New Trim Indicator Design

The Kitfox's horizontal stabilizer has a electric trim which allows the plane to be flown "hands-off" under all flight configurations.  It is important for safety to know exactly what position the trim is in so that it can be set properly for takeoff and landing.  Although this was designed for the my particular kitplane, with changes to the mounting brackets, this could be used on any other experimental aircraft.  The plans call for a mechanical trim position indicator in the cockpit, here is how it works (in theory):

	the moving horizontal pulls the cable
	in the cockpit, the cable pulls a coil spring
	which has markings on it to indicate trim position

This mechanical setup has many problems:

• First, significant strain is placed on the trim motor in overcoming the cable tension.
• Second, cable is prone to eventual breakage because of the high tension
• Third, the indicator is not accurate and has significant historysis.  The dial position for the take off position is quite different if you get there from 'up-trim', versus getting there from 'down-trim'
• Fourth, it is really difficult to see

The solution is to replace it with a digital system, to see a video of it in action click here!  Here's how the new trim indicator works:

	Two strong permanent magnets are mounted on the front of the horizontal stabilizer with an air gap of approximately 3/8" between the magnets and the circuit board with the magnetic reed switches.
	An array of magnetic reed switches placed in front of the horizontal stabilizer senses the magnet when it passes over the reed switch.  When the magnet passes over a reed switch, it closes causing the corresponding LED to illuminate in the cockpit.  For reliability, each half of the sensor array is completely independent - its own magnet and its own cable so that any malfunction will only result in every other reed switch being inoperative still leaving a working system (just with decreased resolution).
	Here the trim motor is shown moving the horizontal stabilizer, the magnet passes over the reed switches.
	The reed switch array data is carried to the cockpit on two cables with gold plated RJ45 connectors
	Where they illuminate LEDs on the center console.  The air gap distance was chosen so that the magnets always have two switches closed, that way if either half of the system fails there is still one LED illuminated thru the horizontal stab's whole range of motion.
	Here is the circuitry for both the cockpit and the tail shown together, some notes are in order.  First Rather than allow the LEDs to dim and brighten with load on the electrical system, I regulated the power supply to 9.1V.  Since the current requirement was so miniscule, I used a 1W zener diode to regulate the voltage.  For dimming, I use a 5.6V zener diode and switch between the supplies with s SPDT switch.  The resistor value was chosen to limit the LED current to 30ma when in the daylight mode, this conservative value should insure the long life of the LEDs.  I kept things as absolutely simple as possible in this first version, no integrated circuits were used.

The advantages:

• accurate and precise indication of elevator position
• low weight
• no moving parts, does not hinder elevators motion
• unaffected by environment, hermetically sealed
• excellent night visibility (has dimmer)
• logical user interface
• inexpensive, costs $30 vs $300 for a sealed linear pot

I view this project so far as a prototype and I wanted to insure everything worked properly before wasting time making it better.  It has passed with gold stars and here is what the next version will do:

• increased resolution, 64 instead of 14 divisions
• increased reliability
  • printed circuit board for more reliable connections
  • full redundancy, results are compared in cockpit and alarm set if they are not identical
  • error checking circuitry to catch any failures
• digitally encode horizontal trim position for use in autopilot and 'auto-trim' feature

   

  This circuitry is in the tail.  Cascaded 74148's do a priority encoding of the 64 reed switch inputs and the result is placed on the A, B, C, D, and E signal lines.  In this schematic I only show two 74148's cascaded, I will need eight for 64 reed switches.  Also not shown are the tranzorbs protecting all the circuitry from lightning.  Power is obtained from the cockpit board.

  In the cockpit is a circuit board which takes the five signal lines (A,B,C,D,E) and decode them into 64 LEDs.  Again, in this drawing I only show the circuitry for 16.  The decoding is done with a 74154 and to drive 30ma thru the LEDs, the outputs of the 74154 go thru a 74F244.  It is very important that any errors be caught, the only thing worse than a malfunction instrument is not knowing that you have a malfunctioning instrument!  A very bright red LED illuminates if: either cable is disconnected, no reed switch senses a magnet, or if the redundant circuits do not give exactly the same elevator position.  For dimming, a 555 is used to lower the duty cycle and dim the LEDs.  The frequency is kept low (~1Khz) to avoid producing noise which would affect the avionics.

OK, now getting really out of control here, it is a small step from this to adding the 'auto-trim' function.  Having 3 buttons you can press: one for take-off trim, one for landing trim, and another for cruise trim.  In each case the circuitry would activate the trim motor and move it until the correct position is sensed. 

back