Rotary switches  but not a switch !

      Rotary switch using a potmeter ...

 rotary "switch" assy
Introduction

Some of the rotary switches on the F-16 control panels use a 30° angle to mechanically move to the next position. These rotary switches are cheap and readily available in the local electronics shop. However, the "not standard" rotary switches use a different angle, for example 36° or 45°. These switches can be ordered for example from Greyhill, but those are not cheap ...
The "problem" becomes even worse if the rotary switch has a specific property like a momentary spring-loaded position. It means that you can rotate the knob to that position, but as soon as you release the knob, the rotary switch will automatically move back to the adjacent position. That kind of rotary switch is even harder to obtain and carries an in proportion price tag :-).
So, I propose a mechanical solution which is fairly simple, flexible and above all, cheap!

Electrical implications

Every position that a rotary switch can have implies a wire to the hardware interface so that the software can "read" each position. So, for a 6 position rotary switch you need (at least) 7 wires, and due to the "switch matrix" hardware probably 6 diodes.
The proposed solution does not use a rotary switch, but a  potentiometer  and some mechanical construction work to obtain the position "stops". The choice of a potmeter has a beneficial electrical interface implication. You do not need the digital inputs (and diodes) and a number of wires at least equal to the number of positions, but in principle just one wire which connects to an analog input. So the electrical wiring in the cockpit becomes simpler!

The resistance value of the potmeter is not critical, but a value between 2k and 10k is advised. The potmeter must be of the linear type. As we will use the potmeter only set to specific positions the value will have a discrete small range for each position. Likewise, when the potmeter is connected to an analog input, the voltage read by the sofware will also show specific small ranges for each physical position.

Depending on the length of the cable (location of the potmeter with respect to the analog input hardware) it may be wise to use a simple shielded wire as used for microphones. One center conductor with a braided shield is enough to connect a potmeter. Note that +5 Volt must also be wired to the potmeter, optionally decoupled with a small capacitor to ground (shield of cable).

Mechanical construction

Parts needed:

The construction is fairly simple; the "trick" is the disk and the spring to produce the mechanical feeling of a rotary switch. So, we leave that for last  hurray
 
As you can see, the potmeter is mounted on the "U"-shaped aluminum profile.
  1. Drill a hole through both "legs" with a diameter of the mounting part of the potmeter. This ensures that the potmeter and the hole for the shaft (to the knob on the panel) are neatly aligned.
  2. Demolish (yeah ...!) a second (old) potmeter which has an identical shaft diameter as the potmeter used in step 1. Of this disassembled potmeter we only need the mounting part. It will be mounted on the "leg" opposite to the "leg" that holds the potmeter and serves two purposes:
    1. guide the shaft of the potmeter, and
    2. enables mounting the entire construction on the F-16 panel.
  3. mechanical assembly The tricky part: construction of the disk. This is described in the separate section below.
  4. Assembly
    1. Slide the potmeter through the hole of the first "leg".
    2. Slide the locking nut of the potmeter on the shaft.
    3. Slide the position disk on the shaft of the potmeter.
    4. Slide a locking nut of the mounting part on the shaft.
    5. Slide the mounting part (of the "donor" potmeter housing) on the shaft.
    6. Put the mounting part in the hole of the second "leg" and screw the nut on it.
    7. Screw the nut on the "outside" of the mounting part (no need to tighten it).
    8. Tighten the nut of the potmeter; check that the shaft rotates smoothly.
    9. Install the spring with the ball-bearing ball.
    10. Rotate the potmeter (close) to the physical end position.
    11. Tighten the nut of the disk onto the shaft with the disk at appropriate "index" position!

Construction of the "position disk"

rotary switch position disk The "position disk" is just a plate of metal that can be mounted perpendical onto the shaft of the potmeter. As that may pose the first mechanical difficulty, I chose a brass toothed gear wheel which has a "hub" with a threaded screw to mount the gear wheel on an axis.

To create the "stop" positions of the rotary switch, I used a file to create index holes. First, I made a drawing on paper of the actual rotary switch positions (the red lines), and drew "helping lines" (the blue lines) between the switch positions. No metal should be removed at the helping lines, only at the "stop" positions. The curved black line gives an idea of the needed circumference.

To obtain the mechanical feeling of the rotary switch positions and the transition from one position to the next I made a "lever". That is a small arm attached at the "U"-shaped bracket (with a pivot point) and at the opposite side attached to a spring. I was so lucky to have a microswitch in the junk box that has a detachable arm of steel with at the end a small 'roller'. The steel provides the spring action.

Small rectangular pieces of aluminum (the blue boxes) are attached to the gear wheel to create end stops. Using the arm of the microswitch, the roller will bump against the rectangular piece of aluminum.

Details of the "FUEL QUANTITY SELECT" (FQS) rotary switch
FQS position disk
The FQS rotary switch has 6 positions: TEST, NORM, RSVR, INT WING, EXT WING, and EXT CENTER.
The TEST position is a momentary position. It is spring-loaded, so if you set the knob to that position, the knob will automatically "jump" back to the NORM position.
The momentary spring-loaded position can be realized by incrementing the circumference of the disk. With a gear wheel that is a bit difficult (you could add a plate mounted on the gear wheel). So, for "fixed" positions you can use a gear wheel, but if you need a momentary position, it is better to make a disk from a piece of aluminum and mount that aluminum disk onto a gear wheel (with slightly smaller diameter of course) with two M2 screws. The gear wheel is used to mount the disk onto the shaft. Of course, you can also use a larger gear wheel, and file off a lot more ...
Just make sure that the slope is such that the roller of the arm will force the potmeter to return to the "steady" position.
The picture shows how I intend to make the disk for the FQS rotary switch. The outer "fat" black circumference is the actual outer circumference of the gear wheel. The blue circle indicates where the peaks must be, that is, where the rotary switch does not have a stop position. The red circle indicates where the "valleys" must be, that is where the rotary switch does have its steady positions. As you can see, from the NORM position to the TEST position the distance to the center of the gear wheel only increases, thus the tension of the spring will increase. I hope that this is sufficient to obtain the desired mechanical behavior ...
A small block in the "U"-shape profile with two screws in the gear wheel (acting as paws) will create the mechanical end stops.
Remark on the PHCC analog hardware interface

I received an email from Viperpit member Klaus. He told me that the PHCC Motherboard scans the analog inputs, and sends a measured value for every detected change. As you rotate the knob, the resistance will have (in theory) infinite changes ... this implies a severe impact on performance as many values are sent to the software which in fact are useless. If the software polls the analog input channels at a rate of 20 ms, you will receive all measured values, thus the "final" switch position detection will take time! That time might be too long, causing unwanted or strange behavior. This needs to be checked out for sure!