How to measure the rotational speed of one propeller
Recently we suspected that one motor of our hexacopter had been damaged in a crash. Since removing the motor requires almost the complete disassembly of the copter I started thinking of ways to test the motor in place.
The idea of using a laser and a photodiode came pretty soon. But there were a couple of small details, that made the whole setup very simple and efficient.
Mechanical setup
A small “Eureka!” moment came, when I realized that I didn’t have to mount the laser diode directly over the spinning propeller, but instead could mount it to the side and shine it through the propeller at an angle. Since I was already nervous about building a support structure over the “spinning blades of minor bodily harm”, this helped simplify the build a bit.
you can see the laser diode attached to the beam of the clamp and the photodiode on the table secured with masking tape
Electrical Setup
The setup is fairly simple, fortunately I still had a 3.3V red laser module and a red sensitive photodiode laying around from a different project. The laser is connected directly to the power supply, the diode is connected in reverse and an adjustable resistor is added between the anode and ground, to provide a voltage divider. The middle of the voltage divider is connected to channel 0 on the Open Bench Logic Sniffer. The resistor is then adjusted, so that the Voltage is less than 1V when the photodiode is dark and more than 2V when the diode is lit.
note the reversed photodiode
Computer setup
On the computer side of the whole experiment I used the stock OLS software from jawis. I chose to trigger on a low signal of channel 0 at a sampling speed of 500 kHz. This allowed me to fit at least 4 periods of the resulting signal into the memory. 4 periods correspond to 2 revolutions of the rotor, since the blade will interrupt the beam twice per revolution. A nice side effect of this is, that you can see some irregularities in the motor by just comparing the two halves of one revolution to each other.
My main goal however was to compare the speed of the suspect motor to the other 5 presumably ok motors. The OLS client has a neat feature, where you can set markers in the captured signal, and it will show you the frequency of the signal.
1.5 revolutions of the rotor at a 89% speed setting. 62.5 kHz corresponds to a tip speed of 230 km/h
It turns out that the motor showed no anomalies, and some careful further testing supported this finding. Our guess is that previous problems were just the result of bad calibration on our part.
