motor free amps at voltages less than 12

[Ether]

Does anyone out there have data to share about the free-speed current of FRC motors being operated at voltages less than 12?

For example, let's say you have an unloaded CIM with 12 volts applied. Per the spec, the free current is 2.7 amps. Now, as you reduce the applied voltage, what does the plot of free amps vs applied voltage look like?

... and how might this differ from a Fisher Price 0673, which has a very high free speed and an internal cooling fan pumping air.


PS - yes, I already know that a first approximation is that the ratio of free current to applied volts is constant, equal to the ratio of free_current_at_spec_voltage to spec_voltage.

[vamfun]

Originally Posted by Ether

PS - yes, I already know that a first approximation is that the ratio of free current to applied volts is constant, equal to the ratio of free_current_at_spec_voltage to spec_voltage.

I know you want some data, sorry don't have any.

I think this approximation would be true only if the drag torque was of the viscous friction type. Generally, i_free = T_free/kt. If T_free is from bearing coulumb losses then it would not vary with speed. So when you lower the voltage and the motor speed then I would bet i_free would remain fairly constant. There are some components of friction/motor losses that change with speed but I think they are generally small relative to coulumb friction.

[Ether]

Originally Posted by vamfun

There are some components of friction/motor losses that change with speed but I think they are generally small relative to coulumb friction.

Even for the small high-speed motors with the internal fans pumping air?

I saw a web page a year or so ago (don't have the link, sorry) for the Globe motor saying that the torque losses vs speed should be modeled as linear with a non-zero (positive) y intersept equal to about half the value at spec'd speed.

[vamfun]

Originally Posted by Ether

Even for the small high-speed motors with the internal fans pumping air?

I saw a web page a year or so ago (don't have the link, sorry) for the Globe motor saying that the torque losses vs speed should be modeled as linear with a non-zero (positive) y intersept equal to about half the value at spec'd speed.

No... internal fans are a good example of increased torque loss vs speed. Guess I was think more on the lines of a CIM

[FrankJ]

Mabuchi motors web site has a widget that lets you play with different voltages & plots the curves. It looks like they hold the no load current as a constant (independent of voltage) & current line as a constant slope.

go here and play with it.

[Kevin Sevcik]

Yes, but Ether wants actual hard data for his motor calculator. He won't be satisfied until he can simulate the FRC motors so accurately that we won't need to build robots anymore. Just CAD them and play the game in the Catalyst/Ether FRC Robot Simulator. Virtual regionals are much cheaper and easier to organize after all.

[PAR_WIG1350]

Originally Posted by vamfun

I know you want some data, sorry don't have any.

I think this approximation would be true only if the drag torque was of the viscous friction type. Generally, i_free = T_free/kt. If T_free is from bearing coulumb losses then it would not vary with speed. So when you lower the voltage and the motor speed then I would bet i_free would remain fairly constant. There are some components of friction/motor losses that change with speed but I think they are generally small relative to coulumb friction.

What about armature drag? That aught to be fairly significant.

[windtakers]

Here's some rough data. (Obtained with help from Richard.)

We measured free amps for one CIM motor and one FP -9015 motor, as follows:

CIM at 12V: 3 Amperes free
CIM at 6V: 2 Amperes free

9015 at 12V: 1.1 Ampere free
9015 at 6V: 0.4 Ampere free

These tests should be repeated under better controlled conditions, with several motors, some run-in time for each motor, and better instruments. Our instruments were a 0-50 Ampere old-school Simpson ammeter and a Cen-Tech (Fluke 87 knock-off) DMM. Our power supply was a 12V DeWalt drill battery, controlled by the DeWalt drill electronic trigger circuit. Will try to post some pictures later.

[Ether]

Originally Posted by FrankJ

Mabuchi motors web site has a widget that lets you play with different voltages & plots the curves. It looks like they hold the no load current as a constant (independent of voltage) & current line as a constant slope.

go here and play with it.

Excellent reference! Thanks Frank. For anyone who wants to try it, you must first select a motor from their catalog, then click on the motor model to bring up the Javascript applet.

Attached are 12 volt and 6 volt motor curves for an RS-550. As Frank said, the calculations in Mabuchi's applet show Io as independent of voltage, and No proportional to voltage.

I doubt this is strictly true (especially for motors with internal fans), and I realize motor manufacturing tolerances and other considerations (especially but not exclusively motor temperature and wear) make this inquiry an academic exercise, but discussing it brings light to how motors work. STEM.

[Ether]

Originally Posted by windtakers

Here's some rough data. (Obtained with help from Richard.)

We measured free amps for one CIM motor and one FP -9015 motor, as follows:

CIM at 12V: 3 Amperes free
CIM at 6V: 2 Amperes free

9015 at 12V: 1.1 Ampere free
9015 at 6V: 0.4 Ampere free

These tests should be repeated under better controlled conditions, with several motors, some run-in time for each motor, and better instruments. Our instruments were a 0-50 Ampere old-school Simpson ammeter and a Cen-Tech (Fluke 87 knock-off) DMM. Our power supply was a 12V DeWalt drill battery, controlled by the DeWalt drill electronic trigger circuit. Will try to post some pictures later.

Thanks Blake! I understand the caveats you mentioned about test conditions and instrument accuracy, but assuming for the moment the data is sufficiently accurate:

• The data clearly does not follow the Machuchi Javascript applet model (no surprise there)
  
  
• The reduction in free current for the 9015 was a greater percentage than for the CIM (no surprise there either).

If you decide to take more data, an interesting test to run would be to reduce the voltage until the motor stops spinning and record that voltage and current.