Oricom Technologies
www.oricomtech.com
HOME | PRODUCTS | ORDER | LINKS | ROBOTICS || PROJECTS ||

Return to:
Projects page.
Servo Load Measurements

The measurements described here were taken on a prototype walking platform, Gimlee-U8, which contains 16 servos total on 8 legs. In the measurements shown, 15 servos were active and performing a characteristic walking gait, with the legs in the air - ie, the total effective loading was small, and the servos were mainly going through their paces. The high-low variations are related to the fact that about half the time, half the servos are not moving - specifically, the knee joints are locked during the power stroke of each leg.

This figure shows the battery voltage out of an 8-cell pack of "AA" cells (NiMH, 1800 mAh).

The battery voltage drops significantly under heavy loading due to internal resistance in the cells.

[battery voltage]
The battery pack voltage was knocked down to 5v, for application to the servos, using an LM1084 5A voltage regulator.

The output variations are likely due to the variations on the input dropping down below the dropout voltage for the regulator.

[v.reg out]
This figure represents the current flowing out of the regulator and into the servos, as measured across a 0.05 ohm resistor placed in series in the ground lead to the servos.

Peak current was approx 0.22v/0.05ohm = 4.4 Amp, and average current 2.2 Amp.

[servo current]
Since the peaks correspond to times when all 15 servos were moving, the current per servo was about 4.4A/15 = 290 mA. Remember, these measurements were for light servo loading. Actually, 12 servos were CS-71s (std torque) and 4 were CS-80 (high torgue), so the latter probably drew a little more than this.

Average power dissipation in the regulator was about (8.3-4.8v)*2.2A = 7.7 Watt. It became rather hot, despite our having used a moderate-sized heatsink [0.5"x0.5"x0.75", 24.4 degC/W].

<| TOP


© Oricom Technologies, Feb 2003