0
\$\begingroup\$

I have two identical PMDC motors. I ran each of them on a dyno at the same voltage and the same load. They both consumed the same amount of current, yet one of them runs 300 RPM slower.

The one that is running slower was actually under test for several hundred hours, while the faster one is brand new. Wondering if there's something going on with the brush/com, I put each motor in a motoring dyno to check their BEMF (both motors were driven with the same speed). Not sure what I was looking for, just hoping to see a difference. The graph is below. The blue trace is from the motor that runs slower. There are obvious voltage spikes that the other motor doesn't have. Is this the reason for the motor running slow? If so, why?

enter image description here

\$\endgroup\$
  • \$\begingroup\$ To my eyes, both have spikes. The blue ones are just more pronounced. \$\endgroup\$ – JRE Feb 28 '18 at 20:14
  • 1
    \$\begingroup\$ "... one of them runs 300 RPM slower." Than what? Is that 20% or 1% of the higher speed? \$\endgroup\$ – Transistor Feb 28 '18 at 20:23
  • \$\begingroup\$ @Transistor I was thinking the same thing. \$\endgroup\$ – Trevor_G Feb 28 '18 at 20:28
  • \$\begingroup\$ One runs at 5700 RPM , while the other runs at 5400 RPM. For the BEMF test, both were driven at 5400 RPM. \$\endgroup\$ – gtetil Feb 28 '18 at 20:35
  • \$\begingroup\$ @gtetil run it again at 5700, the blue trace should be quite illuminating. \$\endgroup\$ – Trevor_G Feb 28 '18 at 20:41
2
\$\begingroup\$

The spikes are caused by the commutator. As it makes and breaks contact with the coils, the inductance of those coils causes kickback voltages.

enter image description here

That is why snubbers are important on these motors.

As brushes wear the nature of those spikes will change.

Looking at your curve though, notice that the cyan trace comes up to voltage (current) and "flattens" out within the commutation period while the blue one is still rising when it switches. That suggests the blue motor has a longer coil time constant so I would expect it to give less torque and run slower.

The reason for that could again be brush related or it could also be something in the magnetics. For instance, if the motor suffered a period of over-heating or over-current.


BTW: There is no such thing as "two identical PMDC motors".

\$\endgroup\$
  • \$\begingroup\$ Could this have any impact on the speed issue I'm seeing? \$\endgroup\$ – gtetil Feb 28 '18 at 20:20
  • \$\begingroup\$ @gtetil see edit \$\endgroup\$ – Trevor_G Feb 28 '18 at 20:25
  • \$\begingroup\$ Instead of a longer coil time constant it could be a slightly different brush orientation - this is adjustable on better motors for efficient commutation and minimizing sparking. One thing to check : is the poorer motor being driven backwards with the brushes set for forward running? That would do it... \$\endgroup\$ – Brian Drummond Feb 28 '18 at 21:00
  • \$\begingroup\$ @BrianDrummond true, that's why I termed it "brush related". That's a whole chapter all on it's own ;) Either way, it still affects the time constant, just exact cause can vary. \$\endgroup\$ – Trevor_G Feb 28 '18 at 21:01
0
\$\begingroup\$

There are indeed differences which are subtle to understand. They tell a complex story about the quality of the brush commutation, alignment, symmettry, steel remenance and magnet strength.

An impulse V=LdI/dt is created during brush disconnect and bounce. The gas corona is spun outward with speed. Insulation quality can be observed by these characteristics as well as conductor mechanics.

Without more details it would be hard to isolate the root cause of an unknown machine and measurement method. One can look at the area under the curve to guess the difference effective RMS EMF but hard to guess. This of cource can be related to brush interface quality and graphite alloy quality and thus voltage drop and magnet strength.

enter image description here Loading the output will reveal even more characteristics.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.