1694 Calle Zocalo
Thousand Oaks, CA 91360
1694 Calle Zocalo
Thousand Oaks, CA 91360
The Whistle/Horn button on a Lionel “Multi-Control” transformers is a two-stage switch. The purpose of the two stages is to first generate a large DC component in the track voltage to pull in the Whistle/Horn relay in the locomotive or tender, then drop the DC voltage to a smaller amount that will hold the relay in the activated state.
When the Whistle/Horn button is activated to stage one, a diode is inserted in series with the output voltage AND an extra “Boost” transformer winding of about 5 volts AC is also inserted in series. The overall output drops slightly because of the loss of half the waveform, but the large DC component pulls in the relay.
The switch is then quickly moved to the final stage to bypass some of the missing AC waveform around the diode. The bypass is a resistance wire that takes different forms in the various Lionel transformers, depending upon the transformer’s peak current rating. This bypass, combined with the Boost voltage, actually raises the output voltage above the original non-whistle voltage. This extra voltage boost was deemed necessary to provide extra power to drive both the locomotive and the whistle motor so that the train does not slow down. Note that the waveform is shifted upward slightly by the remaining DC voltage.
The above waveforms will vary for different models of transformer and the amount of current being sourced. These photos were taken with a Lionel Model RW 110 watt transformer feeding a 6.7 ohm resistive load.
The DC-sensitive relay is an interesting bit of engineering. Note the two heavy copper rings around the core of the relay. These shorting rings give the relay a very high AC impedance because of the large opposing magnetic field generated in the copper rings
by any AC magnetic field. The rings, however, do not impede any DC currents that want to flow since the DC currents do not produce any rapidly changing magnetic fields. The DC magnetic flux operates the relay by pulling in the armature (which is not attached in this photo.) This relay will ignore 20 volts AC, but respond to about 1.5 volts DC!
This video shows Joe O’Loughlin’s Novelty Layout using components furnished by MANCO. The Beep automatically shuttles from end to end on the underside of the plate shelf in his living room. The display is a favorite with visitors! Thanks, Joe!
The following discussion will cover the actual transmission of train control information over wireless links for the Lionel TMCC and Legacy control systems. There are four types of communication:
Wireless CAB-1 to TMCC Command Base and Powermaster (and Bridge to Powermaster) at 27 MHz
Wireless CAB-2 to and from Legacy Command Base at 2.4 GHz (covered lightly)
“Wireless” TMCC Command Base to locomotives, operating accessories, ZW, SC1 and SC2 at 455 KHz via the “Track” signal
Wired serial data in RS-232 format for power controllers and accessory controllers (not covered here)
Before we delve into the various links, we should build up a bit of background in how an antenna works. Radio waves propagate through space (including an empty vacuum) by continually exchanging energy between electric and magnetic fields. The electromagnetic wave can be of any frequency higher than DC, including radio, microwave, infrared heat, light, ultraviolet, X-Ray and gamma rays.
All of these waves travel at the same speed in a vacuum – about 186,000 miles per second. A radio wave takes about 3 seconds to bounce off the Moon and return to Earth. We can easily calculate the distance between crests of the wave by dividing the distance traveled per second by the number of full cycles of the wave per second.
Wavelength = Speed/Frequency
For our three wireless links, the wavelengths are
TMCC handheld 11 meters or 36 feet
Legacy handheld 12.5 centimeters or 5 inches
TMCC/Legacy Track 660 meters or 2160 feet
We are going to transmit and receive these wireless signals through various antenna configurations. Fortunately for our discussion here, anything that makes an antenna good for transmitting also makes it good for receiving. This is known as “reciprocity”. The only difference is that some transmitting antennas must carry more current and voltage because of the higher transmitting power.
Our electromagnetic wave is generated around a flowing current. Currents are generated by differences in electric potential (voltage), and the resulting flow of current produces a magnetic field. If we set up a way to pump current into our antenna at a rapidly reversing rate (our operating frequency), we will generate electric and magnetic fields around the antenna that will carry power away from the transmitting antenna as a wave.
Thanks to reciprocity, if our transmitted wave strikes a receiving antenna, the wave’s electric and magnetic fields will generate a (small) current in the receiving antenna. We can selectively amplify and detect that signal to recover our transmitted information.
To illustrate these principles, we will use a simple dipole antenna as our example. We all used dipole “Rabbit Ears” in the early days of television. We soon learned that varying the length, angle and orientation of the two rods could help or hinder the quality of TV reception. The dipole antenna is simply two conductors stretched in opposite directions, with a transmitter or receiver feeding or receiving signals at the midpoint of the array. The dipole is a “balanced floating” antenna, meaning that the two halves mirror each other and there is no antenna connection to earth ground.
Simple dipole transmitter and receiver systems
The transmitter establishes equal and opposite currents in the two radiating conductors. Usually the antenna length and characteristics are chosen to tune the antenna to resonance so that maximum power can be coupled from the transmitter to the antenna. Note that the driving current is passing through the transmitter. Whatever current comes in one side is also going out the other side.
