Quote:
|
Originally Posted by ConKbot of Doom
Here is what I'm wondering, if the OI uses an internal resistor for the "bottom half" of the divider leg, if you were to use a joystick with 10k pots, would the current drawn by the pull down resistor make the output of the joystick that far off? I know it would add some non-linearity. But if there is a joystick you really want, couldn't you fix that in code? I know my saitek cyborg (my own personal one, were not trying to use it...) has the pots wired up as dividers, not rheostats like the white sticks we've had for the past two years.
Of course if you are really really bent on using a particular joystick, just program a PIC to use the ADC to sample the joystick axis value and output the value to a 100k digi-pot.  |
As Dave Flowerday pointed out earlier, the OI is a different beastie than the RC. Joystick ports are designed to take into account the need to produce a solid 127 if the joystick is disconnected from a port so that a robot doesn't drive around by itself. (The actual spec is less than a 0.05 volt input) The 100K on the OI is different from the discussion on analog inputs on the RC. The 1k OI analog input resistance is not a "pull down" but a current limiter. Since a 10k pot willl produce more current, it will be converted differently than the current through a 100k pot. The result is control non-linearity which can be overcome through software, driver practice and speed contoller calibration.
While we are discussing the OI, the current rules are specific about the devices you can and cannot connect to the OI input ports. A PIC controller that interprets an odd joystick and passes it along to the OI must use the joystick power, which is current limited.