Two new Walthers ALCO DL-109 HO-scale locos will be converted to S-CAB radio control and battery power for operation as a permanent consist with number 759 as the lead loco (the A unit) and 749 (B unit) running backwards. An identical set of components will be installed in each loco and the 2-loco consist will be operated by giving each loco the same decoder address. 

There are five components to install, three of which significantly affect space requirements. The BPS circuit board is fixed and the TSU-1000 decoder is required to safely handle power requirements of this heavy loco. This leaves battery, radio receiver and speaker for which we have options.

Here are a few considerations that led to the final arrangement illustrated below.

• For good radio reception, avoid placing the radio antenna against a metal surface.
• The BPS turn-on sensor must be outside the range of magnetic fields created by motor and speaker.
• The speaker should be located so that the motor magnet does not interfere with speaker performance.

The radio receiver is mounted at front of loco and part of the frame is cut away for two reasons:

1. To accommodate the speaker
2. To create some air-space between antenna and metal frame.

By mounting the receiver above the speaker, it is almost out of sight above cab windows.

The speaker is glued to a styrene plastic mounting that is attached to the frame with screws, which allows assembly to be removed for access to drive shaft and front truck. To complete this sub-assembly, the radio receiver will be mounted on back of speaker enclosure with adhesive or double sided sticky tape.

The BPS is installed at the rear of the loco where it sits in a plastic mounting attached to the loco frame with screws so that the rear truck is accessible. By using 1/8" styrene, the two largest BPS components (inductors) are almost flush with the lower surface of the plastic mounting. There is also a spacer below the mounting so that the BPS magnet sensor is positioned close to the loco roof.
This arrangement leaves space above the motor for battery and decoder, neither of which is sensitive to the motor's magnet.

Eliminating battery socket and direct wiring the battery connection is another space saver. Wire both battery and BPS with leads an inch or more in length, then solder and insulate to join leads as the last step in making the connection. This reduces the risk of accidental short-circuit while working with a charged battery and also provides extra wire to simplify battery replacement, should this become necessary.

There is nothing quite as frustrating as completing, testing and tidying up wiring, only to discover the loco's body shell will not fit with everything installed. To avoid this disappointment, it's best to perform several trial fits of the shell with various combinations of components temporarily held in place with electrical or Kapton tape.

As mentioned earlier, locos will be coupled back-to-back to form a more or less permanent consist. Since the locos are mechanically identical and their speed characteristics are similar in each direction, no speed curve matching is required and respective motor control CVs will have identical values. Configuring function outputs is easy, since a headlight (function F0) is the only loco feature to be controlled. One other function output, F5, will be used to turn off battery power.

Assigning the same decoder address to each loco, avoids the need for a separate consist address. Each loco's decoder, having the same address, will respond to a radio command for that address. The only reason to set up a consist address is a requirement for each loco to respond differently to the same command. This also applies to CV programming commands and is the reason loco 749's motor was connected with polarity reversed; it runs backwards in response to a forward command. A similar result could be achieved using CV29, but this requires the decoders to have different values for CV29. Except for the bell (which is set to zero volume in 749), both locos have identical values for all CVs.

With radio control, no programming track is required and, once the decoders have the same address, CVs are programmed simultaneously for both locos, using either operation or service mode programming. To set a CV value (for example, the bell volume) for only one of the locos, just turn off the other's battery power. If using service mode programming, be sure to turn off power for all other decoders to prevent them from responding to a globally addressed command.

To separate the consist and operate as separate locos, the S-CAB throttle can be used to change the address for one of the locos. If desired, CV29 can be programmed (using the S-CAB throttle) to reverse motor polarity of loco 749.

The consist is operated in the same manner as a single loco. Momentum settings should be the same for each loco with values that match acceleration and braking somewhat realistically with diesel engine exhaust sound. This can be done by using the S-CAB throttle to adjust the values of CVs 3 and 4.

Except for one operating scenario, decoder default settings operate headlights automatically. F0 toggles headlights on and off. With F0 on and loco 759 moving forward, its headlight will be on and 749's will be off. However, if the consist is backing a train into a station or siding, 749 will be shining its headlight onto the end of the nearest car. Operation will be a little more realistic if F0 is used to turn off headlights.

Battery power requires a method to turn batteries on and off. If the consist is on powered track, turn-on is automatic when BPS detects track voltage. If track is not powered, a magnet held close to the on-board sensor turns the battery on manually. With some wired connections between locos, both could be turned on using one loco's sensor. However, it's simpler to just turn on each loco separately (so long as we remember to do it).

