Jump to content

Sergey_Pe

Members
  • Content Count

    69
  • Joined

  • Last visited

About Sergey_Pe

  • Rank
    Junior Member

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. In case of VKB everything can be fixed (seriously!). The magnets on the axes that are used in the sensors are glues to the shafts (threaded pins) of the axis bearings. In extremely rare cases (especially with an extension/ heavy grip/ strong springs) one of the shafts can get loose so that the shaft + magnet start rotating, hence the calibration shift. You'll need to take the gimbal mechanism out of the body, remove a Y-sensor (2 screws) and tighten the nut that holds the Y-axis shaft in place. Here is the photo; I don't have a Gunfighter, this is a previous VKB gimbal, but it's very simi
  2. Sokol is keeping an eye on everything interesting and useful coming up in simulation area. And he is always eager to share his findings which is great!
  3. Just to clarify- with all my huge respect to Sokol1_br, who is doing a lot for the community, MMJoy2 was written by a different person. It's Mega_Mozg, who is now a chief software designer at Virpil Controls.
  4. All the supported A/D modules are listed under the "Source" tab in MMJoy configurator, and these are MCP3201-3208. MMJoy firmware will not be updated as the project is closed and the developer is now working for Virpil team.
  5. All the axis have to be calibrated. If you choose "auto" then you need to move all the axes to the extremes every time you connect the joystick to the computer which is inconvenient (from my point of view). I prefer "save with center" option which has to be applied once using the "calibrate helper", and don't forget to save the settings to the device as you do with axis/ button assignments.
  6. How did you calibrate these axis? The Axis to Button function is using the raw axis range that needs to be set first. If "auto" calibration mode is used, then the range is set only after the joystick is moved to the extremes. Try using the "saved" mode, calibrating and saving the range first.
  7. It looks like a part of your text is hidden behind the picture. Are you using the short-press buttons or ON-ON switches? It might be that you are missing the logic behind the different logical button modes. "Switch On" means that when a physical button is pressed or held there will be a brief logical button press and then immediately a release. Typically you won't see it in Windows game controllers setup screen as the press duration is very short. Try a Button (norm) mode instead- it will deliver a logical button press as long as you have a physical button pressed.
  8. Hey rel4y, No I haven't. So far I used the blue pill's internal ADC with AS5600's in a throttle quadrant project without any noise-related problems. Maybe I'm lucky with the USB ports on my PC... Anyway I consider the internal ADC as an auxiliary functionality- TLE5011 (or TLE5012 after it's included in the firmware, as promised) would be my sensor choice. And yes, I'm using Sprint Layout but no plans for any FreeJoy shields so far. Best regards
  9. Wrong. Calibration always needs to be done for the actual angle- this is exactly what it's done for. Otherwise you won't get a full range of OUTPUT signal for the particular axis. However the INPUT signal with the 360 deg. sensor will be much smaller, so much of the ADC resolution will be lost, as you correctly mentioned. I would strongly recommend using AS5600 sensor- calibrating it after installation (before MMJoy calibration) you'll get a full 0-5V voltage swing for any angle more than 18 degrees without any hassle with the magnet/ sensor positioning as with SSxx or Allegro sensors.
  10. On your second screenshot- DZ boxes. You'll need to set "DZ center".
  11. FreeJoy project. I've switched to it from MMJoy2 (which is very good) for my own projects. More flexible, works with the same sensor types as MMJoy, much more user-friendly configurator software, well-documented. Constantly evolves based on community feedback, but is 100% stable already now. The board (STM-32 based) is as cheap as ProMicro.
  12. Yes, but it requires changing the controller as well. However it will give you better resolution- minimum 10 bit or 1024 steps with MMJoy2 controller that this guy is using. And this controller is way more flexible compared with Saitek. You can put in the additional buttons, play with the axis parameters etc. He has also installied the magnet, cutting an upper part of the lever bearing. This will lead to an additional "play" of the lever as the leftover part wouldn't sit tightly in the bearing anymore. You'll need to put some additional bushing there which again raises the question- is Sait
  13. 256 steps in total- this is the resolution of Saitek controller. When you use a regular potentiometer (or an analog Hall sensor) the resolution fully depends on an ADC (Analog-to-Digital Converter) at the controller's input. But to utilize this resolution you need to ensure that the sensor provides a full power supply voltage range to the input. If it is less than the resolution will degrade proportionally. This is why its's very important to use a pot with the same working angle as the original one- smaller angle will reduce a throttle lever's working zone, bigger angle will degrade the resol
  14. True. But pricey, although you can order practically any configuration according to your own layout. And the built quality is perfect- I'm using his flaps control mechanism, it's excellent.
×
×
  • Create New...