Traumflug
commented
10 years ago After replacing the capacitors on my Gen7 board with 3300uF capacitors, the 12V rail now dips to 11.7V under load, and the 5V rail does not dip at all (±5mV). To my surprise, the motors are also noticeably quieter now, and I cannot explain why this has reduced audible noise.
That's a very interesting observation, thank you for sharing it. I have to try myself. Definitely on the TODO list.
To prevent dangerously high LC spikes, Pololu recommends placing a 100uF capacitor across the rails of each driver, as outlined here.
Currently, a single 100uF capacitor decouples all four (or 5 if you have two z-axis motors) stepper motors, and it may be wise to spec for a much larger capacitor with +1000uF, such as this one. While a high value cap is more expensive, this should attenuate LC spikes and provide sufficient instantaneous current for the motors.
My Gen7 is being powered by a standard 500W ATX power supply and my average power draw is below the power supply maximum specs. Using an oscilloscope, I found that my 12V rail was frequently dipping below 11.2V when my printer was moving. This was also introducing noise to the 5V rail, which would dip below 4.5V during movements.
After replacing the capacitors on my Gen7 board with 3300uF capacitors, the 12V rail now dips to 11.7V under load, and the 5V rail does not dip at all (±5mV). To my surprise, the motors are also noticeably quieter now, and I cannot explain why this has reduced audible noise.
Of course, my power supply is partially to blame for these issues, but it still might be a good idea to increase the value of the motor decoupling capacitors (and perhaps heaters).