Since all the motors tested have been used in powering my pesonal locomotives and work done for others, the collection of data was done without bias under the same standard conditions. They represent a typical small random sample, not an average. However this does show the general relationships among the various parameters. As more data is collected, a more thorough analysis will be possible.

Some test results, in stall torque and current, reveal surprising results in comparing magnet upgrades and some common flat can motors. The order presented is roughly by similar size groups. The RAT is the stall torque increase ratio of the NdFeB over the original alnico. The dynamometer, stall torque drum and spring balance scale were used.

The noload values were measured with a bare motor. A digital laser tachometer was used to measure RPM. To evaluate parameters, motor graphs, of the quality supplied by most manufacturers, can be plotted from this data.

The first tests were at the standard 12V, while for those interested in lower voltages, the second was run at 5V to satisfy those, who doubted that the same relationships would be generally valid at reduced voltage. Since the startup voltage of a loco is dependent on total friction from motor, gear train and load, lower voltage testing is moot. Other than for bragging purposes, the actual value is not important; unless it is excessively high enough to reduce control range. Very rarely would this be due to the motor alone.

NOTE: Motors should be well lubricated with a light oil such as Labelle 101. New or old, they should be run-in until the noload current drops to a stable minimum or the RPM hits a maximum. This may take several hours on some. Lack of break-in becomes very evident during low voltage tests.

Maximum power in watts was calculated from 1/2 the stall torque and noload RPM as developed inMOTOR GRAPHS .

P = K * T * RPM / 4

Since the RPM in the table is stated as X 1000, the formula was rationalized to:
P = T * RPM * .00256725

NOTE: Heat develops very rapidly due to excessive stall current thus increasing copper resistance, which reduces both current and torque. Sufficient cooling time is required between repeated tests. Of note, open frame types cooled far more rapidly than cans.
TEST RESULTS AT 12 VOLTS

MOTOR

AMP

gmf-cm

oz-in

RAT

AMP

RPM

WATT

MP

Sagami 1015 square

.30

18

.25

.03

15.9

.735

.985

Mantua PM-1

1.4

44

.61

.34

18.4

2.07

3.85

Mantua PM-1 w/ NdFeB

1.4

76

1.06

1.73

.24

15.3

2.97

3.98

Pittman DC-60

1.8

74

1.03

.22

14.8

2.80

3.75

Pittman DC-60 w/ NdFeB

2.0

128

1.78

1.73

.22

11.8

3.87

5.19

Jameco #231731 (1525)
Nichibo PC-130SF-09480

0.42

61

.85

.05

5.9

.90

1.31

Pittman DC-70

1.8

120

1.67

.17

14.4

4.42

5.93

Mantua MG-81

1.0

86

1.194

.10

13.4

2.961

3.97

KTM (DC-70 size)

1.8

110

1.53

.30

19.2

5.41

7.25

Pittman DC-71

1.8

150

2.08

.18

13.6

5.22

7.00

New Bowser DC-71

1.8

135

1.88

0.9

.16

14.6

5.05

6.77

Pittman DC-71 w/ NdFeB

2.0

250

3.47

1.67

.16

9.8

6.27

8.41

Cannon 1833

0.9

122

1.69

.06

8.64

2.69

3.61

Sagami 1833

0.9

78

1.08

.06

8.4

1.68

2.25

Sagami 1836

0.8

96

1.33

.04

7.3

1.79

2.40

Sagami 2240

0.9

158

2.19

.075

8.0

3.25

4.35

TEST RESULTS AT 5 VOLTS

MOTOR

AMP

gmf-cm

oz-in

RAT

AMP

RPM

WATT

MP

Sagami 1015 square

.13

7

.097

.03

6.9

.124

.166

Mantua PM-1

.48

12

.167

.18

5.0

.154

.206

Mantua PM-1 w/ NdFeB

.50

33

.194

2.75

.14

4.0

.339

.454

Pittman DC-60

.90

34

.25

.18

8.3

.724

.971

Pittman DC-60 w/ NdFeB

.8

59

.528

1.74

.21

5.8

.879

1.178

Jameco #231731 (1525)
Nichibo PC-130SF-09480

.12

22

.31

.04

2.3

.130

.174

Pittman DC-70

.9

65

.903

.12

6.6

1.10

1.475

Mantua MG-81

.44

40

.555

.04

4.6

.473

.633

KTM (DC-70 size)

1.1

58

.806

.40

6.7

.998

1.338

Pittman DC-71

.50

74

.861

.14

6.0

1.140

1.528

New Bowser DC-71

.60

58

.806

.935

.12

6.6

.983

1.318

Pittman DC-71 w/ NdFeB

.9

107

1.49

1.445

.12

5.0

1.373

1.840

Cannon 1833

0.42

51

.709

.03

3.9

.511

.684

Sagami 1833

.46

44

.611

.05

3.2

.378

.507

Sagami 1836

.32

64

.889

.04

3.0

.493

.661

Sagami 2240

.50

60

.804

.045

3.3

.508

.681

OBSERVATIONS

Often considered poor motors by the "experts", an old Japanese KTM (DC-70 size) with 19.2 K NLRPM yielded very respectable results except for high current..

Possibly due to a smaller cross sectional area magnet and winding changes, the new Bowser produced lower torque and power than the original Pittman DC-71.

It appears that in a given size range, the open frame motors with NdFeB magnets excel substantially in stall torque and maximum power. The standard open frames appear to surpass can motors in the same parameters. If they track similarly to normal graphing, they should produce more torque and power in all the running areas. It must be pointed out that most motors can not be run continuously at maximum power. Unfortunately with most hobby usable motors, maximum continuous current can not be found. Only future graphs of running tests with varying torque loads will verify these parameters.

Comparing the two test groups, it is apparent that some motors do not have similar ratios between the two voltage levels. This may be due to different winding characteristics or physical geometry. Smaller motors are more susceptible to bearing and brush friction at lower torque levels. It was observed that these motors stalled before reaching maximum stall torque in the first trials. Rotating the armature position with respect to the drum finally yielded the max. This could be due to the lower acceleration at smaller torque values, with larger armature masses, which may produce cogging. This presents a good argument for higher gear ratios, which permit higher applied voltage, motor RPM and acceleration. Also higher gear ratios reduce the required motor torque with any given drawbar pull. The unused 2240 had a very surprising drop in torque and power at 5V.

As shown in GEAR FUNCTIONS a Tyco pacific with a PM-1 motor, 80" drivers, 26:1 gear ratio and 3.5 oz pull requires 4.45 gmf-cm motor torque. The motor supplies sufficient torque at 5V along with any shown. Since no data have been collected on running drawbar pull, resulting from rolling resistance at various speeds, grades and curves; power estimates are very nebulous. The only obvious conclusion is that more is better, up to a point. Insufficient power will slow a train more rapidly with increasing load. The worst case would stall the locomotive.

Since the clearance between drivers is about 14.5 mm = .57 ", the flat cans will not fit between them on steam locos, thus requiring a gear box. While the open frames were designed to fit. The 18 mm flat cans are actually 23.69 mm in diameter, requiring more height clearance. The diameter of the 2240 cylindrical is difficult to fit into most diesels.

Lower current appears to be the only advantage for cans at the expense of power and torque. As more data are collected, you may judge the merits of the various motors and types more thoroughly.

BACK TO MOTOR GRAPHS

BACK TO MOTOR EVALUATION

BACK TO REMOTORING

BACK TO REMOTORING SKEWED

BACK TO REPOWERING KNOW WHY

BACK TO SUPER MOTOR

HOME

## Authored on the AMITHLON AMIGA

using:

### OS 3.9 -- 2002

 Browser -- Ibrowse V2.3 -- 2003 Text -- CygnusEd Professional V4.2 -- 1999 Drawings -- XCAD-3000 V1.1 -- 1992 Graphs -- Math-Amation V1.0d -- 1988 Rendering -- Image FX V4.1 -- 2000 Digital Camera -- Kodak DC25 -- 1998 Digital Camera 2 -- Kodak DC280 -- 2003 Scanner -- HP Scanjet 6200C -- 2000 HTML and mistakes -- BUDB -- 1931

Hosting by WebRing.