MANUAL TRIM panel Some mechanical solutions required

Introduction

The MANUAL TRIM panel has just a few controls and indicators, but its simplicity is deceiving! There is one "flat-top" switch, 3 potmeters and two small 180° rotation needle pointer indicators. Easy, you'd say. Yes, as far as that 2-position switch and the YAW TRIM knob goes. However, in the real F-16 MANUAL TRIM panel, the ROLL TRIM knob and its associated needle indicator is one unit. Same goes for the PITCH TRIM knob and its associated needle indicator. When you rotate the trim knob, the needle indicator is also moved via some gears inside the unit.
Further, on the Side Stick is a round button that can be moved in four directions, effectively roll and pitch trimming the aircraft. When using the Side Stick, the knob and needle indicator move! So, there must also be some form of a servo mechanism involved.
All that is mechanically way too complicated for me, so I was thinking of a solution that I can build.

The idea, components and operation
motor potmeter

There are two key items in the trim knob. First, there is that needle indicator. Basically, that is fairly easy to implement with small aircore motors.
But the most striking items are of course those ROLL TRIM and PITCH TRIM knobs that can rotate "auto-magically" when the trim switch on the Side Stick is moved (up / down / left / right). That type of knobs are found on some stereo amplifiers, they are called motor potentiometers, or in short, motor potmeter.
I found on eBay a really good deal for motor potmeters. What makes life easier, is that the potmeter is of the "linear" type, not "logarithmic" (used for volume control). These motor potmeters are specifically designed that they can be operated manually (without too much force), and that the potmeter can also be rotated by a small motor. There is no end stop detection. A slipping clutch prevents mechanical damage when run into the end stop. You can detect this event by measuring the motor current.

So to summarize, what do we have for each trim knob?

needle indicator
motor-driven knob
manual control and control via the switch on the Side Stick

The design
The PHCC Motherboard reads the roll trim and pitch trim potmeter setting via analog inputs. The software can convert the read value to an appropriate setting for the needle indicator. That's why it makes life easy if the potmeter is the linear type. The needle indicator is driven by an aircore controlled by the PHCC aircore daughterboard.

Manual control
If the knob is rotated manually, the PHCC Motherboard will detect the changed measured analog value from the potmeter. The software simply updates the position of the needle indiactor, and that's all! Of course, the measured analog setting can be passed to BMS for the flight model effects.
Side Stick switch control
Now it gets interesting. If the switch on the Side Stick is operated, the trim knob should rotate a few degrees every time the switch is operated. Basically, all that we need is to activate the motor a short moment when the switch is activated. As the motor will drive the potmeter, the update of the needle indicator is accomplished using the "manual mechanism".

Realization
trim control design

To keep things simple, I decided to build a separate "PHCC daughterboard" to control the two motor potmeters. This is really a simple design, as it only needs to read the analog potmeter setting and drive the motor. I am using the fact that this motor potmeter actually has two potmeters on the motor shaft, and that I already have the PHCC aircore daughterboard. The extra cost for a "local controller" for the motor potmeters is just 5 Euro. All other components would be needed anyway. Besides, with the local electronics, the wiring to the MANUAL TRIM panel is kept simpler.
A nice feature with the local control is that at power-up, the knobs can immediately rotate to the center position. Yes, that is also possible without the local controller, but now the knobs will rotate the second you apply power!
One potmeter is connected to the local controller, so that the local controller knows the knob position. The other potmeter is connected to the PHCC Motherboard analog input, only used to update the needle indicator aircore (and pass the position on to BMS, if needed). When the PHCC Motherboard detects that a trim switch on the Side Stick is activated, it sends a DOA command to the "PHCC TRIM daughterboard". On the PHCC TRIM daughterboard, the motor is activated for a short moment.

Initially, I planned to use a so-called "H bridge" to control the motor. Using an H bridge, you can control the current flow direction through the motor, making it rotate clockwise or counter-clockwise. However, this is just a small motor and the design of an H bridge using MOSFETs is not that simple, especially the "high-side switching" section. I do not see any reason why not to use a simple relay to reverse the power supply polarity. Likewise, you can use a relay to switch the motor on and off, driving the motor directly from a power supply. However, I like to keep the possibility in my design to use a PWM signal to lower the rotation speed of the motor. Thus, switching the motor on and off will be done with one MOSFET. The PWM is just a matter of switching on and off very fast!

