Thanks to this fantastic medium it is possible to tap from so many brilliant resources and put something together that may be worthwhile for the entire flightsim community.
What my design was really lacking was a once over by someone who knows what he's talking about when dealing with mechanical designs. Dave Best took the Force Stick design and pretty much redesigned the whole thing. In the article below you can read about his thoughts on the concept. You will find that we are getting closer to an actual design that can be build. I am proud and excited that people like Dave put in the effort to raise the Force Stick to the next level!
What I really like about Dave's design is the simplicity of the whole thing. I am sure this new mechanism is cheaper and easier to build. Also it will take a smaller footprint on your already crowded desk. Being an aeronautical engineer Dave did not only redesign the Force Stick mechanism but he also proceded to design the electronic implementation as well.
Introduction - I read on the FlightSim.Com web site about the possible design for a force yoke that would more accurately depict the forces that you feel when flying a real aircraft. I looked at the proposed design and decided that I might be able to add to the discussion my own concept for a force feed back yoke system. Hardware/Software design is out of my field so you will have to help fill in there. I am sure that any competent flight simulator programmer could provide the correct software to drive the device and I am sure that a hardware designer could rise to the occasion with off the shelf hardware. The photos depict the basic configuration in a conceptual manner. I did not size any of the components for actual loads or motion. The springs shown are depicted only to indicate relative strength to help understand the mechanical concept. To describe the device I will refer to the colors of components in the description to identify the parts I am talking about.
Figure 1
Overview - Figure 1 shows two force feedback devices attached to a common yoke to supply roll and pitch inputs. The force feedback unit mounted on the control yoke shaft supplies pitch while the unit located to the left of the picture provides roll. The roll inputs are made through a sector gear and rack. The roll control feedback unit does not have a trim system in this assembly.
Figure 2
Figure 2 is a picture of the pitch force feedback device with trim system as used for pitch control. The control yoke shaft is depicted in blue. The support frame is gold and supports all of the mixer components. The trim system is depicted in green and the force feedback system is depicted in red and purple. All springs are depicted in gray but the system they work in should be obvious. Guide bushings are yellow and free to slide on the shafts they are installed on.
Control Yoke Shaft - The control yoke shaft (blue) passes through the frame (gold) and can slide fore and aft along its centerline in the frame. The shaft is also able to rotate about its centerline in the frame. A flange attached rigidly to the shaft reacts the spring forces from the force feedback, trim, and passive centering systems to the control yoke shaft. The shaft is moved by actuating a control yoke handle (white in Figure 1) with your hands the same as any real aircraft.
Passive Centering System - A pair of low force springs are placed on the control yoke shaft on either side of the control yoke shaft flange. These springs react against the inside surfaces of the frame and provide a low centering force just like a standard control stick used on a game joystick. This set of springs is provided to remove slop from the system while always providing a minimal centering force. This force does not really exist in most real aircraft but I think it will make the system work a little smoother and it may be required in some low force feedback input conditions.
Trim System - The trim system (green) consists of a threaded shaft located above the control yoke shaft that passes through the frame parallel to the Control Yoke Shaft. This shaft is rotated by a electric motor which is depicted by the larger diameter green cylinder to the right of the frame in Figure 2. An inverted 'U' shaped fitting reacts the forces of two centering springs on either side of the Control Yoke Shaft flange. Holes on the 'U' fitting are threaded to engage the Trim System Shaft. The trim motor rotates the shaft which causes the 'U' fitting to translate parallel to the trim shaft centerline. This preloads the control yoke shaft depending on the movement distance and direction to provide a new centering location for the control yoke shaft. This is the same way a bungee trim system works in a real aircraft such as a Piper Tomahawk. The motor that activates the trim system does not have to be wired to the computer in any way. A separate switch and power supply can activate this system independently of the flight simulator software. This keeps the trim system from drawing on computer resources and also will provide greater control over the trim force than the simulator software seems to provide.
