Test Overview:
Purpose of the test: to verify new motor actuator set up and see if the EC90 600W can output larger required torque values.
3 tests will be performed at each weight and the current will be measured: static at 90 degrees (parallel to floor), velocity moving CW (upwards) from 90 to 0 degrees and stopping at 0 degrees, and velocity moving CCW (downwardsd) from 0 to 90 degrees stopping at 90 degrees.
Set up axis 2 test rig with odrive, encoder, and hal effect sensor with 2 limit switches mounted at 0 and 90 degrees triggered by the bolt. Commands will be sent through CAN commands. 24V power source
Important Specs:
Part | ||
|---|---|---|
Maxon EC 90 Flat 600W 30V | Note: Since we’re running it at 24V, spec sheet info might not reflect actual results Nominal Torque: 1.56 Nm Nominal Current: 10.5 | |
GAM GSL-CS-020 |
Motor: Maxon EC 90 Flat 600W 30V
Weight (kg) | Equivalent Torque (90 degrees, Static) | Expected Current (A) (T/120*0.138) | Current (static ad 90 Degrees) (A) | Current (with -2 velocity) (A) |
|---|---|---|---|---|
5.5 | 27.4 | 3.8 | ||
8.1 | 40.4 | 5.6 | ||
10.4 | 51.8 | 7.1 | ||
12.7 | 64.3 | 8.9 | ||
15 | 74.75 | 10.3 | ||
18 | 89.7 | 12.4 | ||
20 | 100 (Max Acc of Gearbox) | 13.8 | ||
25kg | 130 | 17.9 |
Test Procedure:
Hooked up axis 2 with Saheed’s testing rig consisting of 0.508m metal arm attached to axis where load would be applied that corresponded to torque being applied on the gearbox + motor.
Connected axis to power supply, odrive and hal effect to input closed loop into arm when testing
Current draw from Odrive is monitored from the power supply while current draw fro the motor would be examined from current probe
Test would first measure the current draw at static to just hold up the load, then velocity control of -2 would be applied and highest current draw would be taken
Test results:
Motor performed fine at 8kg and 10 kg, which were below nominal range. When ran at those currents for an extended period of time the motor would heat up, suggesting need for active cooling
When testing at 12.8kg (nominal) motor struggled to lift the load, behavior consisted of “breaks” in movement as it lifted it up. The current draw is not consistent; it spikes and then returns → suspect close to the motor limit will observe unexpected behaviour. (note: Arm can hold just cannot lift) [Smells bad → motor insulator might be damaged]
When testing it at 25kg (competition requirement) the motor could not hold up the weight at all, would collapse in closed loop static loading. Potentially, we could have burned the motor, but when measuring the resistance of phases and running it afterwards, we did not see abnormal behavior or values.
Notes:
Axis 1 was accidentally dropped during testing, motor appears to be broken.
The axis 2 motor could potentially be broken; further examination is required. We might need to buy 3 more motors, and further motor spacing may be done.
Yuming believes we should proceed with weight reduction and torsion spring. Torsion spring could take the weight of the arm leaving the motor to take the weight of the load, which would put it within the nominal range of the axis (61.31 Nm max). Further discussion with robotic prof to learn more about torsion spring implementation and discussing other options such as motor respec needed.
Yuchen: I have disassembled the Axis1 arm due to the motor damage will need to replace the motor; the motor is currently on the table. I also disassembled motor 2 and will need to do an RCA on it. I personally don’t know how to open the gearbox, so I need a mechanical member to help me with that task. I think, based on the observation from yesterday, we want some redundancy + 10% in the motor respec to counter the effect of any potential environmental factor(temperature + long duration)
Yu-Ming He Next time, this will be something we really need. The test plan also need to say what is the hard current limit for the motor, and i will pair up that value on ODRIVE so we are not likely to burn anything up