Building a Load Tester

[lerman]

Has anyone out there besides me thought about building a load tester for servos? Here is my idea.

Get a big honking DC servo (larger than any servo you might want to test) that's rated at a speed higher than any you might want to run. I'll call that motor the load motor. Connect the motor under test to it (shaft to shaft). Mount the housing of the load motor in such a way that it can rotate a small amount and has an arm restraining the motion. At the end of the arm put a load cell so that the force can be measured. That's where you will measure the torque.

Now connect the encoder from the load motor to a circuit that will count pulses. That will be used to measure the RPM.

Connect the drive wires for the load motor to a high wattage low resistance through a PWMed switch. Nichrome wire is pretty cheap and you should be able to make a 1000 watt water cooled resistor pretty simply.

Oh, yes. We will also need a computer controlled power supply to vary the voltage to the motor under test. And an ammeter that we can read from the computer. We might want to also monitor the temperature of the servo. (Or, if we are testing an new AC servo drive, we would use that.)

By varying the duty cycle of the switch, we can increase or decrease the load on the motor. Connect all of this stuff up to a computer. Set the speed of the motor. Vary the load while measuring the torque and changes in speed.

It should be pretty simple to build. And pretty cheap if you can find a suitable surplus load motor.

Ken

[Evodyne]

Lerman,

Hi! Are you wanting a load tester-or a dyno?looked at what it would take: the simplest approach is a inertia dyno. Basically you tie our motor to a heavy flywheel (for which you know the rotational inertia) and apply 100% drive. You need to monitor rpm vs. time as the rotor accellerates. Here is one link, there are many others. What do you think?

Evodyne