After flying with a spring-loaded reinforced gimbal for a while, I noticed that slight backlash has developed in a stick connector part. While it was absolutely tolerable and that part can be replaced easily, I thought it was a good idea to design a new cyclic frame for those of us who value absolute precision over being able to remove the stick when not in use by simply pulling it off its socket. The new stick frame is based on M8 bolts and mounted to the gimbal with nuts. The cool thing about it – you can easily adjust stick grip inclination (you can make a copy of AS350 stick for example) and fine-tune its dimensions to suit your anatomy. It can still be attached or detached quickly but requires 2 wrenches to tighten/loosen nuts properly. I think it will be a new recommended default for a spring-loaded cyclic.
I have finally been able to go to the local car parts shop and buy some springs for the gimbal. Here’s how it looks now:
I should start by saying it works just great. These springs are from some Lada’s clutch (i think they’re for the pedal), they are 55mm long when compressed and about 11mm wide. I am pretty sure any similar spring will work great, just pick one with light or moderate tension.
I have added modified holders for the reinforced gimbal that Guido has asked for. Sadly I can’t test them myself, so please tell me how these mounts worked for ya mate =)
There are 2 versions: the “a” version has the same distance from magnet to the sensor as 6x6x4 one, and in the “b” version, a magnet socket is lifted a little for an extra tolerance (the magnet is 1mm wider so potentially can touch the sensor with its angle when turned, tell me if you encounter this issue).
After finally getting some long enough sleep I thought again on my failed attempts with a Melexis sensor. The sensor itself is quite precise and has some DSP onboard, so I tried it again with a spherical gimbal and noticed that depending on how close a magnet is to the sensor, it outputs not a circle, but a square, inclined to 45 degrees. I thought, “- what if I simply cut excess values and make it a square?”, and did just that. The idea behind it was that jitter of the point in a joystick tester sketch seemed minimal, so the physical precision of the sensor looked like being enough to work well. As long as we have a 15 bit external ADC that powers our gimbal, we should still have something like 4096 points per axis even with this reduced range, which is plenty.
Now, if we use a disk magnet instead of a square one, unarrested twist axis is not an issue anymore, it doesn’t affect X and Y positions in a big way. So this design seems to be perfectly fine for vertical sticks, or sidesticks, and I think with a pneumatic mod it can be as good for everything else. Its beauty lies in its mechanical simplicity, only a few parts are needed, it can be quite small in size.
I think I will make, test and publish both versions so everyone will be able to choose something for himself =) There’s also a third version, the 608 and MLX based one, its fate will depend on a degree of success of the spherical bearing based one =)
Meanwhile, some progress on the 608 and SS495A based version:
I am now working on two reinforced gimbal variants. One of them is based on an ordinary 608 skateboard bearings, another one – GE25 ES-2RS spherical bearing. Both of them will be able to handle springs, rubber bands, pneumatic cylinders, etc. A CJMCU-93 module (MLX90333) is used as a magnetic field sensor.
The fate of GE25 ES-2RS one depends on whether I will be able to prevent it from twisting or not (i have an idea involving _| shaped bracket, 2 pneumatic cylinders, and 2 spherical heads now). If you have some ideas about how to lock “twist axis” of a spherical bearing, please share them in comments =) Here are some pictures of the GE25 testbed and 608-based design.
You can now move cyclic around while trimmed with a hat switch of a B8 stick. This is how it works in a real 407. You can do it in X-Plane with key assignments as well, but Simchair handles this in a better way with its adaptive force trim option. This option will not release trim until cyclic would be in a vicinity of an initial trim position with a given deviation. You can disable this feature when it’s not needed by setting B8_HAT_SWITCH_MODE to “HAT”.
I have been testing new hardware more thoroughly and noticed, that different sims have very different controls sensitivity settings (and ways to adjust it). More than that, some of the sims tend to react on even the slightest jitter from Hall effect sensors and ADC (which is really almost nonexistent), while others ignore it completely. That means that often, there’s a need for per-sim or even per-aircraft setting.
Simchair MKIII cyclic and pedals are sensitive enough to emulate the control response of lightest and twitchiest helicopters out there. But what if you feel it is too sensitive for the particular helicopter model you want to fly? Or, maybe the sim you are flying in treats joystick sensitivity settings differently, or even doesn’t have proper controls settings at all? What if setting curves is not what you want, but your sim doesn’t have any linear sensitivity adjustment options?
While experimenting with adjusting controls sensitivity for different sims, I have found an interesting thing: in fact, the ADS1115 is so precise its values can be used without filtering at a 15-bit resolution! It is rather impressive 32768 points per axis. I can’t say the difference is as huge as between 8 bit and 12, but it depends on the game. For example, while I only felt a marginal difference in DCS and X-plane, it was very noticeable in Arma 3. That’s why I added an ADS1115_RESOLUTION constant to the master controller sketch.
You can set ADS1115 resolution like that:
#define ADS1115_RESOLUTION 15 //from 12 to 15 bits
I am now curious if changing a connection scheme of an ADC to a differential one will give even better results!
You can download an updated .ino file from GitHub.