Goal

• To learn more how engine/motor performance is measured
• Create a basic cheap dynamometer for home project use

Current Design

Components/Sub Assemblies

The dynamometer is currently made up of the following:

• A Feather 32u4 uSD Adalogger with 128×32 OLED Featherwing Display Module – Allows for a quick and simple interface and control of the ESC
• Electronic Speed Controller (ESC) – This is specifically for the BLDC motor under test, and will need to change dependent on the motor’s needs.
• Housings
• Coupling – A TPU connection between the test and brake motors
• A Brake Motor – Provides a variable load and an arm to engage the load cell
• External Power Supplies – For the Feather board and motors.

What Performance It Measures

A motor has several different parameters that define it’s operating point and performance.
The common items provided, for a set voltage, by manufacturers include:

• Current, I (A)
• Output Power, P (W)
• Torque, T (Nm)
• Efficiency, n (%)
• Speed, (rpm)

Typically this is represented on a graph with Torque on the x-axis and everything else on the y-axis.

An example of a motor’s performance curves.

How it Measures

The principle of operation is as follows:

• An operation voltage is chosen (and ESC speed if applicable)
• The equipment is set up as in the picture below, including the motor to test.

• The brake motor provides a variable load for the motor under test
• With the test motor running the load applied by the brake motor is increased
• At various points recordings of measurements are noted for:
  • Current
  • Force on the load cell
  • Speed
• Once you’ve swept the brake motor from no load on the test motor to zero rotations (ideally) the test can be stopped.
• The force on the load cell needs converting to Torque, with Output Power and Efficiency calculated from the measured values.

Video of the system running.

Future iteration ideas/intent

The current setup needs more to become a “usable tool” in my workshop.
Some of the current ideas and thinking I have for upgrades are:

• Integration with power supply
  • Connect to my RS Pro power supply and automatically log the voltage and current linking with the other readings.
  • To integrate this I need to understand what the ports and protocols are capable of communicating and designing how hardware and firmware can be added to incorporate this.
• Integration with the load cell
  • Eliminate user steps and auto read the values for logging.
  • An approach to coordinate with the displays clocked signals needs to be figured out.
    Possibly using the displays row signals to drive interrupts (see the post for more info).
• Integration with speed sensor from the brake motor
  • The brake motor includes a PCBA that is able to measure rotational speed.
    Leveraging this directly would simplify work needed to get these values, or be used as a leaping point for my own design.
• A computer interface
  • Being able to plug into the dynamometer with USB
  • A simple GUI that can be used to
    • start/stop logging
    • drive the ESC
    • drive the brake
    • generate plots and tables of results

Reaching The Current Design

Initial design inspiration came from Jeremy Fielding‘s video of his dynamometer setup: How to Measure Horse Power for Any Electric Motor

Some Versions Made

Housings

• Initial support for the brake motor was too thin.
  It allowed lots of movement and vibrations as the motor could effectively pivot on the walls of the supports.
  • I glued on some plastic sprue I had, this helped a little but not enough
• The main improvement was thickening up the wall so the motor was more limited to:
  • Roll around the axis
  • Lateral movement along the axis

Coupling

• Initial designs and assembly approach were poor
  • The couplings were very rigid which which contributed to lots of vibrations
  • Assembly was hard to get good mounting and alignment
    • Threaded inserts became offset to the axis
    • Press fits became angled to the axis
• Changes made
  • The coupling between the test and brake motors material was changed to TPU. This was the main reduction in noise and vibration
  • An alignment jig was made for assembly, simple but super effective,
    Was just a printed tube that restricted the angle the coupling could be when pressed onto the shaft.

Lever arm

• Single side to dual side
  • Simple change, was already printing other parts so why not
  • Allowed more flexibility on load cell placement

Load cell

Intent was to use a cheap set of hacked kitchen scales to provide the required force measurements.
This development is captured in a separate post here.

Until next time.

H