This is a continuation of my old thread on the old forum: https://forums.robertsspaceindustries.com/discussion/351726/dissatisfied-with-commercial-hotas-offerings-i-ve-undertaken-making-my-own-here-is-my-progress/p1 To get you caught up without having to read all that here is what I've achieved so far and what I plan on doing still: After realizing there isn't a hotas out there I like and the ones that get anywhere close are really expensive (and I'm pretty sure VKB makes 5 or so of each product, and only sell half of those) I set out to make my own, pondered the design of the mechanisms for both for a long time and one day spotted a unused screen arm in a bin, the perfect central mechanism for a lever style throttle. I have a self-built i3 style 3d printer which is the main reason I can set out on a project like this, I design parts to print in CAD and then print them, but I'm trying to use bearings and bolts etc as moving parts to reduce wear and tear etc. (called vitamins in 3d printing circles) I've made a prototype of my throttle design: It's fully functional but it's missing finishing touches like sanding, paint etc, but basically it's finished unless I decide to change the design, still unsure on that. It's about 90% what I wanted it to be, some minor tweaks in the overall shape would improve the ergonomics (it's a tad bigger than I'd like) but that would require re-printing and re-wiring the entire thing, a lot of work and time, for now it's already much better than my x55 throttle which now lives in the drawer of limbo. Next steps are the joystick and one or more auxiliary button panels etc. Been designing and re-designing the joystick's gimbal mechanism for a long time now trying to settle on a design that works and doesn't need hard to get materials. Designing the base gimbal mechanism isn't difficult at all, the hard part is making it center back with the right force, sense the angle of the axis etc, been looking at existing joystick designs for inspiration (great artists and all that) Most of these came from some russian site (no idea how I ended up there) with no source given sadly. VKB's design with a single lamba:ish λ cam per axis is very elegant but if you look closely at the shape of it that's a really difficult thing to emulate, I could 3d print a piece of plastic with close to the same shape but I'm not sure plastic would last, and finding a piece of metal with that shape would be real tricky. Using some other approach to move one lever that's sprung like in the second picture seems easier but again requires a pretty complex shape made from metal. This design seems more promising, yes those 2 cams have a rounded contour but it's a lot more achievable, not entirely sure how much it would differ in feel to have flat sections instead of rounded. Plenty of L shaped metal brackets in hardware stores out there, could use em raw or file down a round section (not sure how to get a smooth finish on metal though after filing, just sand with progressively finer grit?) This design seems even more achievable for someone doing it themselves, basically there is an arm attached to a spring on both sides and each arm only goes in one direction, not entirely sure how it's rigged inside to achieve that but it's not that complex, all you really need is a piece that pushes on another piece but only going one direction, other direction the pin just goes away from the other piece. Could probably be 3d printed pretty successfully. (metal fasteners, bearings etc on moving points, to attach the springs etc ofc) This setup looks very achievable using plastic I think 2 arms that pivot a lot like a scissor is setup, not entirely sure how they spin with the setup in the photo, 2 bearings and they're attached to the outside while the inside is stationary perhaps? Not important I guess. Pins top left and top right stop the arm's when centered so they stay put while the other arm gets pushed by the pin between them to extend the spring. Could line the inside of the arms where the pin pushes with metal and use a metal pin to reduce wear and tear, plastic is great for most structural stuff if you use enough except where there is contact like that, plastic on plastic wears out in a while, metal on plastic wears out the plastic right quick. Especially when it's rubbing and not just hitting square on. This seems the easiest setup by far, just have 2 springs counter each other, but I'm guessing it's pretty tricky to get right, need the right springs mounted at the right angle and distance etc. Still pondering which design to emulate and how to achieve said design, some of those really need some pretty fancy shaped metal cams to work and I'm pondering either finding something close enough to file down or using a design I can do in plastic. For sensing the angle I've bought some MLX90316 14bit SPI rotary hall effect sensors: Works really great in that test setup with a small square magnet on the end of my drill, the SPI intereface is nice because it sends the 14bit angle data digitally to an MCU, that's more than the 10 bits of an arduino or similar's ADC can do and an MCU using analog read is slow because it has to wait a couple ms to sample the voltage. It's nice for joysticks etc because instead of sensing the distance of a magnet like regular hall effects it senses the angle of a magnet, sensing distance requires 2 magnets with a hall effect sensor between them, it's fiddly to get the distances right and you need to get the movement angles right or you smash the magnets into the sensor, or doesn't get close enough to get full range etc, was a small nightmare to get right on my throttle. Going to swap the throttle over to one of these as well for improved performance and fidelity. Since that video I've bought some much smaller surface mount to through hole adapter PCBs to use: