Motor test bed #5- Seaboard SD45 8-28-16

This engine has been tested extensively to evaluate the impact of various motors & electronics modules.

This is one of a series of tests being performed on various motors with and without electronic modules. The purpose in this series is to understand the performance differences with various motors. It is also looking at the impact when an electronic module is in the circuit. These electronics can take many forms. The most obvious is a DCC decoder. It can be as simple as a light card. It can be as complicated as a sound system added to a decoder or something else. It may be one of the new DC control modules.

This is by no means a complete examination of the motors or the electronics. It is intended to understand how the addition of the electronics in the overall circuit impacts the engine performance.

The standard test activity is to be performed on each variation and the results compared and discussed. The intention is to examine this over a series of engine, motor and electronic variations. It is important to verify and define the usual result and the anomaly. This will require several tests. Each of these reports will focus on a test series on one engine. The overall results will be examined as testing proceeds. The various motor options are examined first. Then an electronic module or modules will be added to the end of each of these tests.

For this specific test series, the engine configuration is an Athearn Blue Box Seaboard SD45 engine with the following characteristics:
1. Athearn BB Drive Parts
2. Athearn BB SD45 C-C trucks
3. Stock Athearn BB Wheels
4. New Axle Gears
5. No weight added
6. Motor & Trucks Cleaned & lightly lubed
7. Wheels polished with Kadee wire brush
8. 5-Wire connections

This engine is a stock configuration showing the impact of adding a can motor compared to the stock Athearn blue box rectangular motor.

In this test series there were five variations tested. This consisted of examining two motors and three DCC decoders. These configurations are:

1. Athearn BB Rectangular motor #2, No electronic module
2. Mashima 1833 Can Motor. No electronic module
3. Mashima 1833 Can motor, Lenz DCC decoder (number unknown)
4. Mashima 1833 Can motor, NCE D13SRJ DCC decoder
5. Mashima 1833 Can motor, NCE DA-SR Card DCC decoder

This engine is shown in the following figures:
sd451
sd452

This discussion will focus on the performance measurements of the unit. The basic test requirements are as follows:

1.) Powered with DC voltage with no external pulse wave modulation

2.) Running on a level surface (that is measured and adjusted as required weekly).

3.) The same 8-foot segment of track is used for all the basic tests.

4.) the track is cleaned before each test.

5.) the unit is run without external load for all but the max draw bar force tests.

6.) Each track running data point is the average of three measurements.

These discussions do not deal with accuracy of the shell, the location and nature of the details or the lettering. In fact, because it is a test bed, several details have been left off for expediency.

To appreciate the results, the question of relative to what always comes to mind. For this reason, the data is compared with the results from the full set of nearly 300 engines tested in the data base.

Basic function results are presented in the following charts. The data in each category is compared to the total set of engines.

1- Scale velocity vs voltage of the engine running alone on straight and level track. There is no external load.
sd453
This data is shown from minimum to 16 volts. The actual minimum values are included here because they show how the DCC modules are impacting the shape and position of the curves. The color code is maintained throughout these charts. The grey lines are from the data base of engines tested in this manner. These include all the engines in the data base with no attempt to segregate by some feature or era.

This chart speaks volumes to the impact of the Motors & electronic modules compare.

In this chart a reference line is shown (in black dashed) that is intended to represent the goal of best speed voltage function. Basically, the desire is for the unit to crawl as slow as possible at low voltage and be able to replicate the full size unit at high-speed. For this purpose, 80 smph at 12 volts was chosen as the value. One could argue that it should be higher. True, but without any external load, why would it be lower? Notice a significant number of engines from the data base run slower than this goal level.

In the DC configuration, the Athearn blue box stock motor runs above the goal level at all voltages, except the very lowest speed. The Mashima 1833 can motor in the DC only mode runs very near the objective level, falling short at 6 volts and below.

When the DCC decoders are added to the circuit, the velocity voltage curves markedly change. On the high voltage side, the speed at voltage is significantly lower. Particularly for the Lenz decoder option. On the low voltage side the shape really changes. Visually, the engine is pulsing in this region. With all the decoders the pulsing is noticeable from the initial movement until 9 volts. The Lenz and D13SRJ are both used and the DA-SR was new out of the package. This is not meant to be a condemnation of any type decoder. The same has been seen on ESU, TCS and some others. As will be the outcome of this activity, these changes seem to be part of a DCC module impact. Apparently independent of the manufacturer.

Focusing on the high voltage side of the chart, a vertical line has been drawn at the 12 volt power supply level. This is approximately the DC portion of the DCC signal. Additionally, two horizontal lines are shown intersecting the DA-SR/ D13SRJ and the Lenz data at this voltage intersection. Another vertical line is drawn where this horizontal line intersects the DC version. The voltage difference between these vertical lines is the implied voltage “loss” of each decoder module. Here for DA-SR and the D13SRJ the loss looks to be 1.1 volts. The loss for the Lenz decoder is slightly over 2 volts. The power represented by the DC version curve above the horizontal line has been lost to the system. Any loss is a penalty.

As you come down in voltage, the module versions seem to parallel the DC versions down to 6 volts on the supply for the two NCE decoders. The Lenz decoder would no longer sustain movement below 7.6 volts. Below 9 volts it is obvious that the module is pulsing the motor. In all decoder tests this is visually apparent. The two NCE module units can not sustain movement below a significant power supply voltage, 5.2 volts.

The shape of the curves drawn on the chart were verified taking individual points at 0.1 volt increments. It is clear that the DCC module is creating a PWM signal to the motor even though the supply is not producing one. The resulting velocity voltage curve is very steep with the independent parameter of voltage. From a classic stability consideration, this is very bad.

As shown in the module data, the initial sustained velocity is 5.4 SMPH. The supply only has voltage increments of 0.1 volts. The actual minimum probably occurs at an increment smaller then the minimum of the supply. The problem with this is with any voltage drop due to conditions, this module version will stop. What this says to me is that the PWM while able to sustain lower velocities, they actually can not achieve those levels in the practical sense. With some margin for actual conditions, the DC version will achieve lower minimum speeds.

2- current draw vs voltage for the engine only operating on straight and level track.
sd454
As shown in the chart, both these versions are competitive in current draw, (low). They have some interesting variations with voltage.

Keep in mind that the input voltage times this current is the input power to the system. If the efficiency of the versions were the same, this would indicate how the output power would look. Fortunately, they do not have the same operating efficiency. That was indicated in the velocity chart and will show in other parameters as well.

Except for the Lenz decoder, the curves tend to follow the power loss indicated by the reduction in the measured velocity/power. The Lenz curve would imply that it runs faster than the other two decoders. It clearly did not, so it is losing more power for some reason.

3- Starting velocity. This is the minimum velocity that will sustain movement. This occurs at a discreet input voltage. This voltage varies from motor to motor and drive to drive. It seems to be a function of engine weight and motor capacity. For these charts it is shown as a function of weight.

The back ground data on this chart has been segregated by era, pre 2000 and post 2000 approximated release date.
sd455
As has been indicated in the earlier discussion, the steepness of the velocity slope with voltage for the DCC module versions may be biasing this result.

The data here shows the no module versions sustains movement between 5.1 and 7.6 SMPH. Here the Stock BB motor has the advantage. Both have a similar slope and still are running under 10 SMPH with some voltage margin.

In the case of the options with a DCC decoder on board, the NCE decoders initiate movement at 4.6 and 5.3 SMPH. The Lenz as indicated earlier measured an initial velocity of 28.7 SMPH. If the Power supply could be set at a finer voltage increment the decoder cases may have been able to sustain velocity closer to zero. However, the sensitivity of a 0.1 increment in voltage should be a concern for performance stability in this region of the operating range.

These results are likely dominated by the drive and trucks as well as the DCC module. While the parts were cleaned and lubricated, they are pre dog bone designs. This is being examined in other engine test beds to see how it varies.