If there is no track power, a decoder F5 command turns off batteries of both locos (since their decoders have the same address). Function commands toggle decoder outputs on and off. It may not seem intuitive, but turning F5 on, turns battery power off. However, the presence of track voltage overrides decoder F5 output and keeps battery power on. This can be the basis of a useful procedure as follows. 

Converting this loco to S-CAB radio control and battery power supply is a relatively easy project. Since there is plenty of space, I devoted extra effort to building a framework out of styrene plastic to hold electronic components and battery. A SoundTraxx TSU-1000 sound decoder, the S-CAB radio receiver and the BPS battery power supply will sit on a shelf above the motor. Two 850 mAh batteries will be mounted on edge; one each side of the rear truck assembly.

Removing 4 screws and the front coupler allows the loco body to be removed from the chassis. The rear truck-mounted coupler does not obstruct removal of the body shell. Each truck is secured by a self locking nut, which, when removed, allows the truck to drop from it's mounting. In the photo below, I've also removed drive shafts as well as some wiring for track power pick-up and headlight. The die-cast metal frame is simple and functional, providing a good foundation for constructing some plastic framework to keep the rest of the installation tidy and secure.

The entire installation is two assemblies; battery power supply and batteries at the rear of the loco chassis; decoder, radio receiver and speaker in the middle and the loco cab is left unobstructed except for a frame (to mount a marker light) that is barely visible above the cab windows. For easy access to motor and drive shafts, the plastic framework is mounted with screws accessible from underside of the loco. The retaining nut for the rear truck can be adjusted by moving the BPS circuit board out of the way.

I forgot to remove the blue masking tape when taking the photos below, but it's something I do quite frequently to keep unruly tangles of wire under control during assembly. Paint masking tape has the advantage of being sticky enough to hold wiring out of the way and does no damage when being removed. 

Since batteries don't last forever, it's best if they are accessible, as illustrated in the photo below. The BPS circuit board should be mounted so that the magnetic sensor is close to roof or side of the model where it can be activated by a magnetic wand/pencil. If the BPS board must be buried deep within the model, the sensor can be removed from the board and glued to the model's shell with wiring back to the BPS.

Below is a view from the front of the loco showing the factory installed headlight, which I've wired with a 4,700 ohm series resistor and connected the the decoder headlight output (white and blue leads). I've also added a small surface mount LED as a marker light (also with a 4,700 ohm resistor) and powered directly from the BPS 12 volt output. It's primary purpose is to indicate when battery power in turned on. Sound from the decoder will also confirm power is on, but not if sound has been muted.

Barely visible below the radio antenna is a model DS1425-8 speaker from RailMaster Hobbies. It's already assembled in an enclosure (16x27x9 mm) and produces very good sound for its size.

A neat aspect of this assembly is no components (e.g. lights) are mounted on the loco shell and therefore, there is no wiring between chassis and body of the model. All testing and initial programming of decoder CVs is done before closing up the model.

Operating Tests: Since Tsunami CV default settings are compatible with S-CAB, initial test runs are done before adjusting decoder Configuration Variables (CVs). Since I'd previously checked BPS operation, I began this running test with no track power and turned on battery power with the magnetic sensor. Using an S-CAB throttle and factory default decoder address "3", the loco moved slowly at speed step 1, direction control operated correctly, headlight control worked as expected, sound volume was good, and function output F5 turned off battery power as planned.

Battery charging: For this test, I connect a variable voltage DC power supply with current measurement to the rails and slowly increased voltage while watching current measurement. At about 5 to 6 volts, the decoder came to life and the marker light indicated track power has been detected. Current reading (typically about 200 to 400 mA) confirmed battery was charging. When DC power supply to the track was switched off, decoder and marker light continued to operate on battery power. All was good.

Decoder CVs: The basic decoder CVs can be programmed with an S-CAB throttle using broadcast mode. This requires that the "ALL" key be pressed immediately after switching the throttle to CV mode. I set CV values as follows:
CV1, decoder address, new value "6"
CV2, minimum speed, new value "1"
CV3, acceleration, new value, "20"
CV4, deceleration, new value "20"
CV5, maximum speed, no change
CV6, mid-speed, default value, "0"
CV29, decoder configuration, set to S-CAB default value "0" (The actual value stored in the decoder is "2")

Replacing the loco body: In this case, the body shell slips over the fully wired and tested chassis. Before attempting this, I used electrical insulating tape to tidy up any wiring likely to cause problems by getting stuck as the shell is maneuvered over the plastic frame holding the components. This is also the right time to make sure all wiring is properly insulated.

Once the loco body is securely in place, there is one more check to be sure the magnetic sensor is located close enough to the underside of the loco hood to reliably detect the magnetic wand. (Magnet actuation range is between 0.375 and 0.5 inches.)

Load Test: I performed this test with track power on. The idea is to prevent loco movement without applying pressure that artificially increases traction. The objective is to measure current consumption at the point where the loco loses traction and wheels begin to slip. This loco has plenty of traction; I had my hand blocking forward motion and the loco simply slide the track across my workbench. With track secured and the loco wheels slipping, current consumption was approximately 800 mA. This is right at the short-term overload capability of the BPS when operating on battery power. However, without the assist of track power, BPS current limiting would automatically lower voltage to reduce load and prevent BPS overload.

Narrow hood loco models with heavy metal castings are difficult candidates for battery power and radio control. This Life-Like Models GP9 is a classic example. It's assembly is simple; basically three sub-assemblies held together by a few screws. Remove couplers, undo 2 screws and the plastic shell lifts off in one piece. Undo 4 more screws and a large die-cast molding lifts off the chassis. This metal casting is designed to fill available space and maximize loco weight. Fortunately, it's robust and nothing is directly attached, so there is minimal risk to removing metal from top of the casting to create space flush against the roof of the engine hood. The objective is to remove a little metal as possible so as retain enough weight for reasonable traction. 

A non-sound decoder will be used for this installation, although a sound decoder could be used if more metal is removed to make space for a speaker. However, further reduction of weight is not desirable unless sound is a high priority.

This is all neat and tidy. No work is required except replacing motor and track power connections with more flexible wires that are less likely to fatigue and break as I struggle with reassembly.

Carving (machining) a flat top surface on the metal casting creates space and provides a convenient mounting for battery power and radio control components. Electrical insulating tape is used between components and bare metal as well as holding everything in place. Since the metal casting is a snug fit in the body shell, there is no room for wires between the casting and sides of the body housing. The only space for wiring is between top of components and hood roof.

This photo shows arrangement of components with wiring complete except for front and rear lights (white, yellow and blue leads). LEDs with series resistors are already mounted in the body shell. With so much metal, I was concerned about radio reception and mounted the radio receiver with antenna at front of the loco above the driver's cab. The battery is above the motor and battery power supply at the rear of the loco. The narrow hood limits battery choice to 550 mAh storage, smaller than I would prefer, but probably adequate given the motor's low power consumption and the fact that most of the owner's layout is powered. The BPS has also been trimmed by removing wire stress relief holes and eliminating the battery plug and socket by wiring directly to BPS terminals.

It has become common practice for manufacturers to use clear plastic light diffusers or light pipes for loco lights. To save space, I use a tiny surface mount LED glued directly to the diffuser or light pipe with clear epoxy cement. I also use warm white LEDs instead of the more readily available blue-white LEDs that are totally unrealistic representations of loco lights. With a 12 volt supply, I use a 4.7 K-ohm series resistor, which means the LED current is about 2 mA.

The battery turns on and stays on when track power is detected. A magnet (mounted on a pencil) and two magnetic sensors are used to manually turn battery power on and off. Since a sensor's activation range is about half an inch, (I aim for about 0.375" to make sure switching is reliable), it should be mounted as close as possible to top or side of the loco shell. The "on" sensor is already included on the BPS circuit board. I generally mount the "off" sensor on or near the radio receiver because the radio antenna is best located as far away from the motor as possible. This same precaution applies to the sensor because the motor's magnetic field will falsely activate the sensor.

Since this is a medium sized steam loco with a good-sized tender, conversion to radio control and battery power looks a relatively straightforward project. Modelers familiar with Bachmann products will be aware of phrases such as "DCC-ready", DCC On-Board" and I can now add a new one; "DCC not ready". I suspect this loco is a relatively recent release of what appears to be an old design. It's clear that DCC was not considered when producing this product.

Most of the project involved the tender, which I'll discuss later, but the engine was the real drama. Its design is conventional; a split-frame metal chassis with the rest of the model mostly plastic. Right and left-half frames are electrically separated and measurements indicate that motor terminals are connected to each half-frame. Fortunately, rail pick-up via driver wheels is insulated from the loco frame and rail power is routed to the motor through phosphor bronze contacts between wheel wiper assembly and frame halves; highlighted below.

A cardinal rule of decoder installation is isolate the motor terminals from everything except decoder motor output (orange and gray wires). There must be someone in the world who knows how to remove this loco's body and get access to its motor. Removing the bottom cover plate (in photo above) allows wheels to drop from the chassis but valve-gear is still attached to the rest of the loco. Taking the valve gear apart is often a path of no return and all other options risked serious damage to a brand new loco. Without access to the motor, the only option is leave the motor connected to loco frame.

In fact, the contacts were not removed, but simply reversed and brought through holes drilled in the bottom cover so rail pick-up is wired to the tender where track power is used by the BPS for battery charging.

Connection from decoder to motor must be through the loco frame. For this, two brass shims were added and held in place by the bottom cover.

Once obstructions are cleared from the tender, installation follows a familiar arrangement with BPS and its magnetic sensor flush against the top of the tender. By mounting the decoder and radio receiver on edge against the side of the tender, there is room for a double-size battery. The only compromise is a smaller, low profile speaker, which fits on the tender floor below the battery. It's an acceptable trade-off. A larger speaker would only provide marginal improvement in sound quality, whereas the larger battery doubles loco operating time on battery power.

Here are some close up views of wiring between engine and tender. The loco's headlight is not accessible and is factory wired in parallel with the motor. This leaves 4 wires; 2 motor connections and 2 for left/right power pickup.

The owner allowed me a few days to operate this loco after its conversion. The result is quite convincing, both with respect to performance and sound. I programmed decoder momentum settings consistent with this type of loco and limited the model's maximum speed. This makes loco exhaust more realistic and, to my ear, the small, low profile speaker provides excellent sound. As with operation of a real loco, the engineer needs to think ahead and anticipate braking distance. There is, of course, the totally unrealistic emergency stop initiated by the 'Halt" button.

Despite my earlier complaints about this model's readiness for radio control and battery power, I see real value in working with an economical product to achieve the same results as models three and four times its price.

This is a large HO scale brass loco; a 4-8-4 with a humongous tender. The objective is full S-CAB installation; both radio control and battery power. It's obvious, everything will fit into the tender, but the loco is an oil burner and the tender is all brass. Since there is no radio reception inside a closed brass box (the tender), the antenna must be mounted externally, but there's no coal load, which is my favorite antenna hiding place. Other concerns are the model's age and its solder joints, which have weakened over time, and avoiding damage to paint-work, which was/is still in pretty good condition.

Since loco owner Bill had already decided to replace loco motor and gearbox drive shaft, we agreed to split the project; he would work on engine renovation and I would install electronics and battery in the tender. We just needed to agree on wire connections between tender and engine, and this, I have described in a previous blog. What follows here is the rest of the tender installation, as illustrated in the following photograph.

All electronic components are mounted on a piece of electronic prototyping board and the whole assembly fits flush against the top of the tender. The radio receiver and BPS circuit board are out of sight behind the mounting board and the decoder is visible to the left side in this view. The tender is wide enough to accommodate 1000 mAh LiPo cells and deep enough to use a triple-pack (3 cells in parallel) giving 3 AH (amp-hours) storage capacity. Connections to the loco and speaker were temporarily insulated with green electrical tape while photographing.

I considered two possibilities for the antenna; a length wire (approx. 3") disguised as a handrail, or a very small antenna from Linx Technologies (https://www.linxtechnologies.com/resources/data-guides/ant-916-jjb-xx.pdf)

One version of Murphy's Law says "if its not tested, it won't work" and its a lot easier to test and fix problems on the workbench. Since I had not previously used the JJB antenna, I was concerned about its performance and tested its range using an S-CAB throttle. Results were excellent. However, the critical test was performance with components in the tender and for this test I did a trail installation. Results were good.

With Bill installing a new motor, I don't know its power consumption, but I'm quite confident 3 amp-hours battery storage will keep the loco running long enough to test the stamina of the most enthusiastic operators. The battery fits neatly in the available space and is easily removed.

The antenna sits flush on top of the tender, held in place by the soldered connection to the radio receiver. Since brass is non-magnetic, it does not interfere with operation of the BPS magnet sensor positioned against the tender's top where it's conveniently operated with a magnetic wand used to turn on battery power. The component mounting board is fixed in place by a couple of dabs of hot glue; just enough to prevent movement. The battery is wrapped in foam plastic to prevent it rattling around once the tender body is mounted on the chassis.

