The Boeing 737 brake lever is located in the throttle quadrant. In my cockpit it is part of the mechanics panel that is also home to the landing gear switch, the flaps lever and other assorted controls.
While the airbrake lever is operated manually during decent, a servo will move it automatically to the “Airbrakes Extended” position upon touchdown. I wanted to repicate this behaviour in my cockpit and partially employed the concepts used to build my autothrottle – hence the similarities between the two designs..
The general idea is that an electrical motor and gearbox move the lever when needed. Separation of gearbox and lever (for manual operation) is done with an electromagnetic clutch. The whole system is controlled by EPIC code.
Here is a Corel drawing of V.1 of the auto brake. Since the first version had some significant design flaws, I soon had to build V.2.
V.1 used 4 micro switches that allowed only four different lever positions and states of the airbrake system. The micro switches were actuated by a spring loaded steel ball, taken from a ball bearing. The steel ball moved inside a casing attached to the lever. The state of each micro switch was monitored by the EPIC card.
It soon became evident that the 4 positions of the lever did not allow precise control over the planes airbrakes. As a matter of fact, all they offered were the positions “Down”, “Arm”, “Flt Up” (rather arbitrary) and “Full Up”. Besides, the steel ball tended to get stuck at in-between positions, resulting in anything but smooth lever action and severe strain on the motor.
V.2 incorporates two major changes: Analogue control over the spoiler axis with a hal sensor, and a complete redesign of the lever travel.
Now a spring loaded ball bearing rolls over a specially shaped, curved surface with indents that produces tactile feedback for the various lever positions. The movement of the lever is now silky smooth. It stops in the places where it is supposed to, but can be moved to the next detent without much effort.
Since I was at it, I decided to add backlighting for the airbrake scale.
The lever gearbox seen from the side.
The red box represents the electric motor. It came with an integrated 148:1 reduction gearbox. With the different gear wheels a further RPM reduction is achieved, so that the lever – when driven by the servo – travels at roughly 50 degrees per second.
ear wheels are in different shades of green (nylon) or purple (steel). All axis are 6mm silversteel, mounted in ball bearings on both sides.
A view from above.
The drawing might look confusing, but there is no way of building something like this without detailed planning in advance. Without these drawings, it would not have been possible to construct or buid this system. Even when I had all the pieces in front of me, ready for assembly, I often needed to consult the drawings in order to remember which part goes where.
Since I am not an engineer or CAD specialist, a number of things still went wrong, no matter how hard I tried to foresee all angles of my construction. Luckily, it was just some minor adjustments here and there. The basic design made it to prototype status without much alteration.
The gearbox walls are made from 3mm plywood. It is light and can easily be worked on with simple tools. The box obtains its strength from the L-shaped aluminum profiles that hold it together at the edges.
As a first step, I extracted the gearbox wall dimensions and positions of holes from the above drawings and printed them on paper. The printouts were then glued on the plywood. I found this to be a simple way of making sure that all holes would be drilled at their exact positions. I made sure the holes were a bit smaller that required, so there was room for minor adjustments during assembly.
Three walls of the gear box are joined, and the electric motor is in place.
The spacers that hold the axis in place are made from nylon on a lathe.
The clutch assembly.
The left gearwheel rotates freely on the axle, with the help of two miniature ball bearings. The armature, which is nothing but a 3mm metal disk, is bolted to this gear wheel. When the clutch is activated, this disk is pulled against the electro-magnetic coil of the stator. This establishes a firm mechanical connection between the two gears on this axis, and consequently, between the electric motor and the airbrake lever.
The nylon gear wheel on the right is fixed to the axle with the help of a nail (rather crude, I admit…).
The guide for the lever, made from several layers of transparent, 3mm acrylic glass. Note the irregular indents on the lower side. Here is where the ball bearing of the plunger rolls along. Since the plunger is spring loaded, there is clear tactile feedbeck when the top of the ball bearing sinks into any of the indents.
The two claw-shaped spaces will later hold microswitches that sense when the lever has reached its max. or min. position.
Three Spider LEDs are already glued into the openings in the upper part of the guide.
The lever housing, fully assembled.
The plunger guide, with the spring-loaded plunger inside. This piece is bolted to the airbrake lever. I used a bit of grease around the plunger to ensure extra smooth movement.
With the plunger housing bolted to it, the lever is installed on the 1/4 section of a 120-teeth gear wheel. On this photo it becomes clear that it is actually the plunger housing (and not the lever) that actuates the position limiting micro switches.
This picture is actually taken during the assembly of V.1, but it shows beautifully how all parts fit together.
The first two gear wheels (motor and first transmission gear) are made from steel, the rest is nylon. All are Module 1, so they are rather big and sturdy. That makes them easy to work on, and they can transmit quite a bit of torque. Also, Module 1 gear wheels do not require utmost precision. A welcome thing for a home builder without access to sophisticated tools.
Since this is a prototype, some last minute adjustments were often unavoidable. So there are plenty of out-of-place holes in the walls – it happened several times that pieces did not fit as planned and had to be shifted around a bit.
But this box will disappear under the mechanics panel, and all it is supposed to do is to shift a lever back and forth. Once installed in the cockpit, nobody will know – or care – how it looks…
The hal sensor with it’s linkage to the lever axis. I prefer hal sensors over potentiometers because they are more linear, more precise, produce less spikes and – since they are contactless – last much longer.