If we were to scan along the length of the antenna conductors, we would find there is a voltage wave and a current wave. In essence, the current wave charges up the voltage on the antenna, with the limitation that the current is always zero at the far ends of the conductors. Electrons don’t jump off the ends of the conductors into nearby space. All of the radiating of power is accomplished by the varying electric and magnetic fields.
Our receiving system operates exactly in reverse. The incoming electromagnetic waves generate a voltage difference along the receiving antenna’s elements, and that voltage difference causes a current to flow through the receiver at the middle of the antenna.
Unfortunately, nobody operating a TMCC system wants to have a pair of Rabbit Ears sticking out of their CAB-1. The usual choice is to replace the dipole antenna with a monopole or single-rod antenna. This category includes collapsing whips, rubber ducks and even the large transmitter towers used by AM radio stations (yes, size matters!)
In reality, the rod is only half of the antenna system. We still need a second component so that we can have current flow through the transmitter and receiver. If we only had the rod, there would be no way to force current to flow into a dead-end circuit. The answer is to add a ground system on the other side of the transmitter or receiver connection so that current can flow back and forth between the ground and the rod, passing through the transmitter or receiver. The term “ground” should not be taken literally. For an AM radio tower, ground may consist of a radial star of copper conductors centered at the base of the tower, buried in the dirt. For the rod antenna on an automobile, ground is the metal body of the car.
An ideal monopole antenna would be a rod one quarter wavelength long, mounted on a ground surface (ground plane) that is highly conductive. The ground plane acts as a mirror for the antenna rod, creating a second
Monopole antenna with phantom mirror image
“phantom” rod that would extend into the earth from the base of the rod. This creates the equivalent of a vertical dipole. The current flow between the ground and the rod is equivalent to the current flow at the center of a two-rod antenna. (Ham operators – For simplicity I am ignoring things like characteristic impedance.)
For maximum efficiency our CAB-1 antenna should be a quarter-wavelength rod nine feet long, surrounded at the base by a metal disk 18 feet in diameter. Not too portable and convenient? We bow to practicality and do the best we can in the available space. The collapsible antenna is far less than a quarter wavelength (minimum length = 5”) and we use the ground traces of the printed circuit board as our ground plane. As a result, we have a very inefficient antenna system, but it is good enough to cover the area of a normal train layout comfortably. (The capacitance of the operators body may help somewhat, but there isn’t a good connection between the RF ground inside the case and the operator’s body. The electrical gap through the plastic is approximately ¼”, and using the area of a hand palm as 30 square inches, the capacitance would be about 60pF, yielding a coupling impedance of about 200 ohms at 27 MHz.)
The receiving antennas in the Command Base and Powermaster are planar coils of wire that are 28” long. This configuration may be dictated by the space available and the resonant frequency of the coil.
The bottom line is that the CAB-1 communications is much less than optimal. The transmitter’s output power and the receiver’s sensitivity make up for the inefficiencies of the antenna configurations.
We can make some useful observations about the antennas. The CAB-1 antenna radiates sideways from the rod with an annular wave similar to sliding a donut or bagel (choose your favorite) over the rod. The antenna does not radiate much power off the end of the rod. Pointing the rod at the Base is the worst case orientation. At the receiving end the coil of wire picks up best when a wave encounters the loop of wire edge on. Pointing the CAB-1 rod at the top of the Base would be the worst relative orientation.
The Legacy system has a wavelength of only 5”, allowing the use of more ideal transmitting and receiving antennas with higher efficiencies. Since Legacy uses bidirectional communication, the antennas in the Base and handheld are used for both transmitting and receiving.
Everything was fairly obvious in the previous discussion, but things become much fuzzier when we tackle the Track signal. The best way to approach this problem is to work backward from the locomotive to the Base.
The locomotive contains a receiver that is tuned to the 455 KHz track signal. The front end of the receiver couples the transmitted airborne signal to the receiving circuitry. As in our antenna theory discussion above, we need to have two parts to the receiving antenna, and the receiver connected between the two halves. What are the two parts of the antenna? The locomotive has an antenna in the form of an insulated handrail or wire or foil strip. For locomotives with cast metal shells, the antenna is on the outside of the shell. For plastic shelled diesels, the antenna can be mounted under the roof of the shell.
The antenna connects through a wire to the input stage of the radio receiver. As we saw earlier, we also need a second half to the antenna system so that we can get current to flow through the receiver. For this application we use the frame of the locomotive as our ground reference. The received radio signal will flow between the antenna and the frame of the locomotive. Note that the frame is also connected to the wheels and hence the track outer rails. That means that our receiver is sensing the current flow between the antenna and the outer rails.
We must stop here to dispel Myth Number 1 – “The antenna on a TMCC locomotive picks up the Track signal.” As we just concluded, one side of the receiver’s input connects directly to the outer rail, and hence the Track signal. If the antenna also picked up the same signal, there would be no voltage difference and no resulting current flow. The antenna IS NOT PICKING UP THE TRACK SIGNAL. If not the Track signal, then what is the antenna receiving?
To answer that question we must jump over to the other half of the system – the Command Base. The circuit that transmits the Track signal is shown in the following drawing. Transistor Q4 is the final stage that drives
the “U” thumbscrew terminal on the back of the Command Base. This signal is conducted down the outer rails, through the wheels of the locomotive, into the frame and then to the “ground” side of the radio receiver’s input circuit. This “ground” isn’t sitting still at zero volts, but rather it has part of the 455 KHz signal moving it around electrically.
We can consider this path from the thumbscrew through the track and into the locomotive frame as one half of the transmitting antenna for the Command Base’s output signal. We know that an antenna must have two parts so that current can flow back and forth. Where is the second part of our transmitting antenna?
The output signal on the emitter terminal of Q4 feeds positive current out from the collector of Q4, and sinks negative current through R24. Both of these points tie back to the local signal ground in the Base. R24 is directly tied to ground, and the collect of Q4 is tied to ground by the big bypass capacitor C30. That means that the other side of our transmitter signal is the local signal ground in the Base. We need to connect the other half of our antenna to this ground. Now note that one side of the incoming AC, J5-2, is directly tied to this signal ground. This AC lead comes from the wallwart that is plugged into the AC wall receptacle, and in the wallwart the wire to J5-2 is jumpered to the U-ground safety ground pin.
The other half of our antenna is the house wiring! The track signal generated by Q4 flows back and forth between the house wiring and the outer rail. The house wiring radiates the signal through the air to the antenna
on the locomotive, and thus feeds the other side of the receiver’s input circuit. The current in the receiver is an image of the current that flows between the house wiring and the outer rails.
The house safety ground green wire eventually makes its way back to power panel and there it is tied to the neutral and the earth ground rod or clamp on a water pipe. We now have the other half of the TMCC signal radiating from all of the house wiring – safety ground, neutral and even the hot wire thanks to cable capacitance and any electrical loads that connect the hot to neutral through a resistance, and from the earth. We will refer to this connection as “earth ground” to distinguish it from layout common, which is sometimes called “ground”. Earth ground and layout common SHOULD NEVER BE TIED TOGETHER!!!!
Although the TMCC system is intended to have a conduction path through the wallwart via a wire jumper, I suspect that some TMCC signal will be fed into the house wiring just through the winding capacitances of the wallwart’s transformer. I have not measured the effectiveness of this coupling path.
Problems and Solutions
Now that we know how the TMCC Track signal is transmitted and received, we can look at some common problems and their solutions.
1) Insulate the top part of the tender shell from the frame with tape and connect the antenna to the insulated portion.
2) Poke a hole in the shell and route a section of antenna wire outside the shell.
We must avoid configurations that block out the airborne Track signal by having too much outer rail Track signal. Imagine what would happen if you put your locomotive inside a metal box that was connected to the outer rail. You would have lots of Track signal conducted through the wheels, but there would be not airborne signal on the antenna to cause current to flow back and forth through the input stage of the receiver. This is the situation you create if you have overhead bridges and/or trackside metal structures connected to the outer rails of the track, or many parallel tracks.
Now imagine that we drill a hole in the metal box containing our locomotive, and we insert a wire connected to the earth ground signal. Now the locomotive’s antenna can pick up some of the earth ground signal and create a voltage differential across the receiver’s input. We can pick up the earth ground signal from the center screw on a grounded wall receptacle, a metal water pipe or electrical conduit, or Pin 5 of the 9-pin connector on the back of the Command Base.
Note that this overwhelming is not what we would normally consider to be interference or cancellation. The multiple Track signals are not interfering and reducing anything. They are all adding up quite nicely, too nicely in fact. (There is no “interference” because a wavelength .4 mile long is much too long to create cancellations in a space the size of a layout. We are dealing with a simple imbalance.)
Many folks talk about a “halo of track signal”. Indeed there is a signal radiated by the track, but this is really an unwanted signal. Our antenna doesn’t want to pick up this signal, and we are already getting all the Track signal we need by conduction through the wheels.
Years ago a “TMCC Signal Enhancer” was sold to overcome signal weak spots. The Enhancer merely added some of the TMCC signal from the outside rails to the center rail. As we have seen, this is quite unnecessary, and any signal problems are probably due to too much Track signal rather than too little.
I have heard from people who run TMCC outdoors and in other environments that aren’t surrounded by house wiring. There still must be a second half of the antenna system, possibly though an extension cord into the house wiring and then to earth ground at the power panel. The layout outdoors is still over good old earth.
Many of the ideas presented here run contrary to the folklore, but I believe that this is a cohesive argument that will hold up against testing. I don’t claim to be a certified expert, but I do hold a MS in Electrical Engineering, and I have been a ham radio operator for over 20 years. You are welcome to disagree with me and my ideas, but please present factual evidence for any competing theories.
The Lionel TMCC Command Base (Base) receives, interprets and retransmits commands from a TMCC CAB-1 handheld controller. The Base outputs include a signal that transmits through the outer rails of the layout at 455 KHz to locomotives and accessory devices, and a 9-pin pseudo-RS232 port that can drive power and accessory controls that utilize the RS232 protocol at 9600 baud. The 9-pin connector can also be used for bidirectional communication with a computer, in which case incoming TMCC commands generated by the computer will be echoed on the outer rail signal and the 9-pin output.
