How to find max power curve of motors.

The FAQ said Vex would have an expert evaluate the motors,
and come up with detailed numbers, but all they show is
~100 rpm and 6.5 in-lbs of stall torque.

From an design engineering point of view, ie trying to calculate the desired results before just doing “guess and check”,
it would be helpful to have a torque/speed/current graph, similar to that shown here:

Even better would be to calculate the Power curve and Efficiency curve,
and plot those as well, so we can know what motor speed provides the max power. However, there may be significant differences between motors which would make a standard table misleading, and it would be good to be able to measure this yourself, so you can check the RMA (Real Mechanical Advantage) vs the IMA (Ideal).

A cheap Centronics DVM ($4) from Harbor Freight on sale, as a shunt in the battery line (or PWM cable) will let you measure motor current, in case you are concerned with battery life or motor efficiency.

To find the power curve:
Connect a drum/pulley/wheel to wind up a string lifting a plastic bottle with various amounts of water (weight), and use a stopwatch to measure the time between two heights (=speed). With programming, two limit switches, and an optical shaft encoder, you could probably get more accurate results automatically.

WIth this data, you can measure the effective power output of each motor and motor+gearing combination. At light load and near 100 RPM, there is low power output. At stall condition, there is high torque, but low speed, which is also low power output. Somewhere between will be a maximum power = torque * speed. Once you know that, you can try to design (or guess and check) your robots to run at that motor speed to get the maximum effective power out of the motor.
For best reliability, dont forget to design in some torque margin in case your replacement motor is weaker, or the load heavier, or the gearbox less efficient, than you thought.
A homemade power curve table and graph in your engineering notebook would probably be a plus in notebook competition.

The person we asked to help measure the motor curves dropped the ball on us. If someone in the Vex community has time to measure/calculate the detailed motor specifications, we will get them posted ASAP.

Well, its not like it is that hard to do.
I’ve made a first pass characterization of one of my Vex motors, based on protocol derived from the BEST robotics website.
Protocol summary:
Motor+clutch with direct drive of cantilevered dual 4" wheels winding up a string guided by a bearing. I used a 9 foot lift, but timed only the middle 7 feet. For a weight, I used a plastic jug with various amount of water, weighed accurately to a few grams. Repeat each timed lift three times and take the average time. Controller setup was for tank-drive. Torque = wheel radius * weight. Dist/(2piR) = rev count = 6.8

Data summary:
No load (string only) speed was 110 rpm.
Stall torque with this setup was estimated at 4.3 lb-in, well below nominal.
The last weight in the series went slower and slower and then stalled.
Power curve was flat between 2 lb-in and 3.5lb-in, so if you aim for the lower end, you will have the same power, but more torque margin for emergencies.

  • RPM or RPS is easier to measure than torque;
    So for this motor and setup, any RPM between 50-80 will be close to max power output of the motor.

Lessons Learned: Next time build a bunch of gear combos between two rails,
and measure them all. Put the motors on a third rail, or at least use a longer axle so there is room to get access to the motor mount screws, so you can easily swap in each motor to compare them. Double check the controller setup to make sure you get 100% to the motors.

Fun things to try next time:
Different gear ratios (hypothesis: gears tradeoff torque <> rpm, but they always reduce power output)
Non-cantilevered wheels (hypothesis: bearings on both sides of wheels may cause less power loss)