It is the time of the year when a young man’s thoughts turn to home building their own helicopter sim controls.
Doing some reading on what others have done, one promising options is to use a pair of nested hemispheres for the cyclic joint. A thick grease would serve to both lubricate the movement, and act as a friction mechanism to hold things in place. Apparently some models of helicopters use a very similar system.
The question I was then pondering, was how would you measure the deflection to generate a control input? One thought was having the stick attached to a piston (or something similar) which then could monitor the linear motion. I could use an optical encoder on the piston, but the need to have a highly mobile joint seems like it would have a lot of slop to hinder precision. To get rid of that I think would be too expensive to be practical for a sim rig.
Any suggestions on sensor options would be welcome. I’m not set on the sliding hemisphere joint, but I want to work through it to see if I can see if something will work.
I once (a long time ago) modified a cheap racing sim handbrake to work as a collective, using a hall sensor and a neodymium magnet… I cannot now remember the part number, but if it comes to me I will come back and let you know.
The Thrustmaster Warthog has something like this… It’s basically just a ball with a hemispheric cup around it. It uses a 3D hall sensor at the bottom. Maybe the easiest way would be to use the electronics from a Warthog and copy the principle of operation?
Do you have any pictures or links to similar solutions?
I’ve met these guys at Oshkosh. To me, this is a better solution than Hall effect sensors or the like. Why? Because with this kit you can focus entirely on designing your stick and collective. Any size. Any throw. Once you’ve achieved mechanical perfection, you hook up the lines and you’re pretty well set.
For buttons I’d use an arduino clone of similar from Adafruit. Easy for me to say but I’ve never dealt with DX programming. It doesn’t sound very hard though.
In this drawing it’s a 12" hemisphere (blue) with a 1/8" gap to the outer partial hemisphere (gray). There would be some kind of grease between the two to both lubricate the movement as well as provide the “stickiness” to hold the cyclic. At the moment I’m looking at having the outside partial hemisphere(gray) be the mobile portion that the cyclic is attached to. I just need a way to track the motion of the gray hemisphere in X and Y. The radius’s are much smaller than I think I’d use in the full build to illustrate the concept. I think I’d probably end up with about a 36" radius for the blue portion, to give it a feeling or pivoting like a normal stick, but in actuality I think it’d be closer to sliding.
I had always assumed hall effect sensors only worked in one axis sort of like using pot’s. Need to do some reading!
Hall effect sensors are used everywhere these days. For example in the automotive industry. Every car has them and not just a few. They’re everywhere. Sometimes used to precisely detect a magnets position in space or some applications require incredible precision and mustn’t be influenced by magnetic fields other than the magnets field which is used to control the sensor. The way you do this is by having 3D sensors (basically multiple hall effect sensors on one chip) that can detect magnetic fields on three (or more) axis. That way you can detect the position of a magnet in 3D space and cancel out stray magnetic fields which can happen if you have multiple sensors in very close proximity or even electromagnetic fields which in this day and age with WiFi, mobile phones etc are everywhere and can influence hall sensors if they are sensitive enough.
Probably 3d print, but if I am feeling industrious form them out of aluminum, planish, and then lap smooth (probably the first, but maybe the later). Good point about the yaw, depending on exactly how the sensor setup works, it may not be an issue even if it does yaw. My initial idea of using some kind of linear sensor would have the advantage of preventing it from yawing. I may 3d print a mockup of the linear sensor setup and see if it’s possible to make to make it tight enough to work.
Correct. I actually have an idea about replacing the stepper motors with a mechanical system. It would remove the ability to do beep trim, but it would remove the need for the stepper motors and the associated electronics. In essence replace the stepper motors with a disc that is mechanically braked (thinking basically an inverse bicycle rim brake system). When braked it functions liked the locked stepper, when released it and the axis it’s on are free to rotate as desired.
I need to watch his videos in depth to understand why he uses mobiflight, just in case I’m missing something about the interaction of the stepper motors and the sensors.
Update on Mobiflight - unless I am using one particular function of the collective, mobiflight isn’t needed at all. It can make buttons way more useful in MSFS and some other sims (it can do way more than just keypress emulation), but ultimately I could just out put everything off the Arduino’s and Bodnar boards (and I’m thinking it might be possible to combine those with a higher end Arduino or Bodnar board).
That’s cool. I think the problem with it might be the same problem that one has with an xbox controller. Left/right and fwd/aft you get 100%. But at the corners you get something less. With helicopters this isn’t a problem. But in airplanes when maneuvering hard, it can be.
Rigging up a “magnetic brake” on that would be pretty simple too, same idea I mentioned above with basically a pair of spring clamps, you can pull open via a lever. Hmm, I need to get the sensor portion sorted, but that would be an extremely cheap build for a collective with only the sensors needing electronics. One old bicycle brake lever, some hardware store spring clamps, and some PVC or pipe for the stick and you’re in business.
I think I’m taking a long weekend next week, I might rig up something as a proof of concept as an experiment.