The power source is a special “wallwart” AC transformer with a 3-pin molded AC power plug. Unlike most wallwarts, the Lionel unit connects one leg of the low-voltage AC outlet to the “U-ground” safety grounding pin on the AC plug. THIS CONNECTION IS NECESSARY FOR THE TMCC SIGNAL TO PROPOGATE PROPERLY!! This link from the Base to the house wiring is an essential part of the antenna system for TMCC.
The Base output terminal for connecting to the track is labeled “U”. The U terminal should be connected to the layout’s track COMMON at a convenient distribution point on a terminal strip or the COMMON terminal of the transformer or other power device feeding the track. The gauge of the connecting wire is not important since there is very little current flowing in the wire. (Do not use a lengthy wire that needs to be coiled up. Coils create inductance that impedes the flow of high frequencies. The same rule applies to the wire running from the Base to the wallwart AND any extension cord that might be used between the AC receptacle on the wall and the wallwart.)
The 9-pin serial connector uses a basic 3-wire RS232 format for the incoming and outgoing serial data. The Base output driver stage only swings from +5V down to ground, rather than a true bipolar swing defined in the RS232 standard. This is a common simplification, and many RS232 devices will operate with this reduced swing. The pins on the 9-pin D connector are:
Pin 2 Data Output \ 5 4 3 2 1 /
Pin 3 Data Input \ 9 8 7 6 /
Pin 5 Ground View from rear of enclosure
There are no provisions on the connector for grounding of a cable shield. (The threaded inserts for locking screws are not grounded.)
The TMCC Receiver
Both the Powermaster and the Command Base have radio receivers tuned to 27 MHz for receiving commands from the CAB-1 handheld controller. The exact frequency is controlled by selecting a pair of crystals for the CAB-1 and Base. The crystal in the CAB-1 is set to a frequency that is 455 KHz higher than the crystal in the Base. The normal frequency pair is 27.255 MHz (Channel 6) in the CAB-1, and 26.800 MHz in the Base. The reason for the 455 KHz stagger will be described later. (Note that the plastic holders for the crystals both say “27.255 MHz” even though the receiver crystal is really 26.800 MHz. The actual frequency is printed on the shell of the crystal.)
The receiver’s antenna consists of 28” of 26 gauge wire wound in a rectangular spiral pattern. (Always check the antenna’s attachment wires after servicing to verify that the wires have not fractured where they enter the PC board due to twisting and turning of the PC board to replace components.)
Command Base circuit board
(The switch and wires at the lower edge of the photo are not standard items for the Base.)
The antenna feeds a one-transistor RF amplifier configured as a Common Base stage, with the collector load tuned to 27 MHz.
Command Base radio receiver circuitry
The boosted RF signal is fed to one port of the Mixer section of a MC3371 single-chip radio receiver. The Local Oscillator section of the chip provides a crystal-controlled 26.800 MHz signal that feeds the other side of the Mixer. When the two signals are mixed together, the resulting signal has frequency components at the sum of the incoming frequencies (27.255 + 26.800 = 54.055 MHz) and at the difference of the frequencies (27.255 – 26.800 = 455 KHz.) A ceramic resonant filter rejects the upper sum and passes the lower difference to the Limiter.
The incoming data is encoded as Frequency Shift Keying (FSK), which means that only two discrete frequencies are used. Since only the frequency information is important and not any amplitude changes, the Limiter greatly amplifies the incoming signal until the positive and negative tips of the waveform are flattened off to create a squarewave with no amplitude variation. The Demodulator uses the slope of a tuned circuit set near 455 KHz to change the frequency steps into voltage amplitude steps that represent the CAB-1’s encoded digital ones and zeros. The Audio Amplifier boosts this small voltage to about .7 volts P-P.
Command Base opamps used as comparators and limiters
displays the signal before limiting, and Test Point #4 is the inverted hard-limited output signal. The conditioned digital signal is fed to Pin 26 of the microcontroller (uC) for final decoding. The uC can accurately measure the difference in arrival times between pulse edges to distinguish between the two FSK data frequencies.
TP3 and RSSI gate with no valid signal present
Demodulated data into edge detector (TP3) top RSSI gate bottom
Data after limiter (TP4)
Note that the left portion of the data word is a series of repeated characters. This “preamble” provides a reference clock that can be used for determining the locations of the data transitions in the latter part of the data word where the actual TMCC command actions are encoded.
Unlike the Powermaster, the Base utilizes the capability of the receiver chip to detect the strength of the received signal. This “RSSI” (Received Signal Strength Indicator) signal is compared to a reference level of .375 volts by U2a (pins 2, 3 input & 1 output). When an incoming signal burst is present, the RSSI level on pin 1 drops LOW for 6.5 ms. This “valid signal received” output is fed to uC Pin 17. (Hint: Use this RSSI signal on Pin 1 to trigger the oscilloscope when trying to observe the data on TP3 and TP4.)
RSSI gate – Low = valid signal
In addition to the FSK data from the receiver, the uC also receives the timing of the zero crossings of the 60 Hz power sinewave. The zero crossing detector applies the full 18 volts of the input power through a 100K resistor to a section of quad opamp U1c (pins 9,10 input & 8 output). A pair of parallel back-to-back silicon diodes connected to ground clips the input voltage to +/- .7V to avoid overdriving the comparator’s input. The opamp’s output produces a 0-4V squarewave synchronized to the AC line on Pin 12 of the uC.
Squarewave 60 Hz
The Base is controlled by a 40-pin PIC 17C42-16 uC. The uC’s clock is generated by an 8 MHz ceramic resonator connected to Pins 19 & 20.
Microcontroller and Serial In/Out
Two LEDs indicate proper operation. The green “Status OK” LED is illuminated directly by the +5V power supply, and the red “Incoming Data” LED is controlled by a pull-down on Pin 24.
RS-232 communications travel through J3. Incoming signals on Pin 3 are limited to a 5V swing by transistor Q7. Outgoing signals are buffered by Q6, a PNP transistor that provides an active pull-up to +5V. This signal typically drives optoisolators in the
Serial data on 9-pin pin 2
various power and accessory controllers that use the 9-pin wiring. No protection circuitry is provided for Q6 except for the limited Base drive of .5 mA. This output is intended to be a current source, and it has very limited current sinking capability through the 1Kohm collector resistor.
The primary function of the Base is to control locomotives via the signal transmitted through the track (and the house safety ground wiring.) The transmitted signal is quite similar to the signal from the CAB-1 except that the carrier frequency is 455 KHz rather than 27 MHz.
In radio circles, 455 KHz is a relatively low frequency that is totally immune to problems created by reflections and multipath that plague TV and cell phone signals. The length of each cycle at 455 KHz is over 2000 feet. A layout would need to be on the order of 500 feet long before there could be self-interference problems. We are only discussing self-interference here, not noise from external sources such as light dimmers, switching power supplies, arcing components and other electrical noise sources.
Oscillator and “U” terminal output driver bottom, Power supply top
A nominal 455 KHz signal is generated by oscillator Q2. The frequency of oscillation is primarily determined by the tuned tank circuit composed of inductor L8 and capacitor C27. To shift the frequency for the FSK encoding, an additional capacitor, C28, is connected across the tuned tank by Q5. This adds about 3% extra capacitance, and this will shift the frequency down by about 1.5% using Frequency=square root(L*C).
The oscillator’s output is amplified by Q3 and buffered by emitter follower Q4. Both Q3 and Q4 are tied to the unregulated input side of the power supply, giving them additional voltage swing for driving the track. Capacitor C32 provides AC coupling to the track.
Track signal at “U” terminal
This output circuit requires more than just a single wire to the track to communicate with the locomotives. The other half of the output signal is provided by the ground terminal of the Base output circuit. Unlike many systems for which we consider “ground” to be a reference plane with zero voltage, the Base ground has an active signal. The TMCC designers utilized a Base power supply with half-wave rectification so that the ground of the output circuit is shared with one side of the incoming low-voltage power supply. The jumper in the wallwart that connects this side of the transformer’s secondary winding to the U-ground pin on the AC wall receptacle feeds the ground half of the output signal into the house wiring network.
The house wiring now becomes a huge antenna that is further augmented by the ground rod connecting the safety wiring to earth ground. The signal radiated by this house wiring and ground is the signal that is picked up by a locomotive’s antenna, NOT THE TRACK SIGNAL!
Command Base Signals
The Base is always transmitting the track signal (unlike the serial output which is only present when a command is being sent), but the frequency is constantly shifting. A frequency counter will display a value that is an average over the sampling period. A single (upper) frequency can be established by grounding the base of Q5, thereby disabling the data line from the uC.
When no data is being transmitted, the track signal consists of an idling No Operation signal that keeps the locomotives in the Command mode.
My test unit measured 457.36 KHz with the base of Q5 grounded, 452.39 KHz with the base of Q5 held high and 454.77 KHz with the idling signal being transmitted. I measured a drift of up to .5 KHz as the unit warmed up.
If the frequency tuning of the Base oscillator shifts away from the nominal value, the receivers in the locomotives may have a reduced sensitivity. Re-tuning L8 is a bit risky, but sometimes this is necessary. Using the least sensitive locomotive as a reference, press the horn/whistle button on the CAB-1. Use a plastic screwdriver to adjust L8 back and forth until you determine the extremes of the adjustment range that will activate the horn/whistle. Now adjust L8 to the middle of this adjustment range. Check all you other locomotives. (You could have just one bad locomotive and a good Base!)
The power supply is a simple halfwave rectifier and a 3-terminal 5V voltage regulator. The energy storage capacitor C23 between the diode and regulator has a 4.5V 60 Hz sawtooth ripple that swings between 15.5 and 11 volts.
Power supply ripple
A shorted output capacitor, C32, can be detected without opening the plastic enclosure. Measure the resistance from the “U” terminal to Pin 5 of the serial port. There should be an open circuit. If the reading is about 100 ohms, the capacitor is shorted and you are reading the resistance of R24.
To check the wallwart’s ground path, measure the resistance from the U-ground pin to the outer ring of the coaxial power plug. This path should be about zero ohms.
If the red light flashes and the serial output works, but the track signal is dead, or vice versa, the problem is probably in the individual output stages. If the red light doesn’t flash and both outputs are dead, but the green light is lit, the problem is probably with the receiver, opamp or uC.
For troubleshooting, you can swap crystals between the Base and a Powermaster.
The CAB-1 controller provides handheld operation of Lionel TMCC Command Control layouts. The pushbuttons and rotary knob provide addressing and control of trains, switches and accessories. A jack on the top end of the unit provides a connection for an external “Last Command” button which retransmits the last active command when pressed.
The CAB-1 is somewhat restrictive in that it cannot access all of the address space provided by the TMCC command functions. The limitations include:
0-99 engines out of 0-127
0-9 routes out of 0-31
0-9 trains/tracks out of 0-15
0-99 switches out of 0-127
0-99 accessories out of 0-127
No access to “Group” commands to activate groups of accessories
The basic subsystems are the keyboard, knob, microcontroller (uC), RF transmitter and battery.
The main keyboard PC is connected to the uC/RF board by an 11-wire flat ribbon cable. For identification purposes, the cable connects to an array of holes marked “J1” on the uC/RF board, and herein Pin 1 of J1 is adjacent to the right end of the “J1” outline in the picture above. Likewise, the rotary encoder is connected by a 5-wire cable to “J2”, with Pin 1 again on the right. An additional small board with the Set, L, M, H and Halt buttons is attached to the main button board by a 6-wire flat cable.
The keyboard consists of an array of buttons that activate conductive rubber pads that jumper between two printed circuit board traces when depressed. The buttons are connected in an array pattern that allows the intersections of the rows and columns of the array to be interrogated.
The PIC 16C57-HS is an 8-bit processor operating at 8 MHz. The pin utilization is:
Switch scanning – pins 12, 13 14, 15, 16, 17, 18, 19, 20, 21, 22, 23
Wheel phototransistor detectors – pins 10, 11
Wheel LEDs – pin 6
Beeper – pin 8
uC clock oscillator – pins 26, 27
Encoded data output – pin 7
Transmitter active – pin 9
uC Master Clear – pin 28 (Reset on bootup)
TOCK – pin 1 (Timer input pin is not used)
+V supply – pin 2
Ground – pins 4, 24, 25
Not connected – pins 3, 5
Rotary control wheel – Red Knob
The rotary encoder consists of two LEDs that shine though a disk with 40 radial slots (9 degrees of rotation per slot) onto a pair of phototransistors. The phototransistors “see”
the LEDs’ light after it passes through a phase grating with fixed slots that are offset by 90 degrees (1/4th of 9 degrees rotation.) This geometry produces two partially overlapping quadrature output signals that encode the direction of rotation, with the direction determined by which of the two signals rises first.
To conserve power, the LEDs are only pulsed ON every 735 us. when the system is interrogating the rotary encoder. The LEDs are connected in series with the top end of the string tied to the battery +V terminal. The bottom of the string is pulled to ground through a 300 ohm resistor to uC pin 6.
The optical components must be properly aligned to produce the necessary 90 degree phase differential and a rail-to-rail voltage swing. The best way to achieve this alignment is to use a dual-trace oscilloscope in “Chop” mode to observe both channels simultaneously by monitoring R7 and R8. The waveforms should extend all the way from +V to ground when the optos are properly positioned. The patterns will vary with very slow knob rotation through the following sequence:
Neither opto on
Only one opto ON
Both optos ON
Only the other opto ON
Changing the direction of knob rotation will reverse the roles of the optos. The optotransistors can be repositioned slightly for alignment by bending their leads.
Note that both the top and bottom halves of the case provide alignment and support to the opto board. When operating the board with the case open, use your fingers to align the opto board to vertical. Always recheck knob operation after the case is closed.
Direct external light may falsely activate the optotransistors. Shield the encoder area from workbench lighting during adjustment and testing.
Radio Frequency Transmitter
The CAB-1 transmits a radio signal at a nominal 27.255 MHz. The actual frequency is stepped between two discrete values, with the transitions between the frequency representing the rising and falling edges of the data being transmitted. This form of modulation is called Frequency Shift Keying (FSK). The TMCC Command Base or Powermaster receiving the signal detects the frequency shifts to recover the data. FSK is a form of frequency modulation (FM), yielding a radio link that is relatively immune to interference.
The two discrete transmitting frequencies are generated by oscillator transistor Q7 and varactor diode D3. The varactor diode changes capacitance when a control voltage representing the encoded data from uC pin 7 is applied. This capacitance shift “pulls” or changes the tuning of the oscillator slightly to created two discrete frequency outputs when the varactor is OFF or ON. Test Point #3 displays the oscillator output voltage of about .5 volts P-P riding on a 3.5 volt DC step.
The oscillator feeds the output amplifier to generate a strong signal to feed the antenna. Transistors Q4 and Q5 are connected in parallel to boost the output power while Q3 provides a feedback loop to limit the maximum power. The waveform from the output amplifier (Test Point #2) is cleaned up by a lowpass filter composed of L3, C13, L4, and C17. The voltage at TP2 is about 8 volts P-P riding on the +V rail, and after the filter the voltage is about 7 volts P-P.
The antenna is fed by loading network L5 and C18. With the antenna full retracted, the voltage to the antenna is about 16 volts P-P; with the antenna fully extended the voltage drops to about 8 volts P-P.
To conserve power, the transmitter is only turned ON when data is actually generated by a button push or knob turn. Transistor Q6 applies plus voltage to the oscillator and output amplifier whenever uC pin 9 goes low.
