routes and approaches, when I was not flying a real airplane. In retrospective, a sim surely does keep you proficient on instruments. The basic structural foundation for the sim was a sheet of 3/4″ plywood, cut in half as a 4′ x’ 4′ square, approximately 1.20m square. I cut the two front corners at an  angle to facilitate reaching the inner components and wiring later on.

Linear pot on pitch axis

The main components I had bought were a pedal/brake set from a Cessna 310, the yoke assembly from an old Mooney and a throttle quadrant from a Seneca. All told, about $200 in 2016. I started by modifying the brake pedals, essentially adding a basic mechanism which would make the two pairs move as they do on the plane. One forward makes the other come backward, on both sides. I linked the pedals to springs so that they would reset toward the center when not pressed and offer a slight “feel”  of resistance. I then connected the pedal left/right movement to one 4″ long 100K linear pot and the two brake pedals control to two additional identical pots. When not pressed, the pedal’s pot is centered (neutral yaw), and the two brake pots are at the end of their travel. Wired it all up (3 wires each) and connected it to a Bodnar BU0836X . This versatile USB interface does a lot for controls, switches, encoders, etc. and it’s virtually foolproof plug and play. Alternatively, an Arduino is almost as good. I then used shelving perforated angle steel to build the basic structure of the sim, supporting the instrument panel and, eventually, the front windshield and the roof of the cabin. Once the basic panel frame was in place, I started working on adapting the Mooney yoke assembly to the frame. It worked quite well with a few adapting brackets and some threaded rods to support the front of the assembly. The yoke movement was immediately quite smooth and I used relatively slack bungees to return the yoke to center. As of today, I would build the entire dual yoke from scratch instead, and possibly add force feedback. I will post an article about it soon. To sense motion, roll and pitch, I used essentially the same technique, linear pots activated by an arm linked to each relative movement and eventually connected to ports on the Bodnar.

Throttle connection to rotary pots

So far I was about 40 hours of work into the project. I secured the throttle quadrant to more angle steel in a center pedestal fashion and I linked (some adapting needed here), each one of the six levers to small RC model clevis arms (eBay) and then to six rotary pots’ each with 2.5″ levers attached to them. The feel was good and the friction adjustment which came with the original quadrant makes it even better. I set the prop lever to start about 1″ higher up to take advantage of the feather detent in the quadrant. Now it was the turn of the elevator trim wheel (I never got around to do the rudder trim) which uses a multi turn 50K pot. I bought one with three turns, very precise, but I would now buy one with just one or maximum two turns. All those connections ended up, again, on the Bodnar.

I’ll end this here for now, with more to come for panel, cabin enclosure and, eventually, settings, visuals and sound, computer hardware and starting it all up.


Comment(Comments are visible only to logged-in Members)

Leave a Reply