Force Feedback System - The force feedback system works very similar to the trim system except that the two springs involved in force feedback work independent of one another. You can see in Figure 3 that the force feedback system is similar to the trim system in that it has a shaft (red) parallel to the control yoke shaft that passes through the frame (gold). This shaft is driven by a motor depicted as a larger diameter cylinder (red) to the right of the frame. The shaft in this system is threaded like the trim system but the threads are different. The threads to the left of the control yoke shaft flange are right hand threaded and those to the left of the flange are left hand threaded. The 'U' fitting of the trim system is replaced with a pair of threaded stop plates (purple) that translate in opposite directions along the force feedback shaft when the shaft is rotated by the motor. In addition to reacting against both sides of the control yoke shaft flange these springs also react against a stop flange on the frame. Both springs react against both flanges only when the control yoke shaft is in a neutral position with the trim set in a neutral position. If the force feedback motor is rotated in a direction that moves the translating stop plates toward each other, the effect is to stiffen the control input force required to move the control yoke shaft along the shaft's centerline. If the motor is rotated in a direction that moves the stop plates away from each other the effect is to soften the force required to move the control yoke shaft along its centerline. If the force feedback motor is rotated enough that the force feedback springs are moved away from contacting the frame stop the control force will resort to the centering spring force provided by the middle set of springs (unless the control yoke shaft is translated far enough to contact a force feedback spring again). The Frame stop is really the key to the force feedback system. Without it you could rotate the force feedback shaft all you wanted and the force on the control yoke would always be the same. This is because the force feedback springs would cancel each other out. The frame stop allows the force feedback springs to work independently, providing increased force resistance when ever the control yoke shaft is displaced from the neutral position and always resisting the direction of motion.
Figure 3
I had problems with the transmission of the roll movement as it would be too expensive and difficult to produce. It took Dave only one day to sort that one out and he send me the following modification. (Paul van Dinther)
FORCE STICK ROLL MODIFICATION
The rack and pinion coupling between roll and pitch may prove to be a bit much for the average person to build so here is a alternate configuration. In Figure 1 you can see a new configuration where the rack and pinion setup has been replaced with an offset rod (white) that engages two flanges attached to the shaft on the roll control force feedback unit. The rod is attached to the force feedback control shaft on the pitch feeback unit by two parallel standoffs. This arrangement allows for free translation along the pitch axis and free rotation about the roll axis at the same time. I would suggest designing the flanges on the roll control force feedback unit shaft to rotate freely around the shaft centerline. This would reduce friction buildup during combined roll and pitch inputs. Just imagine a pair of cheap wheel bearings in place of the flanges and you will get the idea.
Figure 1
Figure 2 addresses the problem of picking up inputs from the force feedback units. I threw something together at lunch but I'm not happy with the concept. A roll control potentiometer (bright green) can easily be activated by inverting the same configuration as the roll to pitch coupling shown in Figure 1. This would work out pretty good if your input device is a rotational one. I put a potentiometer on the pitch shaft (also bright green) to start a little thought in that direction but soon realized that the arrangement of a slotted arm would provide a coupling of pitch when roll inputs were made with pitch in other than the neutral position. I think I would prefer a 'U' joint to drive the potentiometer arm at the end of the control yoke shaft to the arrangement that I have shown. I wanted to point out one thing about a design like this. You design a force feedback unit for pitch and the same parts get used for roll. If you want you can even use the same parts for rudder control. The same software and I/O port could be used to drive all axis as the dynamic pressure build up would be about the same for each axis. A change in spring rates for each axis can be used to roughly match the balance of pressures between roll, pitch, and yaw.
Figure 2
Material suggestions - I used very basic shapes in this concept. I know many people are intimidated by making parts. The frame components (gold) could be fabricated by cutting of a length of rectangular tube and drilling some holes. The shafts (blue) can be tubing or rod depending on how hard you want to yank and crank on them. The flanges on the control rods would be easy to make by fixing cheap wheel bearing to the shafts and letting the springs and coupling components ride on the outer race with complete freedom. The trim fitting (green) can be cut from channel or bent up from sheet metal. The bushings (yellow) can probably be found at a hardware store near by. If you don't have a set of dies to thread the shafts, find a local machinist to do the job; it won't cost that much. Just remember to keep things stiff so that nothing binds. Hope this helps take things to the next level.
The Control Systems
Figure 4 shows the basic architecture for the force feedback system. In the schematic the Force Feedback Unit (FFU) has two motors M1 and M2 and five switches S1,S2,S3,S4, and S5. M1 is the trim motor and is powered directly from the power supply via two relays, R1 and R2, switch S1, and two limit switches. The direction of motion for the trim is selected by activating switch S1. As long as S1 is held on either R1 or R2 turns on to power the trim motor M1. M1 runs until S1 is released or one of the limit switches S2 or S3 are activated by the movement of the 'U' shaped trim fitting (Green in Figure 2) This trim system requires no interface with the PC that you are running your flight simulator software on. The motor M2 is the actual Force Feedback Motor. This motor is powered by relays R3 and R4. The motor uses limit switches S4 and S5 in the same manner as the trim system to protect against over-running the mechanical device. The relays R3 and R4 are activated by a Micro Processor Unit (MPU) The MPU takes commands sent from the input/output port of the PC and combined with position information based on a rotation counter (CNT) determines how many turns to rotate the motor (M2) to get the force feedback springs preloaded to the correct level. The MPU could also use input from the limit switches (S4,S5) to fine tune the zero position for the force feedback unit on the fly by reseting the counter to the limit values whenever a limit switch is activated. The simplicity of this system lies in that the MPU only requires one input and that input can be based on only one parameter, Velocity, Dynamic Pressure, or possibly the exiting force feedback variable from flight simulator software are likely values. The box in the diagram labeled PC represents the Personal Computer that drives the MPU. On that computer Microsoft Flight Simulator runs as normal, constantly updating information variables about the aircraft. The DLL FS6IPC.DLL (availiable for download at FlightSim.Com) reads the variable for Velocity, Dynamic pressure, or force feedback and makes that information available to a driver routine (box labeled "driver"). The driver routine takes the value of the flightsim variable and looks up a corresponding value in a chart generated for each aircraft that contains feedback preload vs velocity information. The driver then supplies this information through the I/O port to the MPU to correctly position the force feedback springs by turning the force feedback motor the correct number of turns to get the desired preload on the force feedback springs. The advantage of using a separate file for each aircraft is that you can add this file into the directory for an airplane and tailor the feedback profile specifically for that aircraft or aircraft class. Some airplanes may behave linearly and others may not so this gives you the same flexibility to customize as Microsoft Flight Simulator already offers. Of course the driver will have to be intelligent enough to find out what airplane you are flying and where to find the force vs velocity file.
