dynamometer
n : measuring instrument designed to measure power

• I needed to test the performance of some small shaded-pole AC motors.
• I could not find a local lab with zero turn around time so I made this setup.
• The method that I used is to provide a variable brake which extracts power from the motor and burns it as heat.
• The force which is collected from the motor by the brake is transferred to a digital gram scale by an aluminum arm.
• I made up the arm to measure exactly 100mm from the motor shaft center to the point which presses down on the scale.
• The length is important if you want to get a measurement that can be referenced to a known quantity like Horsepower, Pound*Feet, running torque (not stalled) Footpounds per Minute") or Newton*Meters
• This type of measurement can be useful of you wnat to determine how fast your car can get to a certain speed, how much your robot can lift, or how much water your pump can pump.
• In this case, I wanted to know the performance of a small motor made in China and it's US (mexico, actually) made counterpart.

The complete setup

This photo shows the test setup (except for the variable lab supply that drives the brake and the variac that holds the 120VAC input constant). Clockwise: True RMS Watt Meter which indicates input power to the motor. The motor being tested Scale which measures the force transferred from the spinning motor by the hysteresis brake. Laser tachometer. It shines on 2 pieces of tape - one black and one reflective to measure the motors RPM. Hysteresis brake. It provides a variable load to the motor.   Load (drag on the motor) increases as the voltage applied to the brake goes from zero to about 30VDC

Brake and Motor

Here is the Magtrol hysteresis brake. (Gold cylinder at left) I got a few of them from a Lucent auction. They used the brakes to regulate the tension on fiber optic fibers during the time when they were getting the PVC jacket extruded. Since the brake is kind of heavy I just let it roll in place. The rolling uses zero measurable force as detected within the range of this setup. Note the fan used to keep the motor(s) under test from overheating.

Lever Arm Lever Arm

Here is the arm that I brewed up to transfer the brake force to the scale. The black tape is a small vibration dampener.

Action shot! DSCN9427.JPG

Here it is in action. In the background is the Watt-meter showing 66W input at 120VAC (60Hz) Note the gram scale shows 48 grams and the next photo shows 3016 RPM.

Since my lever is exactly 100mm long, I can just divide by 10 to get the gm/Meter force.

For the complete test I operated the motor at the design voltage of 120VAC / 60 HZ I started the motor with no brake current and recorded the RPM and power draw. I increased the brake current until the motor slowed by 25RPM, I then recorded the RPM, power draw and force on the scale. I repeated the above over and over (slowly increasing the brake current) until the load was great enough to stall the motor. The following is the chart generated from the data entered into Excel. The max HP of the China motor was 0.0206 at 2875 RPM compared to the US motor 0.031HP at 2900 RPM. The last data point was at the motor stall. If the trend had not been downward, I would have needed to add feedback so the the brake current would regulate to hold the speed into the decreasing (foldback) area of the curve.

OEM China Peak HP 0.030752704 0.020589025 Peak Torque 11.1 7.64 No Load RPM 3500 3450 Peak Load RPM 2900 2875 Peak Output Watts 22.94151688 15.35941233 Weight Grams 969 954 Coil Resistance (ohms) 5.7 7.9 Stack height mm 29 33 Stack width mm 60 60 Stack depth mm 63 61

The above indicates that there is about the same amount of material in each motor. Since the shaft is slightly longer on the OE motor, the difference is moot. There is a higher winding resistance in the China motor which is one of the reasons that the torque is lower. The stack height of the China motor is larger which would normally indicate a more powerful motor. The cause of the difference is most likely higher winding impedance (smaller wire or more turns) possibly required due to inferior steel or maybe it's just a design oversight.

! WARNING SALES PITCH ! I HAVE A FEW OF THE SMALL DC OPERATED BRAKES AVAILABLE AT $20 USD ea. (plus shipping) They seem similar to Magtrol HB 8, but bear the marking HB-100-2 (I guess that they were somehow customized for Lucent) http://www.magtrol.com/brochures/hysteresis.pdf The shaft size is 0.1875 dia. They are used, but spin and operate ok. I can take paypal Future Enhancements to the dyno: PLL feedback of the brake current. The motor would need an RPM take off. The phase detector would have a reference frequency input and the motor RPM. The Phase Detector filter output would feed a power amp that drives the brake. The RPM would be set by the reference frequency input. (RPM = Fref*60 for a one pulse per rev encoder) For really nice system you would use a 60 PPR encoder so Fref=RPM. - or - Analog Error amp method. The motor would need an RPM take off which feeds an F to V converter. The V output would go to an op-amp (-) input. A Vref would feed the (+) input. The op-amp output would feed a power amp that drives the brake. The speed would be controlled by the reference input voltage. - of course either of the above methods might benefit from some PID action to keep instabilities from destroying the planet.

Misc References: from http://www.agimotors.com/pdf/A/29020_25_40-XX_Motor_brochure.pdf Part No. Volts/Freq. Rotation Stack RPM Torque Watts 29020-29 120V/60HZ CW 7/8” 2800 8.1 oz in 76.0 Watts 29025-48 120V/60HZ CW 1.0” 2800 6.5 oz in 45.0 Watts 29025-49 240V/50HZ CW 1.0” 2600 7.4 oz in 63.0 Watts 29025-50 120V/60HZ CW 1.0” 2800 8.9 oz in 74.0 Watts 29025-52 240V/60HZ CW 1.0” 2800 10.5 oz in 49.0 Watts 29040-62 230V/60HZ CCW 1 1/2” 2760 9.4 oz in 56.5 Watts 29040-67 230V/50HZ CCW 1 1/2” 2300 15.8 oz in 94.8 Watts 29040-67 230V/60HZ CCW 1 1/2” 2760 11.9 oz in 83.7 Watts 29040-68 210V/60HZ CCW 1 1/2” 2760 11.3 oz in 91.1 Watts 29040-69 230V/60HZ CCW 1 1/2” 2700 10.1 oz in 80.0 Watts

copyright © 2005 C. Barry Ward       email: dyno(at)barryfish.com