Construction

As you can see looking at the MANUAL TRIM panel, the motor potmeter will be at the same location as the aircore motor. As the aircore motor has a short thin shaft to drive the needle directly, the aircore is mounted on a small piece of circuit board directly on the backplate. Thus, the motor potmeter has to be installed at some distance from the panel, above the aircore.
The trim knob is mounted on a shaft using a 50 teeth gear. The diameter of the 50 teeth gear is sufficient to mount the trim knob on the gear using three M2 screws. This shaft is driven by a 40 teeth gear. On the motor shaft is a 22 teeth gear. To obtain the required distance from the backplate (to create space for the aircores, an additional intermediate 40 teeth gear is used. The ratio is 22/40, which is good, because the motor potmeter can rotate 300°, and the trim knob on the panel seems to have a rotation range of approximately 120° - 150°. Thus, the trim knob range is translated to almost the full rotation range of the potmeter.

collar type slide bushings , 40 teeth gear , nylon rings , and trim knob on 4 mm shaft with 50 teeth gear

trim mechanism parts At the left you see the parts needed to build the trim wheel. construction for one knob. The trim wheel assembly consists of two aluminum plates and one L-shaped aluminum bracket. The two plates are 37 mm x 58 mm, the L-shaped bracket is 20 x 40 mm, 58 mm long. The trim wheel is mounted one a 50 teeth gear (sufficient diameter, the gear is only used to mount the trim wheel on the shaft). On this shaft is a 40 teeth gear mounted. The diameter of the shaft is 4 mm. The shaft is held between the L-shaped bracket and one metal plate using collar type brass slide bushings (which are oiled). The collar of these bushings "locks" the shaft between the bracket and the metal plate.

The "intermediate" 40 teeth wheel is mounted on the L-shaped bracket, also using a collar type brass bushing. The pictures shows a metal bushing with an inner diameter of 4 mm to fit on tha shaft and an outer diameter of 6 mm to fit on a 6 mm ring with a setscrew. However, I found some 4 mm rings with a setscrew, so the bushing and 6 mm ring are replaced with the 4 mm ring (which is better!)

The other metal plate is mounted on the opposite side of the L-shaped bracket using two M3 screws. The motor potmeter is mounted on this metal plate. The complete assembly is mounted on the backplate of the MANUAL TRIM panel with the L-shaped bracket using two M3 screws.

Aircore indicators

The two "half moon" openings in the MANUAL TRIM panel are for the trim indicator. Mike's panel comes with the following parts for each needle indicator (going from the front panel side to the backplate side).

indicator "glass"
indicator needle (you have to paint it yourself)
flexible tube (circumference fits the needle, inside fits on the aircore shaft)
indicator position dial

There is little room between the indicator glass and the indicator dial, and you have to make sure that the indicator needle can move freely between them. The following attention points are crucial for a good result.

the distance from the aircore to the panel, otherwise the shaft will touch the indicator glass
the position of the aircore motor axis with respect to the needle indicator (more about that later)
illumination of the indicator dial

Due to the construction of the MANUAL TRIM panel, all trim indicator parts are installed from the rear side. Thus, the indicator glass goes first into place, then "the rest". As distance is so important, I decided to make separate "modules".

