Test setup and results

Here is my little test area.  The “test stand” is a souvenir rail from 2 ft narrow gauge.
I installed  Spectroid because it is easy to use and does not have ads - it does a lot of math and heavy lifting, I just read a number.
The analyzer automatically picks out the strongest frequency and displays the value
I paused the analyzer to freeze the image, took a screenshot to record the data
Multiply by 60 to convert cycles per second to RPM.

Notice the bit of blue masking tape on each motor shaft.

Screenshot of the spectrum analyzer.  I circled the number it displays

The original post has an attachment with the results.

1. Old Pittman motors have a no-load speed of about 18,000 rpm
2. High speed passenger steamers need a motor with the full 18K to 24K rpm
3. Tiny new can motors with RPM of 6k to 8k are excellent for switchers – much lower top-end
4. Heavy haulers can use an 18k rpm motor and drag at lower RPM, higher amps

Back EMF, unexpected voltage jumps, and compensating

The data is attached to the first post.  Each voltage and RPM has two values,

• with Back Electromagnetic Force 
• and adjusted for BEMF.

Originally, I simply set the transformer to 12V and measured the RPM.  I compared it to the rated RPM and it was way off.  A motor rated at 10,000 RPM 12v was running 14,760?  Then I noticed the voltage jumped from 12V to 16.6V.  Oh.  BEMF effects.  I left the leads attached and reduced the transformer until the meter read approx. 12V. Et voila! 10,380 RPM, close enough to 10K.

I re-tested all the motors at 2 applied voltages  

• Voltage before attaching to motor (12V or as close as I could make it)
• Voltage/RPM after attaching to motor, measured V jumped or dropped, according to motor BEMF
• Voltage/RPM after raising or lowering voltage to about 12v:  

The BEMF effect was dramatic with the new can motors.  All raised the effective voltage.  The open-frame Pittman motors dropped the effective voltage by about 10% and I had to increase the output of the transformer.

The current was measured at free running and stall, without adjusting for back emf:

• 12v measured before applying leads to motor,
•  do not compensate for bemf, allow motor to run faster than specification, measure amps
• do not compensate for bemf, clamp the motor shaft to zero rpm with pliers, measure amps

As expected, the can motors used much less current than the open frame.  I was surprised at the difference. 

• Free running : Stall
• Open frame 0.347 : 1.1A
• Highest current usage from can motor 0.049 : 0.66A

Those poor little things got rather warm at stall.  I clamped them very briefly to prevent the magic smoke from escaping.

BEMF affect

i doubt you would measure any difference in voltage if you used an electronically regulated DC supply set to 12V before you attached it to the motor and measured the voltage while the motor was running.    This "BEMF affect" is because your using an unregulated half-wave rectified DC.

but I think the same effect would occur with a PWM output.   The measured output voltage of a PWM output set to 50% duty cycle would be 50% of the voltage supply.   But it would be higher when measured across a running motor.

when measuring the RPM and corresponding voltage of a motor, the voltage needs to be measured while the motor is at that speed and load.

it looks like your notes indicate the stall current is less for the Mabuchi motor, 0.4 stalled vs 0.8 running.   Is this correct?

backwards numbers

I probably reversed the stall and running current for the Mabuchi motor. i will re-run it, and revise.

Thanks for the info on the PWM and peak output.  That makes sense. My ATX power supply blew - I did not have a source of clean 12v.  Probably get another from Freecycle.

I have not devised a method to mechanically load the motors. I should choose a loco as a test bed for actually loading and running characteristics.

Use a motor / generator as a load

I used Hysteresis Brakes on my motor test bench, but I have at time resorted to using generators (motors) as loads. Just wire the "generator" output to some variable power resistors. With some "generators" cogging can be an issue at low RPMs, so skew wound motors with at least seven armature windings would be recommended.