Rudder (the mechanical part)

 

As promised, I’ve finally started working on the rudder pedal project. Here’s a screenshot from X-Plane showing what this post is about:

As I mentioned in an earlier post, I already have a set of rudder pedals that came with my CH Products yoke. 

However, I’ve never really been happy with how they feel or how accurate they are. That’s probably partly because they were designed for desktop use. To keep them compact, they use a sliding mechanism instead of hanging pedals. That same compact design also makes them surprisingly difficult to integrate into the cockpit on the motion platform.

So I decided to design and build a completely new set of rudder pedals from scratch, based on the hanging pedal mechanism used in the real aircraft.

The first step was figuring out exactly where the new pedals could be mounted. There were two things to consider. First, ergonomics. As you may remember, the seats in my cockpit are fixed and can’t be adjusted forward or backward. Second, the mechanical constraints. The biggest issue is that the cockpit can’t extend much further forward because it would run into the wall of my room. Moving the whole simulator further back isn’t an option either, since my wife’s desk is directly behind it. The cockpit sits on a 1.5 × 1.5 m base, and if at all possible, I’d like to stay within those limits.

So I screwed a first prototype pedal onto a temporary wooden frame and clamped it into various positions inside the cockpit. I then climbed into the pilot’s seat and tried each setup, all while doing my best not to accidentally snap the rather wobbly construction. 

That turned out to be more challenging than expected because the cockpit doesn’t have a floor yet. Testing different pedal positions became a bit of a workout, with my legs alternately stretched out and bent, trying not to bump into the temporary pedals too much, while also having nowhere to rest my feet in between.

 

After a few rounds of trial and error, I found a position that not only looked and felt right, but also allowed for a nice, straightforward mounting solution. The pedals could be attached directly to the existing crossbeam at the front of the simulator.

Back in CAD, I was able to take the “PIPER” pedal I’d previously modeled freehand from reference photos and turn it into a complete hanging pedal assembly. From there, I designed the entire rudder pedal system, including the linkage connecting both pedals.

The Pedal Assembly 

Just like the real thing, the pedal is not only supposed to swing forward and backward on the hanging pedal arm. It also has a second pivot point that acts as the brake pedal when you push the upper part of the pedal forward with your toes.

I actually worked out the basic mechanism and proportions directly from screenshots in X-Plane, since that let me inspect the setup from all kinds of angles and even watch parts of it in motion.

Of course, I won’t be adding an actual brake cylinder or rudder linkage, so that’s where the similarities end. I want all movement sensing to be as precise and wear-free as possible, so I decided to use analogue Hall effect sensors for all pedal movements.

Instead of a brake cylinder, my model therefore has a strong magnet mounted to the brake pedal linkage, inside the orange part shown in the image above. Its movement will later be measured by a Hall effect sensor and passed on to the controller.

But all of those electronic details are still work in progress and will be covered in a future blog post.

The pedal arms are made from 16 mm aluminium tubing, which finally gave me an excuse to fire up my lathe again after quite a long break.

To drill the holes in the tubing, I designed and 3D-printed a set of drill guides. They made it easy to drill the cross holes, including the ones offset by 90 degrees, even without access to a milling machine or a rotary table.

The brake pedal pivot includes adjustable end stops. These make it possible to fine-tune the pedal’s neutral position very precisely and, at the other end of the travel, prevent the PLA pedal body itself from acting as the mechanical stop.

Instead, there’s a TPU insert fitted into the back of the pedal, which serves as the contact point for the end-stop screws. In my experience, TPU is much tougher than PLA, and if it ever does wear out, the insert can easily be replaced without having to reprint the entire pedal.

To make the pedal look as close to the original as possible, and even feel somewhat similar, I printed the foot pad from black rubber… or rather, what the 3D printing world calls rubber: TPU. 

I really tried to match the original shape, but after several failed attempts to print the slightly curved surface, I eventually had to give up. The staircase effect you get when printing shallow curves always looked much worse than the flat version I ended up using, which could at least be printed face-down on the textured build plate for a much nicer finish.

One alternative might have been to print the part standing upright on its narrow edge, but printing such a tall, slender part in TPU is asking for trouble.

As a small consolation, I did manage to design the pedal and foot pad with a working snap-fit connection, so at least they can be assembled without any glue, screws, or other hardware.

The Pedal Linkage

As mentioned earlier, the hanging pedal assemblies are mounted directly to the front crossbeam of the cockpit. That leaves just one final piece of the puzzle: the linkage connecting the two pedals.

In my design, the two pedals are connected by a central rocker, which is linked to the hanging pedal arms via ball joints.

The pedals are automatically returned to the center by two additional arms, one for each side, that are pulled downward by extension springs. The exact center position is defined by separate adjustable end stops on each arm.

This part of the design is still very much in the prototype stage.

The biggest issue so far has been the ball joints, which tend to bind at larger deflections depending on their position. Unlike the original CAD model, I’ve changed the real-world prototype so the ball joints are mounted horizontally. That solves some of the problems, but not all of them.

I’m also still not sure whether this linkage can remain entirely plastic in the long run. So far it seems to work reasonably well with fairly beefy parts, but only time will tell how they hold up under continuous use, especially once the simulator is in regular operation.

Fortunately, these components are relatively simple, so if necessary it should be straightforward to remake them from plywood or perhaps even aluminium in the future.

In the photo above, you can see the return springs as well as the adjustable end stops that define the center position. There’s also a temporary Hall effect sensor mounted in place (under the white tape), which detects the movement of a magnet attached to the central rocker.

That’s where the project stands at the moment. 

The CAD models are publicly available on Onshape if you’d like to take a look. Fair warning though: this is very much work in progress, and I haven’t exactly spent a lot of time or effort keeping the feature tree tidy or giving everything sensible names.

Below is a video showing the assembly of one of the pedals:

To be continued…