The installation is wired to charge the battery from track power, if available. As is common practice with brass locos, right rail pickup is from the loco and un-insulated tender wheels provide the left rail connection. The battery power supply turns on automatically when track power is detected. When no track power is present, a magnetic wand waved slowly across the tender top above the Santa Fe name turns on battery power. The Tsunami T-1000 decoder, function F5, is used to turn off the battery at the end of an operating session. Decoder address is "5".

Herb Kephart was an early adopter of "Stanton stuff". His experience with S-CAB evolved along with my efforts to productize what began as systems for my own use. Products got their names as user oriented documentation was prepared and Herb's terminology carries over from his prior experience with radio control; "RC" as used for model cars and planes. "RX" is a radio transmitter; "TX" is a radio receiver. Basically, S-CAB is a hand-held TX communicating with an RX in a loco. What makes S-CAB different from RC is the RX in the loco delivers commands to a standard, commercially available DCC decoder, which controls loco motor, lights and sounds.

What follows is Herb's experience in his own words and photos.

My interests in modeling for years has been trolleys and interurbans in O scale, so all the layouts that I built have been with both rails grounded, and overhead wire. Some 30 years ago I thought that it might be a good experiment to try putting an aircraft RX that I had into some sort of a switching loco, with batteries for power since collecting power from the rails would require totally impractical rework of track, especially switches, which had the rail soldered to brass ties.

With respect to radio, I had envisioned the RX feeding a servo which worked some sort of a transistor throttle. However, it quickly became evident that the size of NiCad batteries at the time made the whole package too large for any 0-4-0 or small diesel; so the project was shelved.

Some time went by, and I saw a system advertised for model railroad use, so I bought a package of one TX and 4RX's. The RX's were small enough that now it seemed possible with NiMh batteries this time. An assembly was lashed up as a test using a gondola car, with wires going to the loco. This was perfectly feasible and worked fine with one notable defect -- control was by pushing buttons for faster or slower. I found this to be incredibly clumsy (for me) -- coming up to a car to couple, so that the couplers just kissed was near impossible. The system went back in the box, where it remains.

Later, I discovered FreeRails, where one Woodie Greene was promoting RC with LiPo batteries, which are much smaller for a given amount of electrical capacity. Woodie was also using RX's from small RC autos and he proved that it was possible to hide the RX board and a small, but sufficient, battery in a reasonably compact space. This had my interest. I bought a 3 cell pack, and a RX board, and put them into an 0-4-0.

It did not work. Back and forth with Woodie who was very helpful, but no success. Finally, 3 RX boards later I figured that the 3-cell voltage was too much, and was instantly frying the boards as soon as the system was turned on. I called technical help at the company that sold the boards, and the 12 year old on the other end of the line had no idea what was the maximum safe voltage for the board. Big help. I tried cutting the 3 cell pack down to only 2 cells, managed to stick a #11 Exacto blade into one cell, which caused it to have a hissy fit; definitely NOT recommended. 

With a new 2-cell pack, I was up and running and this system worked, with one exception. I have a scratch built brass truss bridge, which I discovered was very effective as a Faraday cage—and unless I held the TX very close to the bridge, no signal got through to the loco. Also, there was an annoying buzz from the motor, caused by the low PWM frequency. I tried to get rid of this with various capacitors across the brushes of the motor, to no avail.

Just at the right time, along comes Neil with a pre-production offer of a TX (a wireless throttle) and compatible RX (a radio receiver combined with a commercially available decoder) for an attractive price. I sent money off to Neil and got back what now is the S-Cab system, although I don't think the setup had that name at the time of my purchase. I installed the gear in the 0-4-0 and encountered some low speed control problems. I don't specifically remember what the loco was or wasn't doing, but things weren't right. Several Emails went back and forth with Neil. Then, he phoned and asked me to do a number of tests with the radio/decoder setup trying to get rid of the problem which just wouldn't go away. 

After about 3/4 of an hour, he was running out of ideas. Suddenly, I remembered the capacitor that was connected across the motor brushes, and Neil's immediate response was "BINGO, remove the capacitor". The problem was my fault, and there was nothing wrong with the radio gear. After the call, I thought that anyone who would go to that much time and trouble to try to help out a customer was an alright guy in my book.

With the capacitor removed, the loco ran with no motor buzz and no loss of radio contact on the bridge. This loco continues to perform to my perfect satisfaction.