trim dial backlight The first "module" realizes the backlight of the indicator dial. On a small piece of circuit board are two SMD LEDs soldered with their series resistors. Against the circuit board is a piece of a ruler (!) mounted. The thickness of the ruler should be a bit more than the height of the LEDs, but should not be too thick, otherwise the axis of the aircore motor is not long enough. I drilled and filed openings in the ruler for the LEDs and resistors and cut grooves for the wires, so that the circuit board can be mounted flat against the ruler. The ruler is mounted against the backplate and is also the surface onto which the indicator dial is glued. The ruler makes a smooth surface for the indicator dial, because putting the indicator dial just on top of the two SMD LEDs would not be a "vibration-proof" construction. With this construction, the indicator dial is a bit further back from the front panel. Remember, the total height is limited, because of the length of the shaft of the aircore, but some play is possible with the length of the flexible tube. If the rear side of indicator dial would be against the backplate, the space for the needle between the indicator dial and the front glass is very small.
trim aircore board The second "module" is a piece of circuit board on which the aircore motor is soldered. Before you solder the aircore motor on the circuit board, you must make sure that the "middle position" of the rotation range is at the zero position of the indicator dial. The aircore motor is not a real motor in the sense that you can make it spin round and round. In the 360° rotation, there is a small "dead zone". If that "dead zone" in within the indicator range, you will curse and pull your hair ... and pray that you can remove the aircore from the board without destroying it.
The circuit board with the aircore motor is mounted against the SMD LED circuit board. If you cut the "LED board" and "motor board" with the opening for the axis in such a way that all the soldering holes in both boards align, you have to cut off just 1 mm of the pins of the aircore, because the rest of the length of the pins can go into the holes of the "LED board". This way, the two circuit boards, the ruler and the indicator dial make a compact and strong unit that can be mounted on the backplate using two M2 screws. As the four soldering joints for the aircore need some space, I have cut a bit of the board material on the LED board away with a drill. Do that manually, not with the power tool! A ring on the M2 screws between the LED board and the aircore motor board gives the needed space for the four connection wires of the aircore motor.

And these are all parts to construct one aircore TRIM indicator.

M2 screws , aircore motor board , "wire distance" rings , LED board , ruler and M2 screws

The check that everything fits (it should), put the two M2 screws through the holes in the "ruler plate". Then put the LED board on the screws. Check that the ruler and the board fit perfect against each other, otherwise it will be difficult to get the axis of the aircore perpendicular to the trim dial. As you can see, I used a 3.5 mm drill to manually remove some of the board material so that the solder joints of the aircore and the 4-pin header are not in the way to mount the aircore board also perfectly fitting against the LED board. There is some tolerance allowed, because between the LED board and the aircore board are rings to make some space for the aircore wires to the 4-pin header. The aircore board is the last part, and two M2 screws holds everyting together. As a final step the trim dial is glued onto the ruler surface with a small piece of double-sided adhesive tape.
Now disassemble the contraption ... this was just the check that everything fits the way it should. Using the "ruler plate" with the glued trim dial, you can now determine where you must drill two 2 mm holes in the backplate. For good orientation, mount the MANUAL TRIM panel with the correct screws onto the backplate. If possible, mount the trim wheel assembly also on the backplate (of course before you mount the panel). Now you can put the "ruler plate" on the rear side of the backplate and visually align the trim dial in the correct position. Mark the two holes in the "ruler plate" with a pencil on the backplate. Remove the MANUAL TRIM panel (and trim wheel assembly) and drill the two 2 mm holes. Once again, put everything back together!
I assembled countless times sub-assemblies and partially completed items to see if all fits, and get ideas how to proceed.
Now, put the indicator glass in the MANUAL TRIM panel, and carefully put the trim indicator assembly on the backplate. Check that the end of the axis of the aircore does not touch the indicator glass. If it does, simply put extra rings on the M2 screws between the LED and aircore board. I only needed one ring which is required to get the distance for the wires to the aircore.

Again, remove the trim indicator assembly, and then remove the indicator glass. Install the trim indicator assembly again onto the backplate. Connect the 4 pins of the aircore to the PHCC aircore daughterboard (or whatever hardware you use to control aircores). You have to control the aircore axis to determine the correct position of the needle. You can choose two different methods.

You did check that the aircore "middle position" of the rotation range is approximately at the zero position of the indicator dial ?

The needle points to the "0" when the aircore axis is at middle position.
This probably means that aircore setpoint value 0 is outside the dial at one side, and the maximum value is outside the dial at the other side.
You have to make sure in software that the needle does not move outside the dial range!
The needle points to the last mark on the dial when the aircore setpoint value is 0.
This means that aircore can never move outside the dial at one side.
Of course, it is still possible to move the needle outside the dial range at the other side.
You still have to make sure in software that the needle does not move that far, but the check is simpler.

I have chosen for method #2 ... but the setpoint 0 is just outside the indicator range. That gives me a little "play" when I put the tube with needle pointer on the axis. You do not want to see that at setpoint 0 the needle is just a little before the end of the scale! I do not think that the aircore has enough drive power to damage anything if the needle pointer runs against the side of the front panel.

