jpanhalt
Advanced Member level 3
I am using a Memsic MXD2020EL (100 Hz) accelerometer and a PIC12F683 at 4 MHz to develop an inclinometer for pallet forks on the front of a tractor.
The MXD2020EL outputs a PWM signal whose duty cycle varies with tilt relative to Earth. One can input the PWM signal directly ("digital" input) or use an integrator/low-pass filter to convert the digital signal to analog, and then input that through a 10-bit A/D converter. If I use the digital approach, I will probably use polling of the input pin to capture the data from which the duty cycle will be calculated.
While it is tempting to think the digital input might be more accurate and hence better, I am not so sure of that, even in a static case. Moreover, the front-end loader on a tractor is not a smooth ride. The inherent filtering provided by the analog approach might be an advantage.
At this moment, I am making a board that will allow me to test both approaches. I am asking here whether anyone has experience with similar thermal accelerometers and can relate their experience or advise relative to the two methods of capturing the data.
Regards, John
The MXD2020EL outputs a PWM signal whose duty cycle varies with tilt relative to Earth. One can input the PWM signal directly ("digital" input) or use an integrator/low-pass filter to convert the digital signal to analog, and then input that through a 10-bit A/D converter. If I use the digital approach, I will probably use polling of the input pin to capture the data from which the duty cycle will be calculated.
While it is tempting to think the digital input might be more accurate and hence better, I am not so sure of that, even in a static case. Moreover, the front-end loader on a tractor is not a smooth ride. The inherent filtering provided by the analog approach might be an advantage.
At this moment, I am making a board that will allow me to test both approaches. I am asking here whether anyone has experience with similar thermal accelerometers and can relate their experience or advise relative to the two methods of capturing the data.
Regards, John
