Post by Frisbone on Jul 30, 2013 6:44:13 GMT -5
So on two separate related projects I could find (hacking the controller) the individuals were using PWM to emulate the Joystick. Since buying the PWM controller chip for this purpose and writing the driver code I turned my attention to the actual controller hardware. The sticks are made by a company called Alps and from what I could tell, all they consist of is 2 potentiometers and a push button.
I was racking my brain trying to understand how a PWM signal came out of that thing if its truly just a 2-axis POT. How could a source signal - even if its clocked have its duty cycle affected by a POT? Well, I figured out the answer today. It can't. Its just a POT.
It turns out that these guys are just simulating a pot using PWM. Rapidly changing the voltage on circuitry and smoothing it out with an RC circuit it seems will look like a variable voltage input to the Microsoft MC. So the input to it truly is a variable voltage - probably from 0 to 3V (maybe even less than 3V, haven't measured it yet).
So my extra expense in buying a separate PWM controller chip may have been pointless. Certainly I've been overly concerned with the actual signal - all I really need is a high frequency and a manipulation of the DC in any way I like.
However, a direct solution may make more sense. They make digital potentiometers - even for I2C buses - here is an example:
www.mouser.com/ProductDetail/Microchip-Technology/MCP4018T-103E-LT/?qs=sGAEpiMZZMuD%2f7PTYBwKqeKAyhhR3GLrtrc7NdAsyZg%3d
The only problem with this one is that it has just a 7 bit resolution which would produce 64 steps in each direction. Certainly I have much finer control with the 4096 points in manipulating the start/end points of the duty cycle selection with PWM. So the question is - how necessary is it that we have a high resolution?? Can your average player manipulate adjustments of a stick in more than 64 steps? Seems unlikely. A digital pot might make more sense - and it might consume less power.
Power consumption should be a big concern here since we ultimately would want this to run off of batteries.
I was racking my brain trying to understand how a PWM signal came out of that thing if its truly just a 2-axis POT. How could a source signal - even if its clocked have its duty cycle affected by a POT? Well, I figured out the answer today. It can't. Its just a POT.
It turns out that these guys are just simulating a pot using PWM. Rapidly changing the voltage on circuitry and smoothing it out with an RC circuit it seems will look like a variable voltage input to the Microsoft MC. So the input to it truly is a variable voltage - probably from 0 to 3V (maybe even less than 3V, haven't measured it yet).
So my extra expense in buying a separate PWM controller chip may have been pointless. Certainly I've been overly concerned with the actual signal - all I really need is a high frequency and a manipulation of the DC in any way I like.
However, a direct solution may make more sense. They make digital potentiometers - even for I2C buses - here is an example:
www.mouser.com/ProductDetail/Microchip-Technology/MCP4018T-103E-LT/?qs=sGAEpiMZZMuD%2f7PTYBwKqeKAyhhR3GLrtrc7NdAsyZg%3d
The only problem with this one is that it has just a 7 bit resolution which would produce 64 steps in each direction. Certainly I have much finer control with the 4096 points in manipulating the start/end points of the duty cycle selection with PWM. So the question is - how necessary is it that we have a high resolution?? Can your average player manipulate adjustments of a stick in more than 64 steps? Seems unlikely. A digital pot might make more sense - and it might consume less power.
Power consumption should be a big concern here since we ultimately would want this to run off of batteries.