4- The voltage that is required for the sustained velocity is shown in the following chart:
sd456
Here the DCC Module versions are clearly on the high side of the history. What is being measured is the supply voltage. The actual integrated motor voltage being delivered by the module is likely quite a bit lower. The problem is if you have an engine with a module and another engine without a module, there is a potential conflict when one starts at a power supply voltage of 2.8 volts and another needs 5.2-7.6 volts to initiate activity.

5- Starting velocity variation, implied torque wobble

In every case the data is repeated three times. Running over the same distance. The velocity and current levels are measured digitally. Differences in these readings are an implication of the potential torque wobble of the motor. This also can be a measure of the pulsing level of a PWM impact on the motor.
sd458

Interestingly, The best configuration in this parameter is the Mashima with no module. This same motor is worse with each of the decoders in the circuit. Here the largest is the D13SRJ with a variation of 8.8 SMPH The Lenz and stock athearn BB motor with no module both demonstrate a respectable 2.0 SMPH. The smoothest in this case is the no module can motor..

6- starting current draw

sd457
The initial current draw for the two motors of this engine is interesting. The no module configuration seem to fall where history would indicate they should. The two NCE decoder options compete very favorably with the other engines tested to date. The D13SRJ being near the lowest measured initial current draw. The Lenz option however is above the stock BB motor and definitely high for a can motor expectation.

7- The maximum pull force of the engine is shown in the following figure. These data are taken at 12 volts, when the engine will no longer pull the weight off the floor. Thus it is zero velocity. This may actually be just above the maximum pull force.
sd459
Here all of the configurations are very competitive pullers. In this case, the stock BB motor pulls better than the Mashima can with three of the four options The fourth, with the DA-SR decoder, is the best puller of this series. This is interesting since the motor voltage appears to be lower on these decoder versions for a given supply voltage. Again the results are an average of three measurements and the individual variations are small.

8- Maximum Pull force weight function is shown in the following chart. Here four weights increments have tested for each motor configuration. This is a relatively new test in the series, so the data base does not include as much history. The weight increments are roughly 150 grams. Totaling just over a pound of added weight. The base engine configuration is the left end of the curve.
sd4510
In this case the DA-SR version which showed to be the best puller at the base weight, has a much shallower slop. with weight than the no module versions. At the high weight, the no module versions are clearly the best pullers. All three decoder versions tend to have a similar slope with weight.

9- Current draw at max pull force.
Data does not show anything new, the relationship between configurations is maintained with the added weight. In this case, the Mashima engines have lower current draw. Not the lowest measured, but much better than the Athearn BB motor version.
sd4511

10- Based on the work of others, the maximum pull force can be translated into the number of 4 ounce cars that can be pulled up a 2.5 percent grade. This assumes that the entire train is seeing an integral grade of 2.5 percent. This is a fairly sever assumption.
sd4512
In this case the force curve is mimicked by the translation constant. The beauty of this is that one can see how many cars are implied by the differences in pull force.

Keep in mind, all of the versions can pull well over 12 cars up the grade. The real engines would be expected to pull 2 times the number of drive axles, or 2 X 6 = 12 cars.

11- Taking all of these results into account through the second performance criteria that I have defined in another post, these configurations are compared in the following figure. As a gage, a PC2 value of 10 or more is considered acceptable. Greater than 50 would be very good to excellent.

As has been discussed, the initial velocity level on the DCC module version is questionable because of the sensitivity to the voltage increments. For this reason, the parameter may be overly harsh on that option.
sd4513
With that in mind, the chart shows the no module versions to be the best over the entire weight range. In this case the stock Athearn BB motor version is the best. Beating the Misima motor noticeable. The configurations with the decoder in place all fall down from there. The NCE versions are still acceptable while the Lenz version is not. Keep in mind these all have the same Mashima motor.

So this testing has shown the impact of adding a can motor and various decoder modules to a stock Athearn BB SD45 engine. The results clearly show that there are some penalties that are paid for the control benefits. The most worrisome is the low voltage stability issue. Is this one of the reasons that DCC systems seem to be extremely sensitive to dirty track? Yes, the wheels can and do get dirty faster with the higher voltage levels, but the extreme sensitivity to voltage will drive this problem as well.

Leave a Reply