Since I already had a 2-cell LiPo in the 0-4-0 loco, that installation used only S-CAB radio control. However, Neil had been developing a loco power supply which used a single cell LiPo to produce 12 volts (which became the BPS). At the same time I was building another switcher, this time an Ingersol-Rand Oil Electric, the first diesel loco to have a control circuit that didn't force the engineer to vary both throttle and generator simultaneously when starting a load.  

I got a BPS and another S-Cab radio/decoder and installed it in this model. All eight wheels of the model are geared to the motor, a Maxon, about an inch in diameter. I never measured the amperage of this motor with all wheels slipping. I should -- but this whole setup has continued to meet all expectations. I admit that I was little dubious about runtime with all the power coming from a single cell, but I have never ran it long enough to have low voltage protection shut down the battery. The loco, being a switcher, never moves more than about six cars, although an informal drawbar test suggests it would pull at least 30 cars -- if I had that many! It goes through an hour and a half operating session on a single battery charge and with no power from the track.

Neil sent me the second iteration of the BPS, which will recharge from the rails and asked me to try it on powered track. He wanted me to wire the BPS so the loco could run on track power and only use the battery as a backup power supply. He figured if enough track is powered, the battery can be much smaller. Not having two rail power anywhere on the O scale line (and if I did, some of the cars do not have insulated wheel sets) I looked around for an application. I have been amassing a lot of On30 Bachmann, and also have a brass Plymouth 4 wheel critter, which I selected for the test.

With much cussing, I managed to get everything crammed inside, and an oval of HO track was put together. As I recall, the results surprised Neil. Even with the small battery, run time (loco alone) without rail power was 80 minutes. As I expected, this little critter operates flawlessly, and though I have no layout to run it on, it gets some exercise periodically on it's loop — I'll have to put a coupler on in someday so it can pull some Bachmann rolling stock.

I have inadvertently left the battery on and drained the battery. To remind my self not to do this, I wire the headlights to the main battery switch, Before leaving the layout, a look around tells me everything is off. I'm not keen on the battery-off push-button. So, when the critter gets pulled apart for a paint job, I'm going to substitute a small on/off switch, to break the battery circuit, which is the way I have the other locos wired. I don't use the magnetic sensor, because waving a wand over an engine that is sitting under a bunch of overhead wires (for the passenger equipment's  pantographs) sounds like asking for trouble.

All in all, I'm a VERY satisfied S-CAB user.

Introducing Peter Vanvliet
Peter models S-scale and has now converted three of his locos to S-CAB radio control with battery power. He has meticulously documented these projects on his own website and I invited him to summarize his first installation, the NW2 switcher, as a guest blog. He agreed, and what follows is his description with some Stanton editing. My apologies for skipping several of Peter's detailed explanations. Please visit his website for complete documentation.

Peter continues,
I have decided to convert this engine to use the S-CAB system, and to make it battery-powered. You can read about S-CAB in the article I wrote about it. The first step in any installation is to remove the body shell. I have a separate article on how to do that. After stripping all electronics, what remains is a motor, flywheels, truck drive shafts and wires for rail pickup. The first challenge is to determine locations for all S-CAB components and be able to replace the body shell with everything installed.

The battery:  I'm using NWSL's 1000 mAh battery and installed it first, because it's the largest component. As luck would have it, the battery fits width-wise in the shell and can lay flat against the shell roof in order to clear the mechanism that must fit below in final assembly. This requires the speaker to be removed.

I wanted to install a " battery shelf" mounted using two holes in the top of the truck mounting bridge. These were already tapped (from factory installation of the original circuit board). It was just a matter of cutting a piece of 0.040" thick styrene, drilling two holes, and mounting the "shelf" using the original screws. Since the screw heads are not recessed, I glued a second piece of styrene on top of the first one to create a flat surface for mounting the battery. The shelf sits above the truck mounting and "below" the top of the motor to make sure there is enough clearance to the top of the shell. To secure the battery, I spread some hot glue on the styrene, and quickly pressed the battery onto the glue, which sets in a few seconds. It turns out to be a solid installation, yet somewhat removable should the battery need to be replaced in the future.

The BPS circuit board: There is very little space above the motor and below the inside-top of the shell! The solution called for wrapping the motor with electrical tape and directly placing the BPS circuit board on top of  it. The issue with the BPS is the two inductors that are mounted on one side of the board. The other issue is that the way to turn on battery power is to wave a magnetic wand within 1/2" of a sensor, which means the BPS board must be close to the locomotive shell. So the BPS is positioned with the inductors (its largest components) facing down and the board pushed as far as possible toward the battery. The battery leads can then plug into the socket at opposite end of the board, with wires tucked to one side under the board. Since the motor fits snugly inside the shell, there is no space beside the motor except for a couple of layers of electrical tape.