The indicator needle is mounted on the aircore axis with a small piece of tube. The inner diameter fits snug on the aircore axis and the hole in the indicator needle fits snug on the outer diameter. Hold the indicator needle with tweezers on the axis so that the end of the axis is through the indicator needle just at the surface of the needle to determine the length of the tube. As the aircore itself is at some distance, the length of the tube is not critical, but excessive length will make it more difficult to push the tube with the needle onto the axis.
Before you can put the needle on the aircore axis you have to paint it. First, apply a thin layer of white paint on the needle pointer part (thus, no white paint on the round part for the shaft. This side (with white paint) faces towards the indicator dial. The round part for the axis must remain transparant, so that light (from the dial backlight LEDs) can illuminate the needle pointer. To block that light, the round part of the needle pointer is painted black on the top, the view side facing towards the indicator glass. This way, only the pointer part of the needle is illuminated.

After the paint has dried, put the cut-to-length piece of tube in the shaft opening of the needle pointer. The tube with needle pointer fits snug on the shaft of the aircore. Note that, once the needle pointer is mounted on the shaft, you cannot remove the aircore or LED module without pulling the indicator needle & tube from the aircore axis. The aircore is a delicate "toy" ...

YAW trim

The YAW trim potmeter is a bit special, if compared to a normal potmeter. A normal potmeter has a rotation angle of approximately 300°, whereas the YAW trim potmeter has a rotation angle of some 180°. The solution is simple. Mount the potmeter on a bracket, put an adjusting lock ring on the potmeter shaft and put a small screw on the bracket. Use a long setscrew in the adjusting lock ring so that this setscrew runs against the small screw mounted on the bracket. The screw is mounted in such a way that the rotation angle of the normal potmeter is mechanically limited to 180°. Due to limited space, use a small potmeter with a 4 mm axis.

Parts for the 180° rotation YAW trim potmeter
You need a U-shaped bracket, length 30 mm. As I did not have that, I used 2 pieces of L-shaped aluminum (you may have noticed that I use L-shaped profile aluminum a lot). The long part of the "L" is 40 mm, the short part of the "L" is 20 mm, the thickness is 2 mm. After cutting 2 pieces of 30 mm length, put them together in a vise and drill the hole for the potmeter in both pieces. This way, the two holes will line up perfectly. The rest of the work is as follows.

Enlarge the hole for the mounting ring in the mounting plate. The mounting ring guides the shaft of the potmeter (no axial play) and is used to mount the bracket on the metal backplate of the MANUAL TRIM panel. You can use the housing of an old potmeter for this. As the diameter of the hole in the mounting ring is 6 mm and the axis of the potmeter is 4 mm, I used a clamping 4 mm to 6 mm piece on the axis of the potmeter.
Drill one 2.5 mm hole next to the potmeter in the potmeter bracket. This hole is used for M2.5 screw that will limit the rotation angle of the potmeter.
Drill two 3.2 mm holes in the short leg of both brackets. Make sure that the holes align. Two M3 screws mount the two brackets together to form the U-shaped bracket.

Mount the assembly on the backplate with the mounting ring. Check that the axis rotates smoothly in the mounting ring. Position the potmeter in such a way that the counter-clockwise minimum position is the "upper" position of the YAW trim dial on MANUAL TRIM panel. The mechanical end stop of the potmeter defines the counter-clockwise end position. Now rotate the axis of the potmeter 180°, to the clockwise maximum position, which is the "lower" position of the YAW trim dial. At this position the setscrew msut be against the M2.5 mechanical stop screw. Of course, cut the length of the setscrew to the required length.
Here is a mechanical drawing (not on scale) to copy the brackets. Check the dimensions, they might be slightly different, as I improvise a lot ...
yaw brackets dimensions drawing

Local PHCC "Manual Trim daughterboard"

The potmeters can be connected directly to the PHCC Motherboard, the aircores to a stepper/aircore PHCC Daughterboard and the motors to digital outputs. That implies a lot of wiring. I decided to develop a PIC-based PHCC "Manual Trim" daughterboard for the following reasons.

