Disclaimer: New to this forum; I didn't lurk much yet. I might be doing something already tried (and failed :P).
I've got into building a RepRap (a Prusa 2 FIY) myself, but found the electronics to be quite expansive. So, being a fan of BEAM philosophy, I took on the adventure of trying to do away with microcontrollers (and maybe, eventually computers) to control the RepRap.
The following circuit is what I came up with.
[I'll add pictures at a later edit and schematic if you are interested]
Basically you only need a stepper motor (actually an array, but bear with me) to drive and a clever way to set it on and off and make it rotate prograde or retrograde at the right time. So first, there an OpAmp Phase Shift oscillator stage that generates the sine and cosine waveforms that make the motor rotate. These signals are redirected and switched by a low-voltage relay contraption which is driven by a computer with MOSFETs into two push-pull MOSFET amplifiers (each for each waveform). The motor connects to the amplifier's output.
The computer is a Raspberry Pi running a python script which I won't show as of yet under the embarrassment of its... say, incompleteness...
The GPIO of the Raspi puts on high one of its pins to make the motor rotate in one direction and turn other pin high to make the motor rotate the other direction: both pins low means the motor is stopped by means of no relay allows the oscillator waveform to get to the amplifier.
Does this work? Well...
Given I'm not as much expert in electronics as I wanted, the fact that I only have one stepper motor yet and that this circuit seems to be unreliable 40% of the time... What's you measure of success?
I believe I should be using bipolar transistors instead of MOSFETs, but I couldn't get them work for some weird reason. Also the fact that I'm not using a microcontroller and instead rely on direct computer control can be argued against by saying that computers don't work in real time or are fast enough, but since I've seen some guys remote controlling a quadcopter with a Raspi, I believe anything. I plan to use parallel processing to control simultaneously the various motors and bypass computer control for bed and nozzle heating, by instead using looped feedback circuits (maybe with electronic neurons or differential amplifiers and such). This is unless someone points out a better idea!
I've got into building a RepRap (a Prusa 2 FIY) myself, but found the electronics to be quite expansive. So, being a fan of BEAM philosophy, I took on the adventure of trying to do away with microcontrollers (and maybe, eventually computers) to control the RepRap.
The following circuit is what I came up with.
[I'll add pictures at a later edit and schematic if you are interested]
Basically you only need a stepper motor (actually an array, but bear with me) to drive and a clever way to set it on and off and make it rotate prograde or retrograde at the right time. So first, there an OpAmp Phase Shift oscillator stage that generates the sine and cosine waveforms that make the motor rotate. These signals are redirected and switched by a low-voltage relay contraption which is driven by a computer with MOSFETs into two push-pull MOSFET amplifiers (each for each waveform). The motor connects to the amplifier's output.
The computer is a Raspberry Pi running a python script which I won't show as of yet under the embarrassment of its... say, incompleteness...
The GPIO of the Raspi puts on high one of its pins to make the motor rotate in one direction and turn other pin high to make the motor rotate the other direction: both pins low means the motor is stopped by means of no relay allows the oscillator waveform to get to the amplifier.
Does this work? Well...
Given I'm not as much expert in electronics as I wanted, the fact that I only have one stepper motor yet and that this circuit seems to be unreliable 40% of the time... What's you measure of success?
I believe I should be using bipolar transistors instead of MOSFETs, but I couldn't get them work for some weird reason. Also the fact that I'm not using a microcontroller and instead rely on direct computer control can be argued against by saying that computers don't work in real time or are fast enough, but since I've seen some guys remote controlling a quadcopter with a Raspi, I believe anything. I plan to use parallel processing to control simultaneously the various motors and bypass computer control for bed and nozzle heating, by instead using looped feedback circuits (maybe with electronic neurons or differential amplifiers and such). This is unless someone points out a better idea!