Radio receiver and decoder: As you can see from the previous photos, we are running out of space in the engine's hood section. The cab interior and a substantial weight take up the cab's entire 3D space. So, the interior will have to go in order to make space for the decoder/receiver combo. It was actually quite difficult to get the cab off of the rest of the body shell, but it is possible. 

This is what is left after the cab interior has been removed. The weight is glued in place and is not removable without damaging the body shell. With the shell placed back on the chassis, there is a cavity where the decoder may fit! There's hope!

After studying the opening into the body, I realized that there is space above the motor mount and under the BPS circuit board. However, the heat-shrink wrap makes the combo too wide for the opening, although the circuit boards by themselves would fit. I eventually replaced the heat-shrink material with electrical tape, but first, I wanted to do a test install of the decoder/receiver and see if the system as it stands actually works.

I soldered the two gray wires from BPS board to wires connected to wheel pick-up (so that the battery can recharge from rail power, if present).  I covered the solder joints with liquid electrical tape. I then connected the red and black wires from decoder to BPS. Finally, I soldered orange and gray wires from decoder to motor. I ignored wires for lights for now.

And off to the layout I went. The track was not powered. I swiped the magnet wand over the sensor on the BPS board, turned on the S-CAB  throttle, and moved the speed control. The chassis started running smoothly, in both directions. 

Yippee! What a neat sight to see the engine running without rail power! I took a quick video and posted it on Vimeo (the chassis is running on my desk!!!) Next step, back to the work-bench.

To complete the installation, I cut a piece of 3/4" wide styrene and inserted it under the decoder and above the rear truck mounting bridge to keep everything away from spinning flywheels. After routing wires, I wrapped more electrical tape around these parts to hold everything in place.

Next, I connected the lights, snapped the cab back onto the body, put the engine on the track and tested it again paying attention to the LEDs. They looked good and they matched the engine's direction of travel.

At this point, everything was functional. However, there was one more part to install, the battery off switch.

The battery off switch: While considering possibilities, all of a sudden I had the idea of using the exhaust pipes. They are hollow and open into the body shell. My idea was to remove the push button from the orange wires of the BPS circuit, and solder them to two brass wipers. These are then going to be mounted under the two exhaust pipes. An external tool (shown in the photo) makes electrical contact to trigger BPS shut-down, just as the push button does. 

I put the engine on the track and tested the final set-up.  Everything worked great. I could turn on battery power by placing the magnet wand right behind the front exhaust pipe and operate the loco. When I am done running, I briefly insert my "exhaust pipe tool", and the battery is off! I love it! All with no external modifications to the engine.

I did find that it is easy to forget to turn off battery power, or to think it is off but it really isn't. Here's my process. Turn off track power, if present. Use the above mentioned tool (or the switch, if you use that), to turn off power from the battery. Then use the throttle and increase speed control. If the engine's headlight turns on or if the engine moves, the battery is not really off.

I tweaked some decoder CVs and got the engine to run nice and slow, while keeping its top speed reasonable for a NW2. As a stress-test, I ran the engine back and forth on my layout, which has about a 20-foot mainline. The headlight was on, but it was not pulling any cars. It was tedious, but I wanted to find out just how long a battery charge would last. I let the battery charge overnight so I was sure that it was full. I turned off all rail power and ran the engine purely on battery power, back and forth. I varied speed every so many runs, including maximum speed sometimes. The engine shut down just a few minutes short of THREE HOURS! Wow! The goal of this system is to have HO-scale engines run for one hour. It is fantastic that I got 3 hours out of it with an S-scale loco.

My next test was to see how this engine behaved on our Digitrax-powered club layout (the Houston S Gaugers). We had a one-day show. As my S-CAB article mentions, Neil indicated that the engine's drain on the battery is faster than the battery can be recharged. However, I figured that with the layout providing power and my home layout test showing three hours, I might get at least four hours operation out the engine. I started it with a 15-car train when the show
opened at 10am. At 4:30pm when we started to take our layout down, I turned off my engine. It had run the entire day! It never stopped, and it kept on running at one point when a short on the layout shut down the Digitrax system.

These two test results were way better than I had expected. I would have been happy if I could get a one-hour ride out of the engine on my home layout and maybe a two- to three-hour run on the club layout. This was fantastic. 

This loco was delivered with a non-sound, 2-function decoder installed in its tender and 6 wires between tender and engine. Our objective is to equip the loco with S-CAB radio control and battery power and upgrade to a sound decoder.

Since this model is already wired for DCC operation, no work is required on the engine. Sockets allow the engine to be unplugged and safely stored while all conversion work is performed in the tender.

The tender has a metal chassis and plastic body. Extra parts are included to model the loco as a wood, coal or oil burner. Mounting an oil bunker on top of the tender provides an excellent location for the BPS circuit board and leaves room in the tender for a larger capacity battery (more energy storage).

There's a lot of stuff to fit in a small space and the last thing I want is a tangle of wiring that prevents successful re-assembly at the end of the project. An oval hi bass speaker in a rectangular enclosure is mounted by cutting the chassis round speaker opening to fit the rectangular enclosure so it is recessed slightly into the chassis floor. This allows the battery (not shown in the figure) to be mounted above the speaker enclosure.

Not visible in the figure is a styrene bracket to hold the battery and support vertical mounting of decoder and radio receiver each side of the battery. I've departed here from the usual arrangement of mounting decoder and radio receiver back-to-back.

To fit vertically, I had to shave (sand-paper) edges of the radio receiver board to make it as narrow as possible. The styrene bracket is screwed to the chassis; decoder and radio boards are mounted using epoxy putty, some of which is visible at the far end of the radio board below the antenna. Since the chassis is metal, I wrapped circuit board edges with electrical insulating tape before gluing with the epoxy putty. The wiring panel (which I anticipated would have to withstand some tagging and pushing during wiring) is mounted, never to be removed, with epoxy cement. The lower row of pins terminate connections to motor, headlight and rail pickup. The upper row provides connections for battery and BPS.

I always approach a complicated project with the pessimistic assumption that, if I don't test every step along the way, the end result won't be successful. There's nothing quite as frustrating as devoting hours of work on a loco, only to discover at the end, it does not work. In the photo above, radio control is being tested using a DC power supply without requiring battery or BPS. The latter combo was tested separately.

I'm quite impressed with how well this loco runs and quality of sound from decoder and speaker. In hindsight, a TSU-750 would have been adequate to handle motor load and installation would have been easier with a smaller decoder. At 12 volts and heavy load (my thumb on a driver wheel), the loco required approximately 200 mA. A smaller battery would have done the job, but extra storage never hurts. As a demo to amuse and confuse: sit the tender alone (no loco) on a table; turn it on with the magnetic wand, and send commands from an S-CAB throttle. The only thing we can't do is watch the loco move. All the sounds are there, including chuff.

I recently completed this conversion, installing both S-CAB radio control and adding BPS battery power supply. In HO scale, the engine compartment inside width of 3/4" complicates the project and requires a non-standard battery.
The model (photographed here without handrails and other trim) was manufactured by Kato and is quite "friendly" from an assembly perspective.

The motor sits fairly low on the chassis, leaving useful space above the motor for battery and BPS. The dynamic braking hood is removable and I used this to make wiring more accessible. Decoder and radio receiver are mounted using a piece of circuit board attached to the motor bracket. The radio antenna is mounted vertically.

Since this arrangement requires 7 wires connecting the chassis to roof-mounted components, I added a socket to allow loco body and chassis to be separated. This facilitates testing and simplifies maintenance, should it ever be required.

Several original Kato items had to be removed to clear space above the motor; a wiring bracket (replaced with the circuit board), a metal weight, incandescent light bulbs and plastic "light-pipes" for head and rear lights.

Most of the "light-pipe" was cut off so that only the actual headlight and rear light portions were retained. The end of what was left of the light pipe was polished smooth and LEDs were glued using epoxy cement.

I also added an LED (with 2.2K resistor) in the loco cab and wired it to BPS 12 volt output to indicate battery power on or off.

Non-standard battery: Mechanically, this loco is well-engineered and has a Can motor requiring less than 500 mA full load. Because of space limitations, I used a smaller LiPo with 2 cells wired in parallel. This provides 900 mAh storage in a package measuring 2" long, 5/8" wide and 1/2" thick. Since, stiff battery leads were unworkable in the limited space, I substituted very flexible wire, eliminated the battery socket and soldered battery leads directly to the BPS.

Finally, here's a photo with the dynamic braking hood removed showing the opening I created to improve access to BPS and lighting connections.