# How to measure electric motor torque?

I need to measure torque output of a PM DC electric motor. My plan was to use a current clamp to log the current vs time, a voltmeter to measure volts vs time, and an encoder to log rpm vs time. Then I would use the equation T = (7.0429*I*V)/n to get torque vs time. Where T is torque in ft-lb, I is current in A, V is voltage in V, and n is angular velocity in RPM. This comes from the link between power, torque, and rpm, and the 7.0429 is just to get units of ft-lb. I don't have the efficiency and I can't get the stall or free load values because its installed in a machine. I do have a plot of RPM vs torque and Amps vs torque on the same plot. Can I use this to back out torque values as I log current and rpm?

• simplemotor.com/calculations – JIm Dearden Jan 31 '17 at 20:23
• To make it simple, torque is (roughly) proportional only to the current. It doesn't rely on voltage. Mechanical power is (roughly) proportional to both current and voltage. Multiplying by voltage then dividing by speed does nothing. So you just have to measure the current and calibrate the value to mechanical power you measured outside. – Janka Jan 31 '17 at 20:43
• You are measuring current, and you have a torque vs current plot. Can you not use these to estimate torque? The difficulty with estimating torque based on the input power is you don't know what your losses are. – user28910 Jan 31 '17 at 21:18
• "I need to measure torque... I do have a plot of ... Amps vs torque" - if you can measure Amps then that's all you need! (assuming the plot is accurate....) – Bruce Abbott Jan 31 '17 at 21:18
• I thought about that, but then there is a "rated power" and its listed at a certain rpm, with certain current, and certain torque. If I take that specified current, and use that value on the plot, and then find the matching torque, it doesn't match what is listed under "rated power" values. The plot says its about 3 ft-lb, where at "rated power" the torque value given was 1.91 ft-lb. – mrkevelev Feb 1 '17 at 12:58

## 2 Answers

Keep it simple. To measure the torque simply create a transversal arm attached to the motor using a known length. Let the whole motor free to turn (not the shaft, the outside case) along with the attached arm. The arm tip, opposed to where it's attached to the motor, can then be placed on a balance. You can measure the torque now using the balance measure and the arm length.

For example: if length = 0,3m and the balance indicates 100g than your net torque at the specific engine speed equals 0,1Kg x 9,8m/s^2 x 0,3m, in N.m

Be sure to calculate the arm length considering the motor center point to the arm point that touches the balance.

If you can calculate input power vs. speed, then extrapolate to estimate input power and current at stall, that would give you I squared R losses vs current. From that you can calculate a resistance value including brushes and commutator. You could then calculate I squared R losses at any speed and subtract from input power to get a closer value for mechanical power. Without no-load data to determine friction and windage losses, you probably can not get any closer than that to mechanical output power and torque.

• If it is possible to lubricate the motor, some of the 'no-load losses' can be reduced. – Whit3rd Jan 31 '17 at 21:55
• The details of the plot you have didn't sink in before I wrote my answer. I agree with the comments, the torque for any given current will not vary much when operating on other speed torque curves obtained by reducing or increasing the voltage from the voltage for which the plot was made. – Charles Cowie Jan 31 '17 at 22:08