Results 1 to 6 of 6

Thread: Analogue to Digital Joystick Control

  1. #1
    Join Date
    Sep 2004
    Posts
    1,905
    Rep Power
    21

    Cool Analogue to Digital Joystick Control

    I have tested a few Joystick Controllers on the PC, and I am very disappointed in the level of accuracy that I have experienced. The analogue joystick does not give a smooth range. For example, when you pull down on the joystick just slightly, there is no response in the y-axis output. Furthermore, when you pull down on the joystick with medium force, then there is a full scale response in the y-axis output. This is after the controller is calibrated on a Windows PC.

    When I disassemble the controller, I see a design that is not linear or accurate. The physical resistive force is on the x-axis and y-axis mostly. Thus when you move the joystick controller in a diagonal path, the force feedback that you feel is not as great or consistent as if you move the joystick straight down. The inaccuracy in the position comes from the use of just 2 analogue variable resistors, one for each axis. When the joystick is held down, one of the variable resistor "turns" and outputs an analogue signal. Somehow, this signal produce a y-axis output, but for the same push down on the joystick, there is a different y-axis output.

    A design in which the actual position of the joystick is transcribe into a y-axis output would be far more accurate. A design which uses a capacitive touch sensor or resistive touch sensor like in the diagram below would be much more accurate.

    image hosting

    I hope they build more accurate joystick controllers, and keep them at low prices.
    Let's act on what we agree on now, and argue later on what we don't.
    Black men leave Barbeque alone if Barbeque don't trouble you

  2. #2
    Join Date
    Sep 2004
    Posts
    1,905
    Rep Power
    21

    Default Calibration Issue

    There is an issue in "calibrating" in windows.

    Firstly, the joystick naturally moves in an area marked by a circle. However, it is given an area marked by a square in Windows. I would suggest a circular area to be used. That would be more natural to the actual movement of joysticks.

    Secondly, the joystick position in that square does not have sufficient resolution, for fine grain control of the joystick. I would suggest 2048x2048, or much greater for the resolution of the area in which the joystick can move. The current resolution is 256x256.

    This would be better suited for football, racing, flying, and first person shooter games.

    Furthermore, if the joystick is sufficiently soft and even (in terms of the spring tension being low and uniform in all direction), then even small children could easily steer characters in games.

    This is an example of the current calibration in Windows.









    images hosting
    Let's act on what we agree on now, and argue later on what we don't.
    Black men leave Barbeque alone if Barbeque don't trouble you

  3. #3
    Join Date
    Sep 2004
    Posts
    1,905
    Rep Power
    21

    Default Calibration Suggested Solution

    This is a suggestion of what the calibration window should look like. Note I tried to include an output of the current data (X and Y) for compatibility with existing games. Also, note that the new data output (Magnitude and Direction) has a higher resolution than 256x256. (256x256 is the current resolution for X and Y). The exact resolution decided on should be higher than the present hardware capabilities. Hopefully about 2048x2048 discrete values could be used. Note that the values for X and Y has to be limited or clipped once it is outside of the square area.













    images hosting
    Let's act on what we agree on now, and argue later on what we don't.
    Black men leave Barbeque alone if Barbeque don't trouble you

  4. #4
    Join Date
    Sep 2004
    Posts
    1,905
    Rep Power
    21

    Default

    Added default position



    images hosting

    I hope the controller/joystick that is much more sensitive and accurate gets developed at a low cost.

    I am personally tired of crappy controllers.

    Or expensive controllers.
    Let's act on what we agree on now, and argue later on what we don't.
    Black men leave Barbeque alone if Barbeque don't trouble you

  5. #5
    Join Date
    Nov 2004
    Posts
    5,192
    Rep Power
    25

    Default

    Personally I just use a PS2 to USB converter. But for the level of accuracy you want maybe something else is in order.

    Maybe something like this?
    https://www.amazon.com/Mayflash-F300...words=joystick
    Knowing the solution doesn't mean knowing the method. Yet answering correctly and regurgitation are considered "learning" and "knowledge".

  6. #6
    Join Date
    Sep 2004
    Posts
    1,905
    Rep Power
    21

    Default

    Well. I do have the Mayflash adapter for WiiU Pro controllers, but I have generic WiiU Pro controllers
    I also have the Mayflash adapter for Wii Pro controllers, and both generic and original Wii Pro controllers.

    Yes the accuracy I seek is a bit more.

    For example, if you want to draw a round path with a controller, in the windows calibration tool, then you will feel different levels of resistance as you move upwards, diagonally, horizontally. This is because of the 2 axes that physically pivot the joystick



    Ideally, when you make a circle with your thumb, your thumb goes through an infinite number of vectors from horizontal, to slightly horizontal, to diagonal, and so on. If each vector direction has a different resistive force to your finger, then that would add to the difficulty in making a perfect circle. That could be eliminated with a ball and socket pivot, with a uniform spring set up for resisting any pull on the joystick. Typical controllers have one wheel and axle pivot embedded in another wheel and axle pivot to simulate the full 2-dimensional movement of the joystick. A ball and socket pivot already supports 2-d movement, and it is perfectly uniform. Try to feel it when you try to make a medium size circle.

    It is also difficult to arbitrarily put the joy stick in a position that could give you a desired output. For example, if I want an output of the calibration tool be be X=170 and Y=40, I should be able to get that if my hands are steady, or even, those values +- 1, and hold the controller at that position. And if I choose another pair of arbitrary values for X and Y, then I should be able to find them in 5 secs and so on. For any arbitrary values in the entire range, with ease. This would be better aided with a small digital matrix that can translate position to a sensor value like a capacitive touch sensor. The current var-resistor sensor is not that consistent.

    Imagine playing a football game, and be able to kick the ball into any side of the goal no mater how far away your player is from the goal, just because the resolution in the joystick is present to make that possible. Even racing games would feel much more accurate. But it is especially needed in flying games which require you to be delicate with the movement on the joystick
    Let's act on what we agree on now, and argue later on what we don't.
    Black men leave Barbeque alone if Barbeque don't trouble you

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •