This post describes my most difficult dead-rail conversion to date: an MTH O scale 2-8-8-8-2 Virginian Triplex (MTH product number 20-3101-1) that I purchased on eBay circa September 2019. Previously, I converted a Sunset 3rd Rail Allegheny with an MTH Proto-Sound 3.0 board to dead-rail, but the Triplex was my first complete dead-rail conversion of an MTH locomotive to 2-rail operation, which included lathe turning high-profile wheels to approximate an NMRA RP-25 flange profile (also see NMRA standard S-4.2) so that the locomotive would operate reliably on track meeting NMRA standard S-3.2.
The inside view of the tender below demonstrates a significant challenge: space is very tight with the Proto-Sound board and the large speaker consuming a large part of the tender’s internal volume where we need to install additional dead-rail components: DCC-compatible RF receiver; 14.8V LiPo battery; and switch, charging, and antenna wiring.
The strategy starts to emerge:
Replace the Proto-Sound 2.0 (PS2.0) board with a PS3.0 board that can operate in DCC mode.
Remove the original rechargeable battery and its cradle and locate the 14.8V LiPo battery pack there.
Remove the large speaker and replace it with a smaller 4-ohm speaker so that we can make room for the 14.8V LiPo battery pack and the Airwire CONVRTR-60X DCC-compatible RF receiver that operates in the 902-928 MHz ISM band on Airwire channels 0-16.
Lathe down the high-rail wheel flanges to approximate an NMRA “RP-25” profile for 2-rail, dead-rail operation.
An advantage of this strategy is retaining almost all of the control the PS2.0/PS3.0 provides, including directional head/tail lamp, marker lights, cabin lights, flickering firebox, sound, and fan-driven smoke units.
Proto-Sound 3.0 Conversion
The first step of the dead-rail conversion was easy: replacing the Proto-Sound 2.0 board with a Proto-Sound 3.0 (PS3.0) board from Ray’s Electric Trainworks. As I have mentioned in other posts, working with Ray Manley is a great pleasure. I sent my PS2.0 board as a trade-in to Ray, and he took care of the rest, providing me with a fully-functional PS3.0 board, complete with DCC capability.
The heatsink for the PS3.0 board necessitated drilling and tapping a new mount hole with spacer, as shown in the figure below.
The following photos show the original electrical power inputs to the PS2.0 board and their modified connections for the replacement PS3.0 board.
As you can see below, the AC power from the center rail pick-up (hot) and the outside rails (ground) were disconnected – we will be getting our power from a 14.8V LiPo battery pack in the tender. In this case, there is no Constant Voltage Unit, so no Battery +(14.8V)/Battery -(Ground) connections are required.
The AC power connections in the tender are also disconnected, and the power inputs to the PS3.0 connect to the switched battery power. The Battery +(14.8V)/Battery -(Ground) connection on the wiring harness was NOT required.
Locomotive Electrical Modifications
There were two aspects to the electrical modifications in the locomotive:
Electrical power supply
The original headlamp was a power-hungry incandescent bulb. An LED with a polarity-independent plug from Evan Designs was used to eliminate the need to determine the polarity of the original headlamp wiring.
The power-related modifications consisted of removing the center-rail pick-ups, which is very easy on MTH locomotives and disconnecting any wiring to the center-rail pick-up (hot) and the outside rails (ground).
Battery installation was very challenging since the only practical placement location was the original rechargeable battery and its cradle mounting beside the PS2.0 board. A special-order 2x2x1 14.8V, 2600mAh (38.48 Wh, 5A rate, LxWxT: 133 mm x 40 mm x 25 mm) LiPo battery purchased from Tenergy.com provides the one cell-diameter thickness required to fit the battery pack between the PS3.0 board and the tender hull.
In MTH steam locomotives, the wheel axles insert into a solid cast chassis frame, so the driver wheels must be pulled off the axle before machining the high profile wheels to approximate an RP-25 profile that is compatible with two-rail, dead-rail operation.
Tender Mechanical Modifications
The tender’s mechanical modifications involve adding a Kadee 740 coupler and accommodating additional dead-rail electronics.
The original coupler pivot, rather than using a frame-fixed mounting, was used to mount a Kadee 740 coupler. This strategy ensured that tight curves would not bind the coupler.
A Kadee 740 coupler was mounted on the original coupler pivot, as shown in the Figure below. The brass screw heads were ground down to provide clearance with the tender frame.
Additional Dead-Rail Electronics
The added dead-rail electronics include the charging plug, the ON/OFF/Charging plug, a smaller speaker, and the antenna mount.
The original speaker was far too large to provide clearance for the additional battery, DCC-compatible RF receiver, and other electrical components needed for the dead-rail installation. So a 16mm x 35mm speaker was placed in the bottom of the original speaker’s cavity, and UV glue holds the speaker in place.
With the locomotive reassembled, it’s time to test it out! If your locomotive has a smoke unit(s), always ensure sufficient smoke fluid is loaded. Even if you don’t intentionally turn on the smoke unit – sometimes it’s unexpectedly activated.
After doing about eight or so O scale 3-rail to dead-rail conversions for steam locomotives, some similar features pop out that I will discuss in this blog. As with my other O Scale Dead Rail blogs, I will try to stick mostly to my own experience.
A note of caution: O scale steam locomotives are expensive, and some, to me, are works of art. Consider very carefully whether you have the patience and skill required to make locomotive conversions. I got into O scale dead-rail conversions to teach myself patience, a few skills, and respect for these beautiful models. Give yourself plenty of time to make these kinds of conversions – being in a hurry is a prescription for trouble – I know because I sometimes got in a hurry. Just don’t.
There are some good tutorials on locomotive repair and disassembly. I recommend this one as a good place to start.
Good grief, what do “general considerations” mean? Well, it’s the general aspects that guide the conversion process for both the locomotive and the tender.
First, most, but not all, of my O scale 3-rail to dead-rail conversion experience is with Sunset 3rd rail steam locomotives: Big Boys, Cab Forwards, Challengers, and Alleghenys, so first off you see that I’m an articulated locomotive fan. Articulated locomotives (AL) can be challenging to dis-assemble, and, especially, re-assemble. Heck, they are tricky to handle correctly especially since the front driver chassis must in some way “float” to navigate modest-radius curves.
In general, even though 3-rail locomotives pick up AC from the track, the locomotive motors are almost always DC motors, where a connector from the tender supplies the DC power via a connector from the tender. Usually (always in my experience), the two outside rails are electrically connected to the locomotive chassis, and the AC power from the center roller pickups completes the power loop, fully isolated from the locomotive chassis.
This outside/center AC power inputs are typically supplied to the tender from the locomotive via the same connector to provides DC power from the tender back to the locomotive motor. So right off the bat, you have some bad news and some good news:
Bad news: We will need to eliminate the physical and electrical connections to the center-rail AC picked up by the center pickup shoes from the locomotive to the tender because we’re not using rail power in dead-rail. This will be accomplished by removing the center rail pickup shoes and, just to be on the safe side, eliminating all of the center-rail AC wiring. In theory, once the pickup shoes are removed, all center-rail AC wiring is isolated electrically, but I don’t take any chances, and I just remove all of the center-rail AC wiring.
The good news about the bad news: We can re-purpose the center-rail AC power plug connection between the locomotive and tender (originally sending center-rail AC power from the locomotive to the tender), and instead we send Switched Battery+ from the tender to the locomotive to provide power to the Constant Voltage Unit that distributes power to components such as smoke units, marker lights, and sometimes the headlamp.
Good news: The locomotive’s chassis “ground,” which is electrically connected to the outside rails via the locomotive wheels, does not require electrical connection modification – all we’re going to do is ensure that chassis ground is connected to the tender’s Battery- (Ground) through the original locomotive-to-tender plug. The Battery- (Ground), as the name implies, is also connected to the tender chassis ground.
Good news: It’s easy to get the DC power from the tender to the locomotive’s motor(s) through the original locomotive-to-tender plug without modifications to the locomotive motor’s electrical connections.
Another unique aspect to 3-rail locomotives is the reversing board in the tender that converts the AC power coming from the locomotive to a correctly polarized DC voltage for forward/reverse motion depending on the pattern of interruption of AC to the locomotive. I am not an expert on reversing units – I am in the business of removing them for dead-rail operation. If the locomotive has sound, the sound electronics/card is usually highly integrated (meaning not DCC-compliant) with the reversing unit. So, we must provide a DCC-compliant replacement sound card of full DCC decoder with sound since the radio-controlled receiver boards “speak” DCC in most cases. Removal of the original sound electronics is a shame because it may have some unique/interesting audio we’d like to reuse. If I ever figure out how to reuse the sound from these original sound units, that will, of course, be another blog. My initial attempts to obtain legacy (remember these cards are sometimes over fifteen years old) circuit interface information from OEM sound card manufacturers such as QSI Industries have not been fruitful, even though they were friendly.
All of the 3-rail O scale steam locomotives I have converted to dead-rail had the following aspects of conversion that I discuss in the next three sections: center pickup removal, lighting modifications, and various electrical modifications. Three-rail locomotives are somewhat more challenging than 2-rail to convert because of the unique aspects of power pickup using the third, center rail.
Center Pickup Removal
Removal of the center pickups involves both the rollers themselves and the electrical connections to them. I remove the locomotive pickup rollers because we will usually operate the locomotive on two rails, even though they are unpowered. I suppose you could leave the rollers in place, but this would restrict dead-rail operation to dead 3-rail track. In summary, I think removing the rollers gives you more operating flexibility for dead-rail operation.
See the Figure below for an example of how to remove the undercarriage baseplate so that you can remove the rollers.
Unfortunately, the undercarriage baseplate requires removal to get at the center pickup roller mounting screw. This plate is usually held in place by six to eight retention screws, but even when you remove them, you have several issues:
Often you cannot remove the plate without catching the brake shoes on the drivers. You can solve this problem by removing the brake shoes from one side and sliding the base plate sideways to free to brake shoes on the other side. The wiring underneath the plate connecting to the pickup rollers does not have sufficient slack to allow lifting the cover. You have two options: carefully slip a wire cutting tool in the narrow space between the freed baseplate and the rest of the chassis, or remove the entire wheel chassis from the boiler section and cut the wires from above. See the Figure below. You still need to remove the baseplate to unscrew the center pickup rollers because you cannot get to them even with the wheel chassis separated from the boiler section.
Once you free the baseplate from the rest of the chassis, then it’s a snap to unscrew the roller from the baseplate and to cut/remove any connecting wiring. Usually, I also remove the insulator pad screwed into the baseplate.
Once all of this is done, then, of course, you remount the baseplate and any temporarily-removed brake shoes.
Note: Care is required to remove and replace the baseplate. The spring-loaded driver bearings are held in place by the baseplate, imparting significant spring force on the baseplate, which will bend the baseplate unless you uniformly ease in and out the retaining screws. Slight bending is OK but significant, local kinks are bad. Also, these bearings can easily pop out of place, so inspection is required to ensure the bearings are properly seated before tightening down the baseplate.
Steam locomotive lighting includes some or all of the following:
We’ll discuss each of these in turn
The headlamp is special because we usually want to control its on/off automatically as the locomotive changes direction and possibly according to “Rule 17” which dims the headlamp and tail lamp according to whether the locomotive is stationary (dim) or moving (full brightness).
In my experience with Samhongsa locomotives (Sunset 3rd Rail and Williams), the headlamp is powered by one of two means:
A separate, small rectifying board. It is powered by AC from the center rail and AC from the outside rails and turns on/off power to the headlamp as the locomotive moves forward/backward. This technique is typical of older Williams locomotives.
An output on the Constant Voltage Unit. In this case, you cut the power from the CVU to the headlamp, and you supply power to the headlamp with a two-wire connector, as described in the previous case. Of course, you should directly trace the wires to the headlamp before you cut any wires. This arrangement is typical of Sunset 3rd Rail locomotives. See the Figure below. When I first started these conversions, I was super-conservative and electrically isolated the CVU from the chassis, and I made a direct, wired connection to the Battery- (ground), as shown in the Figure below. But, using chassis ground, which will be electrically connected to the tender’s Battery- (ground) through the loco to tender plug has worked out OK.
Marker lights are almost always LED’s whose power is supplied by the Constant Voltage Unit. No modification of the wiring from the CVU to the marker light LEDs is necessary since once you supply power to the CVU, then it will in turn properly supply the marker light LEDs.
As with the marker lights, once you supply power to the CVU, then the CVU is designed to supply the proper power to the smoke units. A switch is always mounted on either the locomotive or the tender to turn on/off the smoke unit. No smoke wiring modifications are required.
All of the smoke units controlled by CVU’s in my experience are Seuthe smoke units, typically requiring 6V DC. They have the disadvantage that they do not “chuff” in coordination with steam locomotives’ piston action. Replacement of these venerable, but unrealistic smoke units with smoke units that contain a “fan” that propels the smoke in synchronization with the sound chuffs from Lionel, ESU, or MTH is a big topic that will be covered in another blog.
The smoke unit pictured below is very similar to the Lionel 610-8057-200 smoke unit pictured above. It was retro-fitted with a 100K thermistor to emulate the ESU smoke unit inputs for a LokSound L V4.0 decoder.
A final note of caution on smoke units: When using smoke units with metal cases, it’s crucial to ensure that the heater element (resistor), fan motor, and thermistor (if there is one) are completely electrically isolated from the metal casing that will be in firm contact with locomotive ground.You should verify this with an ohmmeter. If you don’t check each lead to verify there is no short to the smoke unit case, you will be sorry. I have forgotten twice (I learn slowly sometimes) to check for electrical isolation of these components from the smoke unit case, and I poured amps from a battery in the tender to the locomotive along a direct path from Battery + to Battery – (ground), melting lots of wiring along the way. This error will cost you hours of work and possibly dozens to hundreds of dollars in damaged electronics.
Original Equipment Manufacturers (OEM’s) often made a point of ensuring good ground contact to some of the smoke unit components – we don’t want this. We want the decoder (or the RF receiver if its handling smoke unit control) to handle how these components are electrically-routed to ground! Yes, these components all need power which we frequently supply from a common source, but it’s the route to ground that we want to carefully control.
As a side note, pay careful attention to the voltage requirements of the smoke unit components. Different voltages almost always supply the heating element and motor! I have never seen a smoke unit motor that didn’t require 5 Volts. I have inadvertently killed one of these cute little motors by accidentally supplying it with 14.8V.
The heating element’s voltage depends very much on its resistance (they vary from 6 ohms to 28 ohms) and interaction with the electronics that control it, such as a DCC decoder or an Airwire RF receiver such as the G3. You need to carefully understand the specs of the heating element and the device that will control it. Spend some time reading. As a general rule, you tend to stay out of trouble with higher resistance heating elements because for a fixed supply voltage, the power produced by the heating element is inversely proportional to the resistance (remember: Power = (V*V)/R, where power is in Watts, V is in Volts, and R is in Ohms).
Even if you are within the proper voltage supply range for a smoke unit, you can still burn them up under certain circumstances. Pay heed to the dark warnings that you should always run your smoke unit loaded with the proper amount of smoke fluid. As the fluid evaporates, it cools the heating element – no fluid and the heating element and the specialized batting that wicks the smoke fluid into contact with the heating element will get hotter. Often, it’s the wick that gets scorched with too high temperatures or no smoke fluid, and this scorching damages the wicking action.
That’s why some smoke unit manufacturers, such as ESU, use thermistors to measure the heating element’s temperature and send this information to the decoder so that it, in turn, can adjust the voltage to the heating element to prevent burn-up. So far as I know, only ESU DCC decoders have thermistor inputs for this smoke unit feedback control. I have reverse-engineered an ESU smoke unit and found that it uses a 100K thermistor that can be easily purchased and retro-fitted into other smoke units if you intend to use the smoke unit with ESU decoders. High-melting point solder must be used when adding a thermistor. Lacking that decoder feedback control of the smoke unit’s heater, I have devised a compact temperature feedback control, but that’s for another blog…
Constant Voltage Unit
We have already discussed the Constant Voltage Unit (CVU) a bit under the section dealing with headlamp modifications. Some locomotives have a Constant Voltage Unit (CVU) that converts AC power from the outside rails/chassis (ground) via the locomotive wheels and center-rail AC via the center pickup shoes, respectively, to a constant voltage for components such as smoke units, LED marker lights, and cabin lights. The CVU usually consists of a rectifier to convert AC to DC, a capacitor to “smooth” the DC, and a voltage regulator to maintain the output DC at a constant voltage, frequently six volts as required for Seuthe smoke units used by Samhongsa-manufactured locomotives (Sunset 3rd Rail and Williams locomotives).
I have encountered three types of CVU’s:
MTH Proto-Sound 2/3 versions that power maker light LED’s and incandescent cabin lights (see my blog on dead-rail conversion of a Proto-Sound 3.0 locomotive).
Sunset 3rd Rail CVU’s that are usually mounted at the front of the boiler and power Seuthe smoke units and marker light LED’s.
Williams CVU’s that are similar to Sunset 3rd Rail CVU’s, but do not control the headlamp, and instead use a separate Headlamp Direction Board.
Important Constant Voltage Unit Control Options: In the figures and discussion above, I modified the Constant Voltage Unit power supply to connect to Switched Battery+ and Battery- (ground) from the tender. These connections will always turn on the devices supplied by the CVU (although the locomotive-mounted smoke switch will turn on/off the smoke unit) when you switch on the battery in the tender. To allow the receiver to control the CVU-connected devices actively, you can replace the Switched Battery+ input to the tender-to-locomotive connector with a relay, controlled by the receiver, that turns on/off Battery+ to the CVU. See the Figure below for an example. You mount this relay in the tender, not the locomotive!
In this example, a 215H-1CH-FC12VDC relay, with a 1N4002RLG protection diode in parallel across the input control (these parts are available from mouser.com by cutting and pasting the above part #’s into their search), will connect/disconnect Switched Battery+ to the locomotive’s CVU according to whether the Airwire G3’s RM1-8 output is on (shorted) or off (open-circuit). Unpowered (i.e., RM1-8 is off), the relay is “normally open”, so no power reaches the CVU from the relay. When the relay activated (i.e,, when the RM1-8 is on), the relay closes to supply Switched Battery+ to the CVU. Setting up the on/off logic for the Airwire G3’s RM1-8 output by remote control is described here, recognizing there are slight differences between the device part numbers and wiring hook-up described in the link and those presented here. It’s kind of neat to hear the mechanical relay quietly click closed and open…
Of course, there are numerous other techniques for implementing a high-amperage switch that is controlled by the receiver or DCC decoder. This particular example is called a “high-side” switch that connects/disconnects battery power to the device, as contrasted with a “low-side switch” where the device’s connection to ground is controlled by the switch.
Locomotive to Tender Connector
Good news: tender conversion is usually much easier and less diverse than the modifications needed for dead-rail conversion of locomotives.
Reverser Unit Removal
Sadly, the sophisticated electronics contained in the tender must be removed almost totally, including the reverser unit. These electronics are not to be confused with a possible MTH PS-2 or PS-3 controller that you probably want to keep as described in another of my blogs.
As I have mentioned in other blogs, it’s a shame that you cannot reuse some of the electronics containing the sound data, but I have not discovered a way to do this. (I still have hope of someday figuring it out, so I have kept and labeled all of the electronics I have removed – you should, too.)
I strongly urge you to take pictures of the tender assembly and its electronics before removal or modification of any components – you might be glad you did. Document everything with a picture – that’s what cell phone cameras are for!
The board removal process first consists of carefully disconnecting any wiring to the boards (with pictures and notes!). Usually, plugs provide these connections, but occasionally, clipping off the wiring becomes necessary. I would suggest that you leave enough wire connected to the board to facilitate a new wiring connection just in case you reuse or sell the boards.
Some of the wiring such as to tender lights and connections to the locomotive need to be carefully preserved.
After wire removal, then the boards can usually be separated from the tender by removing mounting screws or two-sided tape.
As I have mentioned in other posts, I mount the batteries for O scale steam locomotives in the tender. For 14.8V operation, I use 4 LiPo cells in series (the so-called “4S1P” configuration). Regardless of how you ultimately arrange the individual 3.7V LiPo cells relative to each other, you can’t get around the fundamental size limitations imposed by a single LiPo cell on O scale: the cell’s length. Generally, O scale tenders are not wide enough or tall enough internally to provide sufficient clearance for even a single LiPo cell aligned along the width or height of the tender. So, this almost always means orienting the cells’ lengthwise in the tender in some fashion.
Even knowing this constraint, fitting a 4S1P battery pack into the tender sometimes requires a bit of ingenuity/trial-and-error. There are no hard-and-fast rules here, so I will provide a series of figures showing how I’ve “done it.”
Use thin Velcro tape to mount the battery to the tender frame, usually on the chassis and not the tender hull. I mount the “hooks” side of the Velcro to the frame, and the “loops” side to the battery – I suppose this is an aesthetic choice, but I believe consistency is good: so my rule is all Velcro on the frame are “hooks,” and all things that are mounted using Velcro “loops.” See picture below for the kind of Velcro I use.
Use three position, ON-OFF-ON switches when connecting the battery “+” side (the battery “-” should be well-connected to “ground”, which should include the tender and locomotive frames):
ON: power supply for all items on the tender and locomotive
OFF: “neutral” that completely disconnects the battery to prevent slow discharge when not in use
ON: recharging jack
Use high-quality switches and recharging jacks. Cheap ones available on Amazon are generally foreign-made junk. Spend two or three dollars apiece and purchase them from reputable electronics distributors such as Mouser and Digi-Key. I have been sorry when I used cheap hardware.
Placing an antenna and external electrical connections is never an ideal process: compromises must be made that try to minimize the impact on aesthetics and maximize RF reception and accessibility.
While far from ideal, I frequently place the antenna in the dark reaches under the tender chassis. The RF engineers are going to jump all over me for doing so because there’s a lot of metal in a nearby plane parallel to the antenna axis – the location “works,” but the reception range takes a hit. The Figure below is a typical example of mounting for the antenna, an electrical switch, and a charging plug.
Sometimes, when the tender has a coal load, the antenna can be hidden underneath a “coal cloth” on the top of the tender. This strategy sometimes improves reception when there is an opening beneath the coal load where you can place the antenna. Frequently the original coal “load” is either mounted with screws (if you look carefully) or is easily pried off with the mounting board. See Figure below.
At the risk of offending some purists, sometimes an opening can be cut in the tender hull where the coal load goes to afford both better antenna placement and internal access to the electronics in the tender. The opening will be covered over by the coal cloth. See the Figure below for an example of this strategy.
Antennas themselves involve some compromise because they must be short to fit within the confines of our locomotive or tender or in the area about them. Very generally speaking, shorter antennas are less efficient than longer ones. Most dead-rail transmitter/receivers operate (in the US) in the band ranging from 902 MHz to 928 MHz, which is one of the so-called “ISM” (Industrial, Scientific, and Medical) bands that does not require strict licensing to use. This frequency converts to wavelength in vacuum is about 33cm, which at even “quarter-wave” makes for an antenna of roughly 8.25 cm or about 3.2 in unless, for instance, you reduce the length by making the embedded antenna wire helical.
Compact, high-quality, quarter-wave 915 MHz ISM antennas are manufactured by reputable vendors such as Linx Technologies in convenient right-angle form-factors with RP-SMA connectors (see Figure below) and are available at Mouser or Digi-Key. Again, my suggestion is not to purchase cheap antennas of unknown quality: when range performance is poor it will be very difficult to know what to blame. Just don’t worry about the antenna and buy quality: it’s only going to cost you seven to 11 dollars US.
Because this topic is not unique to 3-rail, please see this post for a discussion of Transmitter/Receiver Addition.
Three-rail locomotives have a variety of means for providing two aspects of lighting: constant output with changes in track voltage, and whether to turn on or off with direction or independent of direction, especially since AC operation makes it more challenging to determine direction.
We have already talked a bit about the role of constant voltage units (CVU’s) for lights in the locomotive, and tenders often have colored marker lights and almost always have a white tail light. AC voltage from the rails usually powers these lights, so almost all of the electrical “infrastructure” used to supply AC lighting power must be replaced for battery-powered DC operation.
These lights are often older incandescent bulbs that are more power-hungry than LED’s. Since dead-rail conversion involves the use of DC battery power around 14.8V or so, LED’s with the proper form factor and protection resistor to replace these incandescent bulbs are easy to find for colored marker lights or the rear light. The figures below are an example of LED installation in marker lights and tail lights.
If the original marker or tail lights are LED’s, then retaining most the wiring supplying power is desirable. The wiring to the LED’s must be cut somewhere to convert to battery-powered operation, but look carefully for the protection resistor in series with one of the power leads to the LED, and do NOT cut the wiring between this resistor and the LED. As a side note, if you connect power directly to an LED without a protection resistor in series with the LED: if the power supply “forward biases” the LED, you will burn it out with too much current – a brief, tiny nova will ensue. Frequently, the protection resistor is near a plug, making it a convenience to retain this plug.
For 14.8V operation, the LED’s I have used require protection resistors ranging from 470 ohms to 1.2 Kohm. The specific value is dependent on the LED: smaller LED’s generally tolerate smaller maximum currents, so larger protection resistances are required.
LED’s produce light only if you apply a positive voltage to its “anode” relative to its “cathode.” It’s the electrical equivalent of a stop-valve. In essence, current flows easily through a diode in only one direction, so even in the original wiring, either a positive DC or positive rectified AC voltage supplies the LED anodes.
Ensure that the original DC or rectified AC voltages are reasonably close to replacement battery voltage so that you do not need to modify the resistance of the protection resistor. Make sure that the replacement power connects with the same polarity as the original power supplied to the LED. If powered “backward” for very long, damage to the LED may result. Brief reverse biasing with the protection resistor in series with the LED to test for the proper polarity is not harmful, just don’t apply it for long. The 14.8V batteries I use are generally not vastly different from the original voltages supplied to the LED’s, so increasing the protection resistance is usually not necessary.
After connecting the LED to battery power, if the LED seems too bright, by all means, add a resistor in series with the LED and the original resistor to reduce the current through the LED. A higher-than-necessary value for the total series resistance will not cause problems, and frequently LED’s are just too darn bright using the “proper” value for the protection resistor.
Caution: This topic is somewhat controversial. Some may not like the suggestions under this topic. Please try to remember that the discussion is only offering possibilities, not hard-and-fast rules.
The most important mechanical modifications to 3-rail locomotives deal with the “high-rail” flanges on all of the wheels. If you intend to use high-rail trackage, then the discussion below is probably irrelevant.
However, if you, like me, dislike the very non-prototypical appearance of the high rail trackage and the large wheel flanges that go along with this kind of trackage, or if you want to operate on 2-rail track that conforms to the “Scale Track, Standard Scale” NMRA Standard S-3.2, then the high-rail flange profile requires modification to better match “Scale Wheels” in NMRA Standard S-4.2 and “Wheel Contour” NMRA Recommended Practice RP-25.
There are no doubt numerous strategies for replacing the high-rail flanges, and I’m sure I haven’t thought of them all. The two approaches I will discuss are wheel replacement or flange profile modification.
A final note of caution before describing details: Please determine how you will use your dead-rail locomotive. If you will be using the locomotive strictly on un-powered rails, as I do, then proper wheel insulation is not an issue.
However, if you intend to run your locomotive on powered trackage, then proper wheel insulation may be required. If the powered trackage is 3-rail, then insulation is not an issue since the wheels are by design uninsulated. However, if the trackage is powered 2-rail, then proper wheel insulation is required. Operating 3-rail locomotives on powered 2-rail requires wheel insulation, which is a major hassle, especially when dealing with locomotive drivers.
The powered 2-rail standard is that the locomotive frame is electrically-connected to the right rail, so all of the locomotive’s left wheels must be insulated; and the tender’s frame is electrically-connected to the left rail, so all of the tender’s right wheels must be insulated.
My suggestion: don’t operate dead-rail O scale locomotives on powered trackage at all, but most notably not powered 2-rail trackage. But, this may not be an option for you, so you must address proper wheel insulation for powered 2-rail operation, and you must ensure the locomotive and tender frame do NOT share a common ground. Otherwise, a short across the two rails will occur.
Wheel replacement may involve completely replacing the high-profile wheel with an RP-25 conformant wheel, or, in the case of steam locomotive drives, replacement of the outer rim or “tire.” We will discuss each in turn.
Complete wheel replacement is a very feasible option for leading/trailing locomotive trucks and tender wheels. Specialty vendors such as Northwest Short Line (NWSL) and Precision Scale offer numerous sizes and styles of RP-25 compatible wheels and axles. I will warn you that navigation of these companies’ offerings is either fun or tedious, depending on your level of desperation and patience. While looking for what you need, remember that this is supposed to be a fun hobby.
The knottiest problem with wheel replacement is dealing with the axles. You can completely replace the wheel axles along with the wheel replacement, or you can bore out the replacement wheels’ inner diameter to match the original axle’s diameter.
For instance, NWSL almost entirely offers both replacement wheels with an inner diameter of 1/8″ for O scale and axles that match this diameter. High-profile wheel axles are almost always a larger diameter. If you can find a 1/8″ axle that matches the length and end-style of the original axle, by all means, go with this option.
Unfortunately, you may not always find a suitable axle. If you are handy with a lathe, then making a new 1/8″ diameter axle that matches the original axle’s length and end-style will be a fun project. If machining is not your forte, then you might try boring out the replacement wheel to match the original axle’s diameter.
I will warn you, however, that boring and mounting the replacement wheel on the original axle is somewhat challenging, and requires the right tools, technique, and skill.
As a final note on wheel replacement, ensure that if you intend to operate your locomotive on powered 2-rail trackage (which I don’t recommend), then proper wheel insulation is required. The vendors above offer insulation options along with replacement wheels and axles.
Flange Profile Modification
The NMRA Recommended Practice RP-25 and NMRA Standard S-4.2 help define the wheel flange profile that will operate reliably on two-rail trackage meeting the NMRA Standard S-3.2. These wheel profile and trackage standards are more prototypical-looking than high-profile flanges and track, and RP-25 profile wheels will work on high-profile rail if it’s well laid, but high-profile wheels will not operate reliably on NMRA Standard S-3.2 compliant turn-outs.
RP-25 compliant replacement steam locomotive drivers are generally not available, so modification of the original drivers is usually required.
In some cases, it is possible to remove the driver center from the “tire.” Then an RP-25 compliant tire is machined and press-fit onto the driver center. Some modelers, such as Joe Foehrkolb and Glenn Guerra excel at this process, but I don’t, so my alternative is to machine the tire while still integrated with the center driver until its flange profile approximates the flange depth (approx 0.036″) and flange width (0.039″) dimensions of an RP-25 profile.
The following photos demonstrate my process. Of note is that the most critical dimension to “get right” is the flange depth, which means reducing the flange diameter with a series of small “side-cuts” (parallel to the lathe spin axis) on a lathe until the flange diameter is about 0.072″ larger than the “tire” diameter. Then a series of small face-cuts (radial to the lathe spin axis) reduces the flange width until it is approximately 0.039″. However, this dimension is less critical than the flange diameter, so the flange width should not be reduced by so much that it leaves a sharp flange edge.
One of the most vexing challenges for me when doing dead-rail conversions on O scale steam locomotives is what battery to select. As nearly as I can determine, lithium polymer (LiPo, see LiPo Wikipedia) batteries are the almost-universal choice among dead-railers. So, the choice of battery technology was not much in question for me, but the battery voltage and size is still a choice to be made. I’ll deal with each of these in turn.
I’m not sure how I fell into the 14.8V “camp” for LiPo batteries on O scale dead-rail conversions. Maybe it’s because CVP pushes them for their Airwire products (CVP Airwire). I think CVP has valid admonitions about using higher voltages in regards to radio control range performance and cooler operation (see for instance CVP G3 Decoder User Guide, page 12). And, and almost all of the O scale operating modes (2 rail DC, three rail AC, TMCC, DCS) seem to have 24V or so maximum operating voltage, which is getting close to damage threshold for some radio receivers such as the CVP Airwire receivers. So, would 11.1V (3 LiPi cells in series, “3S” as discussed later) work? Probably just fine in most cases. The plus for 14.8V LiPo is that vendors offer a variety of battery physical and power configurations at 14.8V, which I’ll address next.
There is probably a good deal of lore and religion surrounding the specifics of which brand and configuration of LiPo battery to choose for dead-rail. I’m going to try to stick with what I have tried, not what might theoretically be “better” or “worse.”
Let me start with the less controversial part of my decision process: configuration. LiPo batteries come in a large variety of sizes and configurations (see for instance this site), but in my personal experience with dead-rail, the LiPo’s seem to have a single cell size roughly that of an AA cell with 3.7V output with a charge capacity of around 2000 to 2600 mAh. The individual cells are connected to achieve a total output voltage of about 14.8V (four connected in series, thus the term “4S”), but what varies is the total charge capacity and its ugly handmaiden – physical size.
This is the bear: We want large charge capacity for longer running times, but we in O scale must fit the batteries in often-tight spaces (at least compared to G scale) such as tenders where we also put radio control receiver boards, sound cards, speakers, etc. Of course, HO scale has even more severe volume constraints, but with approximately one-sixth the mass the locomotives must pull.
Personal experience here: I tried to fit an eight-cell (two rows of four vertically-stacked cells, 2.6″ x 2.8″ x 1.4″, 6000mAh, CVP BATT2) configuration into several O scale tenders, and that battery pack just flat-out would not fit. Even though this configuration was only one cell high, O scale tenders are just not tall enough to fit even one cell oriented vertically – the cells must lie sideways to fit. This cell-length limitation is important to remember for O scale. You might find space in diesel locomotives, but not steam locomotive tenders. These limitations were a big disappointment for me since I wanted to cram a large storage capacity battery pack in the tender and run “forever” (forever being at least four hours).
Sigh… Backing off in size, I have found that 2x2x1 LiPo battery packs will fit in O scale tenders with the individual battery cells running along the length of the tender. See the Figure below (pardon the body parts). You can see that this configuration also comfortably fits the width limitations of O scale tenders.
A note of caution: you cannot stack too much on top of this configuration before you lose vertical clearance inside the tender. For instance, I thought this was going to work:
Stacking the receiver board on top of the battery would be a useful space-saving strategy, but this configuration would not clear vertically in all tenders I tried (Big Boys, Cab Forwards, Challengers, and Alleghenies). The receiver board manufacturers would probably dislike my mounting electronics on top of batteries, even with sufficient clearance.
My experience is confined to three battery suppliers of 2x2x1 LiPo battery packs: CVP, Tenergy, and “HJE.” All come with a Protection Circuit Module (PCM) that provides:
More creative 14.8V battery configurations are possible that include one battery-diameter thickness solutions such 1x2x2 (Tenergy.com) with dimensions of 131 x 36 x 23mm (LxWxT), so it’s approximately one battery-diameter thickness, two battery-diameters wide, and two battery lengths long (~5.2″). See the picture above for an example of using this thin configuration in a very tight mounting configuration on an MTH Virginian Triplex with a PS-3.0 board operating in DCC mode. See this blog for more details.
Don’t jump to conclusions; MTH did NOT manufacture the C&O 2-6-6-6 Allegheny with MTH Proto-Sound 2.0 (PS2.0) I found on eBay. Instead, it’s a brass locomotive produced by Sunset 3rd Rail (Figure 1) and converted to PS2.0 (see MTH PS2.0 Upgrade Manual). See Figure 2. The retrofit replaced the original tender QSI-OEM Digital Soundboards, the wiring harnesses for the tender and locomotive, and the Suethe smoke units and their voltage regulator board in the locomotive.