Pin 8 of the uC feeds a squarewave signal to a bimorph disc (marked “LS1” on the PCB silkscreen – LoudSpeaker 1?) mounted on the back shell to produce beeps to create user tactile feedback. Note that the black wire is actually connected to the plus side of the battery. Pin 8 is active Low to draw current through the bimorph.
In the picture above, the bulge in the black wire between the hot glue globs is a diode in series with the battery to avoid damage from reversed battery polarity. This diode is not present in all CAB-1 units. If the diode is present, the apparent battery life will be reduced somewhat because of the extra voltage drop across the diode.
The jack at the top end of the case can receive a 2mm mini plug that connects a simple external pushbutton to the uC. When the button is pressed, the uC re-issues the last transmitted action command.
The 4-cell battery pack is connected in series. Looking at the back of the unit, the terminal in the lower left is the minus terminal, and the terminal in the upper right is the plus terminal.
Part 1 A New Kid on the Block
It all started very innocently. In the Fall of 1987 I was attending a monthly meeting of the San Fernando Valley Toy Train Club when one of the members wanted to show off an engine that he was going to race at our annual Cal Stewart train meet in Pasadena. As I recall, his locomotive was a blue Lionel 8769 Gas Turbine switcher that uses a DC can motor. He had added a bridge rectifier to allow it to run on AC power. Although the motor was spinning fast, the low gearing did not provide much speed. He made a couple of test runs and concluded that he was ready to race.
I had never seen the train races before, and I didn’t know what to expect. Our Cal Stewart races are drag races, with two parallel tracks that are slightly longer than ¼ scale mile. At the ¼ mile mark, an insulator is installed to kill power to the racing trains for
coasting to a stop after the speed run. The power is supplied to both tracks by a postwar ZW transformer. Both racing locomotives as aligned at the start mark, and then the AC input power to the ZW is switched on to energize the tracks. The first train to cross the ¼ mile mark wins.
Inspired by the demonstration at the Valley meeting, I decided to try my hand at building a racing entry. I didn’t see much hope for something with small wheels and low gearing, but the rectifier idea was interesting. I dug out an MPC 8902 steamer that most
folks would consider a piece of junk – a loco with a plastic body and a DC can motor. I figured the plastic would be light, and I could soup up the DC voltage source.
My first try was to add a full-wave rectifier so that the train would run on AC. Those DC motors are probably intended for 12 volts, but the ZW puts out about 20 volts AC. After accounting for the voltage drop through the rectifier, I was probably getting 18 volts to the motor.
Testing the racer became a problem because I didn’t have a straight stretch on my layout that came close to being a scale ¼ mile (27.5 feet.) As a result, I never see my racers make a full high-speed run until I get to the Cal Stewart race track.
Applying the old adage “If some is good, more is better!” I thought that I could boost the voltage to the motor by adding some capacitors across the output of the bridge. I proceeded to pack 10,000 microfarads of electrolytic capacitors into every available space under the locomotive’s shell. The capacitors charged up to near the peak voltage of the sinewave power input, adding about 40% more voltage. Now the poor little 12V motor was running on about 25 volts. And run it did indeed! On startup the wheels would spin, throwing off an impressive shower of sparks, and down the track it would fly. Again, I couldn’t run it very far, but things did look promising.
When Cal Stewart came around, I packed my racer into a shoulder bag and headed for the race track. I got there a bit late, and the races were already in progress. The little turbine didn’t stand a chance against the “fast” steamers that were vying for the top spot. The racing was fairly close, but they finally determined who was the fastest.
At this point I stepped up and asked if I could run my train against the winner. When I pulled out my little plastic MPC racer, I could feel the contempt for this piece of junk from the other competitors and maybe it was my imagination, but were they laughing up their sleeves? I put my racer down at the Start line, and the previous winner was placed on the other track. The official started the countdown – “Three, Two, One” – and the power switch was turned On.
The other train began to budge, but my train was so busy spinning its wheels and throwing sparks that it looked like it was just standing still. And then it started to move. Off it went in a big “whoosh”, shooting down the track at speeds nobody (including me) had ever seen before. There was a collective gasp from the crowd and a lot of “Holy $***”s were uttered. My racer crossed the finish line before the competitor was halfway down the track. It was no contest. When my racer reached the power cutoff point, it wasn’t ready to quit. The capacitors were fully charged, and they continued to supply the motor with current on the “dead” cool-down track. Fortunately, someone caught the racer before it shot off the end of the table.
We ran a couple more runs, and I did have some problems with the capacitors vibrating loose and dropping down out of the shell. I was declared the winner, and my prize was a gift certificate for some Gargraves track. It was a great day for me, and train racing at Cal Stewart would never be the same.
Part 2 Upping the Ante
There are four requirements for winning a drag race. In drag racing lingo, the first is a good launch or hole shot. The racer must start moving immediately with full power and not lose time with excessive wheel spin. In automobile drag racing, the initial “reaction time” or time between the green light and initial vehicle movement is measured in hundredths of a second. For train racing there is no human reaction time involved, but there can be delays in ramping up the voltage to the engine’s motor.
