I’m trying to not make this come off wrong, I really mean to ask if it was worth it compared to the analog of flying IRL…other than a fun project? I see you fly in real life and was trying to see how close it got you?
I did a much simpler build in the sense that I avoided having to dial in a PID loop and to me it feels spot on enough for GA that it just feels like getting in another plane…going from even two skyhawks same model, same year at my local FBO feel quite different even though they share similar characteristics.
My build was on the basis that unlike driving FFB most of the time your yoke inputs are pretty minor so you don’t percieve a continuous gradual change in force. I.e. I trim as I go so I remove the control force as quick as possible and if I need to push or pull against the yoke I don’t have some gradual comparison of relative force changes, so I designed it around phases of flight though it is still gradual it doesn’t require PID to continuously adjust force. There is a little bit of a sacrifice in that the force is in steps but I also avoid having any unrealistic motor jitter feel and I’ve made it such that it blends the steps well. It is 90% mechanical with some solid state support and basic IO using SimConnect. I stumbled across this looking to adapt my trim to a motor that positions itself based on sim connect IO.
My build consists of a yoke with pitch axis slightly inclined with some ballasts to about match the weight of the elevator with engine off and fixed to track roller bearings. So without a gust lock in, my yoke will slowly drop like in real life. Then have 2 types of resistance I add for pitch, hopefully I can explain it well - its complicated to explain but its actually quite simple:
Stage 1 is just like a commercial yoke with bungies and provides the fine tuning but only for the pitch axis because in a real plane I never feel like the yoke wants to self center per se in the roll axis - which makes sense that’s the point of the aileron weights to prevent the ailerons from being blown around in the wind. During ground movement I have a linear actuator that applies tension based on engine rpm…as you do the runup youll see the elevator move to your trimmed neutral and it bounces around just like IRL. This stage is also active during stage 2 to provide pitch centering to neutral and to aid in that region of very light controls around your trimmed point.
Stage 2 uses linear dampers (think cabinet drawer soft close but bigger)which to me has a very similar feel to pushing through the air because its mechanically similar (don’t use a shock absorber because those don’t have a linear characteristics and you’ll feel them compressing the fluid which doesn’t feel the same. This is the hard part to explain picture a v pattern of the dampers and a tab that slides with the yoke and bumps into the dampers, and to avoid a moment, a matched damper weight is on the left and right, this is mirrored for the opposite direction of elevator deflection. This means as you push in the yoke for more deflection you will push against an increasing amount of dampening. Then the magic comes from a rotary actuator in the center that attaches to two mirrored work gear setups that pull the V into a flatter V so you hit more actuators sooner (and more total because some are out of reach at slower speeds so you get more resistance total not just more sooner). Think of a double door closing…each side is hinged and gets pulled down uniformly. The stage one bungee provides the holding force and is attached to the same brackets so as you apply more pushing force you apply more holding (self centering force). Sounds complicated but its one motor simply looking at IAS and RPM (for some tweaking with slip stream and ground
movement) from sim connect. Each damper and bungee is mounted on a rail with set screws for fine tune adjustment.
Trim function: my trim wheel moves a gear that physically moves the resistance section of the yoke to align with the position of the yoke so you literally trim out the resistance. The hall effect sensor for the pitch axis moves with the resistance so zero is at zero, and moves another hall effect to measure how far from center the trim is (i.e. hall effect trim axis). This was neccessary to avoid thinking too hard about programming required to make MSFS do what I wanted. This is kind of where I am looking to improve and maybe you have figured this out (admittedly I didn’t read every post in the thread and may have missed this). MSFS assumes your controller is going to spring to zero and as far as I can tell doesn’t have an option for your controller to follow simulated elevator position in the sim. In other words trim physically moved the elevator IRL based on the drag it adds to the control surface and that physically moves the yoke. So my issue lies in say for instance I add more power, my elevator is going to move to pitch plane up to use up the extra kinetic energy which leads to a constant speed (eventually after it stabilizes but with a climb component added), mine doesn’t - the plane moves but the yoke takes on an incorrect relative center. It would be nice if you could turn off MSFS math to account for a zeroed control surface but for now, unless you see another way, I am going to decouple the physical trim connection, put another rotary actuator that backs out the math to produce a relative elevator center position while physically in the position it should be for. Right now I account for it with a center reset similar to a seat lock slider with a button that releases the trim gear just like with track IR or VR having to reset a relative center.
My roll axis is even simpler, I used a friction motor to create a weighted feel with inertial resistance and light bungie tension to eventually center it but you have to rotate against the weight of it a little bit in either direction which is more linear like the cable system it is vs variable tension on a bungie. I have a solenoid that engages a rocker arm for 7lb linear dampers when air speed is alive or at higher rpms.
All of the dampers have spring plungers on them which provides a great solid feel without the wierd springy bungie feel of a retail product and without gear lash from motors. That is also what allows for that small area of control with very little actual force you get in a real plane (as least that’s the best I can describe it).