It was a well-done conversion, so I was very reluctant to tear out the tender PS2.0 control board, and the wiring harnesses in the tender and locomotive. The PS2.0 conversion used an MTH smoke unit that has both fan speed and smoke intensity controls.
The CVP Airwire receiver boards I typically use for dead-rail conversion don’t have this level of smoke unit control. And, the PS2.0 board used a speed encoder on the locomotive motor’s flywheel to synchronize the PS2.0 board’s sound. See Figure 5 for the speed encoder reader and flywheel strip and 5b for additional electrical connections.
All of these built-in features were pretty nice, but I still wanted a dead-rail conversion.
Hmm… Looking around, I found out that Proto-Sound 3.0 (PS3.0) had DCC/DCS control options and the wiring harnesses are the same for PS2.0 and PS3.0 boards (see for instance MTH PS3.0 Upgrade Manual). Things are looking up. Then I found a great site, Ray’s Electric Trainworks, that provides PS3.0 replacement boards and great support.
My thought was this: if I can upgrade the locomotive to PS3.0, then I can take the following steps. Jumper the PS3.0 board to DCC operation, disconnect the original rail power/communication wiring, and re-connect the rail power/communication wiring to the DCC outputs of a CVP Airwire CONVRTR-60X receiver (CONVRTR Users Guide). Easy, right? Not so fast.
Ray at Ray’s Electric Trainworks was a great help: he steered me to the right PS3.0 card that I needed for the tender, and he loaded the Allegheny sound file for me. Otherwise, I would need a bunch of DCS infrastructure to load the sound file. And, he gave me a rebate for the old PS2.0 card! Great guy.
OK, I have the PS3.0 card from Ray. The PS3.0 card came mounted in its plastic carriage that is screw-mounts on the tender chassis through pre-existing holes. The heatsink orients a bit differently between the PS2.0 and PS3.0 – no big deal – I just needed to drill a hole in the tender chassis in a slightly different place. The PS3.0 doesn’t use a Ni-MH battery, so out it went. That was a good thing, too, since I needed the real-estate for the replacement LiPO battery that would supply power to the control boards, lights, smoke unit, and the locomotive.
Since I wouldn’t be sending track power to the tender, I cut the Ground Lead and Pickup Roller Leads wires. I re-purposed them by connecting these harness wires to the Constant Voltage Unit, which is a heat-shrink blob whose input leads were cut from their original chassis connections (see Figure 5 again).
While I was at it, to reduce power consumption, I removed the incandescent cabin and headlight bulbs, and I replaced them with Yeloglo LED’s (see Yeloglo description) whose + input was in series with the Yeloglo’s 470-ohm resistor for 10-16 Volt operation. Yeloglo LED’s have an excellent yellowish output reminiscent of incandescent light.
The diagram below shows the tender’s wiring harness with my modifications.
Continuity testing revealed that the locomotive Roller Pickup Leads and the Ground Lead connected to Pin 1 and 3 of the 7-pin connector, respectively.
Note that BOTH locomotive and tender Ground Leads (both are black) that are input to the PS3.0 on pins 3 and 4 of the 7-pin connector, respectively, MUST BE DISCONNECTED FROM THE LOCO/TENDER FRAMES AND CONNECTED to the “B” DCC output of the CONVRTR. This pair of connections was the trick. I initially connected ONLY the Ground Lead coming from the locomotive (pin 3 of the 7-pin connector) to the CONVRTR, and it did NOT work! I don’t know if only connecting pin 4 of the 7-pin connector would work – I didn’t try it.
Similarly, I disconnected the red wire that is input to the PS3.0 on pin 1 of the 7-pin connector (which originally connected to the locomotive Roller Pickup Leads) from the plug bundle that connects the tender to the locomotive, and I connected it to the “A” output of the CONVRTR.
The “harness side” of the two cut wires originally going to the Pickup Roller Leads (red) and Ground Lead (black) on the locomotive, were connected to switched Battery + and Battery – (ground), respectively, to provide power to the locomotive’s Constant Power Unit.
Note: “Switched B+” means battery power coming from the Battery’s + terminal that is turned on or off with a switch (you want to be able to turn off the power!). “Battery – (ground)” means the connection to the Battery’s negative terminal that is usually grounded to the tender chassis by a battery charging plug.
A picture is worth a thousand words, so studying Figures 6 and 7 will give you the story of what wiring cuts and re-connections that are needed to convert a Proto-Sound 3.0 steam locomotive to dead-rail.
Other Dead-Rail Conversion Details
Of course, there are other aspects to the dead-rail conversion that are required. These aspects include the addition of battery power and CONVRTR connections and removal of center-rail pick-ups and electrical connections that are part of a typical 3-rail to dead-rail conversion for an O scale steam locomotive. These conversion aspects are discussed in another blog.
In summary, if you have a locomotive with the PS3.0 installed, conversion to battery-powered DCC operation and radio control (dead-rail) is straightforward once you know the few wiring cuts and re-connections you need to make. The DCC operations for this particular locomotive can be found in the MTH document “Premier 2-6-6-6 Allegheny Steam Engine .” What you preserve with the PS3.0 is good DCC functionality, the original sound, and coordinated smoke – and that’s a pretty nice combination.