Figure 4
MPU - the MPU must control the position of the force feedback mechanism with a bit of accuracy. Figure 5 shows the basic architecture of the MPU. Of course the figure only shows the logical devices and data links not the actual components or circuit needed to build the MPU. An optical pair acts as a pickup that is triggered by a disk with a hole in it that would be attached to the force feedback motor shaft. This pair signals a driver that triggers a counter. The purpose of the counter is to keep track of how far the force feedback shaft has compressed the springs by assuming 1 rotation is equal to one pitch length in movement. An I/O buffer supplies the output from the PC driver program from the PC I/O port to a comparator with a coded signal made up of 8 binary bits. The first six bits indicates the position to move the force feedback springs to in 'turns of the shaft' or 'number of pitch increments' as a binary number. The seventh bit indicates the direction to rotate the shaft and also the direction the counter should count up or down. The eighth bit indicates when to turn the force feedback motor on or off. The position bits are compared with the counter bits in the comparator and if the numbers don't match the motor is turned on by signaling a solid state power supply that controls the relays R3 and R4 in the indicated direction with the select line until the numbers match. When the numbers match the motor is turned off by signaling the power supply again to turn the relays off. The shaft system will have to be rotated fast enough to respond to airspeed changes but not so fast that it 'coasts' for a bunch of turns after the MPU turns the motor off. Also the software driver would not want to supply commands faster than the system can respond. I am sure that there are much better feedback and positioning system that would work such as those on radio control aircraft servos. I am just trying to get across an idea here in simple terms.
Figure 5
About Complexity And Springs - I know that this system requires more mechanical complexity then the system that you proposed. However, this system uses mechanical complexity to avoid creating a large burden on the computer. I think the trade off is worth it. Instead of having to constantly calculate where the trim system should be moved to for proper feedback the system simply monitors velocity and adjusts only as needed one portion of the mechanical device. The trim system is independent and does not affect the calculation of force feedback pre-load when the trim or control position is changed. The springs selected are compression springs as they a less likely to fatigue and break under continual use. Even the tailwheel springs on my aircraft are compression springs used in a tension application. Saves ground looping! Tension springs fatigue at the attach hooks due to the large local moments at the attachments. I assumed that everyone would figure out that the control shaft is used to drive the potentiometers for the pitch and roll controls from a conventional game controller so I didn't show them in the pictures to keep from cluttering things up any more then needed. Another thing, this system would more than likely create a lot of noise, so plan on headsets for your PC sound output to avoid the whine of motors and the clicking of relays and switches. On the control shaft I think I would use a ball bearing fixed on the shaft in place of the flange to allow the system to rotate easier under load. As for deciding how strong the springs should be, Federal Aviation Regulations (FAR) part 23 has design criteria for smaller aircraft and has a very useful appendix with simplified design criteria for aircraft with lower weights. This document should provide maximum control forces for pilot inputs that would help to select the right springs. Pilot reports in magazines like Sport Aviation from the EAA also report stick force per 'g' which would be a good measure for tailoring feedback. Another thing, a control yoke with real input forces will have to be nailed down pretty good.
About Me--I am an engineer at an aerospace company in Southern California. I got my PPSEL when I was in college and have flown off and on over the years as money allows. I built an experimental aircraft and flew it until a few modifications got too time consuming and I put the project on the back burner. I have been in between flying periods right now and the flight simulator has been helping me to brush up on procedure and navigation within my budget. Since I have benefited from many free downloads from FlightSim.Com I am providing this article and its contents for all to use. Just remember me if you make the big bucks off of these concepts!
