SERVO 08.2016 35
SimonK. This forces the robot's speed to track your radio stick
position tightly. The robot won't coast, and as a result, the
controller does not have to fight the motor just to slow it down.
Finally, to improve low speed performance further, I read
the transistor switching times on an oscilloscope and turned
the minimum PWM duty cycle of the controller as low as it
could go while still turning the transistor on. This meant the
motor could spin very slowly to start with, giving the mass
of the robot more time to accelerate.
The combination of these modifications resulted in a
250 lb drivetrain that closely mimicked how Overhaul drove,
using motors that weighed only four pounds each. There is
still significant work to be done, but I think that with tuning
to your needs, you can get 99% of DC motor performance;
the one exception being "holding torque" where the motor
isn't moving but has to apply torque. For that, more complex
control schemes than what SimonK and BLHeli can offer are
It also necessitates a slightly different driving style,
which I would describe as "avoiding center stick;" meaning,
trying to avoid having the robot stop and start repeatedly,
and instead choosing to turn instead of stop. Luckily, I was
Replacing the stock low quality capacitors with high grade
The system piloted by Sadbot installed fully in Overhaul 2.
Will it perform?
Scoping the rise- and fall-time of the transistors to adjust
the minimum running speed and braking power as closely
as I can.
A Turnigy SK3-6374 motor mated to a BaneBots P80
gearbox comprises the drive unit for Sadbot and
Sadbot, the 250 lb drivetrain mule for Overhaul 2.