DYNAMOMETER
There is much hype on the quality of new motors used in expensive locomotives, but very little data on their true performance or ratings. Over the years, there have been claims and arguments about the superiority of one type of motor over another. But to my knowledge, no one has proven any of these to be factual. Many have been mislead by so-called experts, who have expounded on second hand statements; resulting in disappointment. With the development of DCC, there has been too much stress on low current draw to meet the limitations of modules. But what is the current at desired operating torque and power in road operation?
A motor drawing .2 amp @ 12 volt with a very high, 80% efficiency would produce 1.92 watt output. This is barely enough to handle a small HO switcher with a few cars, excluding gear train losses. The final decision should rest on the marriage of the proper gear train and a well fitting motor to supply the speed and drawbar pull at the desired voltage. The current will fall where it may; hopefully within the maximum motor rating.
One of the main instigations for this development is to compare some of the expected improvements in motors with new NdFeB magnet replacements against the original and others on the market. Others may be enticed to fabricate a dynamometer to extend testing data. Although this can not be considered a high precision, calibrated device, the relationships among motors should be accurate enough for comparisons and selection. With only a small motor sample, results can only be close approximations to published data. Most do not state whether the data is typical or minimum and what the variation range is. To this end, this device will permit relative comparisons to be made under constant conditions.
Used to measure the output of a prime mover (motor or engine), an electric motor dynamometer consists of three parts: a brake to measure torque applied as a resisting force to the shaft, a tachometer to measure RPM and an ammeter to measure current. The product of the first two, with some constant, yields the output power, while the current and applied voltage yield the input power. By using a fixed regulated voltage source and a good ammeter, efficiency and other data, necessary for graphs, can be calculated. Since stall current is very high and way beyond the continuous current rating, damaging heat develops rapidly. To reduce this, a stall torque drum was added with a fixed hook and string connected to a balance scale, for rapid readings.
The first dynamometer developed was the Prony Brake of the late 1700’s. This used two blocks of wood clamped about the shaft with adjustment screws. These were affixed to a long lever arm, with its end connected to a balance scale to measure the force exerted. There were many variations over the years. But most of these arrangements require adjustments and apparatus, which are not practical for very small applications.
About 1858 Lord Kelvin developed the “rope” brake, based on the earlier design by Prony, by replacing the wooden friction blocks with a length of rope coiled around the revolving shaft. Variations of these are still used in many engineering school laboratory exercises. Designing a small, inexpensive unit, eliminating torque lever arms, presented problems.
Many years ago, Robert Higgins collected data on Sagami motors for NWSL, but the actual set-up or calibration was not found. After considering and trying many possibilities over the years, none proved totally practical. Thanks to some recent research on the Web and elsewhere, a simple inexpensive idea was developed.
Analyzing the problems
First, motors vary in size and shape, but a Micro-Mark, or similar, vise with rubber jaw pads could clamp them for easy alignment. Since this vise places the shaft center at a height of about 7″, a simple 8″ “L” bracket could position the dynamometer vertically and horizontally by clamping it under the vise base. Second, Motor shafts vary in diameter and in many cases, the actual measurement is quite different from the nominal value. Modified NWSL universal joints can accommodate them and ease the alignment. The remaining problem is to make the dynamometer.
Since most new motor specs are rated in the metric system with TORQUE in gram (force)-centimeter (gmf-cm) and power in watts, this will be used. Fortunately 1 ounce-inch = 72.00775 gmf-cm ? 72 for easy conversion. Also 1 mousepower (.001 horsepower) = .746 watt. A set of slotted, gram weights measuring up to 610 was purchased to measure running torque along with two spring scales of 250 and 500 gm to measure stall torque.
A prime item required is a well regulated 12 V DC power supply that will handle at least 4 amps. Reasonable computer switching types can be found at Jameco, All-electronics and others. Most of these will provide 5 V DC at higher currents for lower voltage evaluation.
There are three distinct types of measurements required to fully graph a motor, all of which include current measurements. The first is the noload current and RPM, which for greater accuracy and to avoid loading, should done directly on the motor shaft without the dynamometer coupled. A small drill chuck, with a contrasting stripe painted across it, can accommodate different shaft sizes. For most practical purposes, the slight increase in current and decrease in RPM due to dynamometer load are minor.
The second is a single, short time length, test to determine stall torque and current. Since the current value is several times that of the maximum operating current, time must be kept at a minimum, with success measurements spaced to permit cooling. A string, affixed to and wrapped around a 1 cm radius drum, will pull against a spring balance and read directly in gmf-cm. With the addition of the noload current and RPM, abbreviated graphs can be plotted based on theory. Theoretically the maximum power is at one half the stall torque and maximum efficiency is somewhat less; but in practice they may vary. Frequently they are beyond the maximum continuous operating current rating. Although this abbreviated method is the most common type of data presented by manufacturers, it only yields a fairly close approximation.
Performing this test at low voltages can reveal data close to the motor’s start-up torque.
In the third running type, current and RPM are measured at known torque values based on the stall torque. Plotting the current and RPM data at each torque increment will produce a graph of the actual motor characteristics for comparison.
With very small motors, the friction present in the dynamometer bearings may produce sizable errors in the noload current and RPM. This may effect the abbreviated graphing method, but complete running data, in the third full graph type, will compensate for this. These measurements should be taken on an unattached motor shaft with an optical tachometer.