Due to the local hardware, the wiring becomes simple.
Local hardware is required anyway, to control the rotation direction of both motors.
Due to the local "intelligence" it is possible to do all kind of nifty things, such as
- let both trim wheel knobs automatically rotate to the mid position at power-up
- implement a speed control for the motors using a PWM signal

In the final design, the control logic for the MANUAL TRIM panel is a mix. The motors are controlled by the local processor, and one of the two potmeters of each motor is read by the local processor. The second potmeter of each motor is connected to the PHCC Motherboard so that the trim wheel setting is available for BMS. The aircores are controlled by a PHCC stepper/aircore daughterboard (with aircore firmware).
I have chosen hex 54 as the PHCC address for the "Manual Trim" daughterboard. Hex 54 is ASCII "T" (for Trim). To address the required functionality the following sub-addresses are implemented in the firmware.

Click the image for a detailed schematic

sub-address	data byte	description
0	0 1	ROLL trim adjust request decrease (rotate conter-clockwise) increase (rotate clockwise)
1	0 1	PITCH trim adjust request decrease (rotate conter-clockwise) increase (rotate clockwise)
2	0 ... 255	Motor (trim) speed
3	0 1 2	Diagnostic LED usage mode LED off PHCC message acknowledge heartbeat mode
4	0 ... 255	8 digital outputs
5	0 ... 15	4 digital outputs
6	70 ... 180	ROLL center position definition
7	70 ... 180	PITCH center position definition
8	don't care	Rotate ROLL & PITCH trim to center position
9	0 ... 255	Watchdog time-out on trim wheels centering action. 0 :: time-out disabled 1 ... 255 :: time-out in increments of 100 ms
10	0 ... 255	Motor running time per trim adjust request. 0 :: return to default setting (200 ms) 1 ... 255 :: running time in increments of 10 ms

Backlight of the MANUAL TRIM panel
Manual trim backlight PCBs

The MANUAL TRIM panel has two areas with text that can be illuminated. The middle region of the panel has no text so there is no reason to cut away all the material at the rear side of the panel. For that reason I decided to make two separate boards for the SMD LEDs and resistors. Two small slits are cut to connect the power supply to the boards.

I was glad that I made that choice, because the shape of the boards is not quite a standard rectangle. I made an error in cutting the "island" for one LED of the left side board, but that is solved using a simple wire to connect the LED to +5 Volt.

The method used to make the backlight with thin copper-clad circuit board is described in the description of the EPU panel and the AIR COND panel.

Finishing up: the connection bracket

There are quite a few connections to and from the MANUAL TRIM panel. As with other panels, I try to keep sensitive signals as much as possible away from potentially disturbing signals. This is what I have chosen.

one 9-pin sub-D for sensitive signals
- DOA connection (GND , +5 Volt , Data , Clock , Reset)
- switch matrix (1 "row" signal, 2 "column" signals [the "row" signal and one "column" signal is passed through] )
one 5-pin "CB" connector for analog signals
- GND, +5 Volt
- ROLL , PITCH , YAW analog signals
one 8-pin "CB" connector for ROLL and PITCH aircore control
one 15-pin "High-Density" sub-D connector for 12 spare digital outputs
one 9-pin sub-D connector for illumination
- +5 Volt, backlight PWM - for panel backlighting
- +5 Volt, instrument PWM - for ROLL and PITCH dial illumination

I have the panel backlighting and dial illumination wired separately to the connector, because there are plenty of pins left. However, when the panel backlight is on and the dial illumination is off, the PITCH dial gets a lot more stray light from the panel backlight than the ROLL dial. Although the panel backlight and dial illumination will probably always be on at the same time, I do not think that it looks nice when the panel backlight and instrument illumination are not "up" at the same time. So, I will probably combine the two in the connector (not in the panel). I like to have a choice without having to do some modification in the panel itself.

The result

The picture at the left side shows the completed MANUAL TRIM panel. The wiring from the control board to the controls and motors is done, but this picture is more clear without the wiring.
The picture at the right side shows the illumination with the backlight LEDs on at maximum brightness. Note that the illumination of the dials is not turned on. The dial illumination is wired separately, so that the backlight and instrument illumination can be controlled separately. However, there is a clearly visible difference between the PITCH TRIM and the ROLL TRIM dial. I will probably wire the panel backlight and the dial illumination in parallel in the connector. So, the MANUAL TRIM panel itself has both options available.