Managed to do some testing on my motors today. I took out the bolts from the end plate that carries the brushes and used a couple of sash cramps to clamp the brush plate at various angles to see what effect advancing and retarding the brush timing has on rpm.
To advance the timing you rotate the brush plate in the opposite direction to the rotation of the motor, and to retard it you rotate the brush plate in the same direction as the motor rotates. Advancing a motor increases the voltage it can be run at without severe arcing or flashover at the commutator. It also increases the rpm and reduces the torque slightly. To run these motors at 108 volts rather than the rated 36V, it's been recommended that I advance the brushes 8–10 degrees. (This is stuff I've gleaned from slightly obsessive reading of everything I can find online, so if anyone knows better please shout
![Smile :)](./images/smilies/icon_smile.gif)
)
The resolution of my rev counter isn't great but it's good enough to show a trend in readings. I adjusted the brush plate angle in 2.5 degree steps to +10 degrees and then -10 degrees, and measured the rpm with the motor rotating clockwise and anticlockwise at 12V with no load. It turns out that the motors consistently rotate about 10% faster clockwise than anticlockwise at all timing angles. It also looks as if the motors were originally neutrally timed, as rpm increased gradually as the timing was advanced for both clockwise and anticlockwise rotation.
The difference in speed in either direction is a bit of a pain, as I want to run one motor clockwise and the other anticlockwise. I'm hoping though that this is just the result of small differences in brush pressure due to the angle that the brushes are mounted at. When the motors are under load this difference will hopefully become insignificant.