On the other hand, wheel spin or loss of traction can be a major factor with trains. Steel wheels on steel track have relatively poor traction. A common enhancement is to add rubber “traction tires” to the drive wheels. If all the drive wheels have traction tires, a supplementary set of wheels (or a sliding contactor) may be required to make electrical contact with the outer rails.
The second requirement is a fast top speed. Getting a good jump off the line won’t win a race if the other vehicle can pass you before reaching the finish line. The ultimate goal is to keep accelerating all the way down the track.
Our trains usually reach a maximum velocity partway down the track. When you manually turn a motor, the motor winding generates a voltage, which is known as “back EMF”, where EMF (ElectroMotive Force) is just a fancy word for voltage. The faster the motor spins, the greater the back EMF. As the EMF ramps up with speed, eventually the back EMF will become nearly equal to the applied voltage from the rails. When this happens, there is only a small difference in voltage across the windings of the motor, and the current through the motor falls off. The current in the armature windings is the true source of the magnetic power that creates torque in a motor, and when the current drops, so does the motor’s power.
Third is clean run without running out of bounds. For automobiles this means staying in your own lane and not weaving around. For our trains that means keeping all of the wheels on the rails. If you look back at the first photograph of this article, you will note that the racing tracks are not perfect straight lines. What appear to be innocent wiggles become serious obstacles to a high speed train. In extreme cases we have resorted to auxiliary guides that keep the racer pointed in the correct direction.
The fourth requirement is reliability. Racing usually puts excessive stress on most of the running gear, leading to premature failure. If the racing engine breaks down before the series of elimination and final races is complete, there will be no trophy today!
I needed help with my hole shot. I had one traction tire and three metal wheels driving the racer. Fortunately, the sparking on the wheels was creating mini craters, which actually worked to enhance traction. The single traction tire was also steering the locomotive to one side, causing derailments. Several races were won with my racer flying down the last part of the racetrack at an angle.
The simplest solution was to add a second traction tire opposite the existing tire. If I owned a metal lathe at the time, that would have been quite simple. As it was, I had to resort to using a Dremel grinder to cut a groove in the driver while the motor was spinning its wheels under power. The workmanship wasn’t beautiful, but the results were adequate. (By the way, the traction tires do need to be replaced periodically. They tend to wear and to craze, leading to thrown or broken tires.)
The second traction tire improved my launch. I still had two metal wheels providing electrical pickup with the outer rails.
Our next racing outing was the TTOS National Convention in Long Beach. Today the races are called Steve Marinkovich Memorial Train Races because Steve supervised the races for a number of years. Steve was a serious competitor who expected to win.
The racetrack is usually available to contestants prior to the race for testing and tuning their entries. Sometimes two competitors will have a mock race to get a feel for the level of competition. Steve and I made test runs, and Steve realized that he wasn’t winning.
The races (under Steve’s supervision) were supposed to start at the posted time, probably 1 p.m. My young boys and I showed up at the racetrack at the appropriate time, but there was no Steve! We waited and waited, and finally Steve showed up. He had gone off and rebuilt his racer with extra features, including traction tires, to try to match my speed. The world had to wait in the interim.
We raced, and indeed his racer was improved, but then his modifications began to fail and fall off. In the end, he lost.
I remember one run very vividly. My boys were assigned the job of catching my racer at the end of the track. Remember the charged up capacitors? Well, one time my boys missed, and the locomotive shot off the end of the track and hit the floor with the motor still cranking out power. The train scooted across the Convention Center floor, flying between the legs of people, with my boys in hot pursuit. They finally caught up with the runaway, but it was about 100 feet from the end of the table by then! I was very relieved. There would be no headlines in the newspaper “Innocent person injured by runaway racing train.” After that we worked out a catcher box with foam rubber cushioning. These trains were getting to be dangerous!
Part 3 More Power, Scotty!
Having helped my hole shot somewhat with the second traction tire, I now turned my attention to the top end speed. More voltage would mean more speed, but we were limited by the output of the ZW, and both racers got the same voltage anyway. One day I realized that my combination of full wave rectifier and filter capacitors could easily be turned into a circuit called a voltage doubler.
Rather that charging all the capacitors simultaneous to one maximum voltage, the doubler charges half the capacitors on the positive peak of the sinewave, and the other half on the negative half. The voltages of the two set of capacitors are added together to give a doubled output voltage.
My next realization was that I could easily have both circuits available. With the flip of a simple switch I could go from regular to doubler and back again!
This was the answer to my quest. I could convert the ZW’s 20 volts AC to a DC voltage of over 35 volts! I liked the idea, but I bet my little 12 volt motor was having nightmares! Once again, I couldn’t give this modification a real test at home. I could run the motor on the workbench, but I was risking blowing up the motor because there was no load or drag to slow it down.
My technique wasn’t the only solution to the top-end speed problem. Another method to increase the maximum speed is to reduce the back EMF by removing a few turns of wire from each of the motor armature’s coils. This technique is commonly used for Marx racing engines. At least one of my competitors was using this technique.
The next round of racing was not very satisfying. The drive wheels had so much power that they would climb off the rails. Several times my engine crossed the finish line sideways at a 30 degree angle.
I tried to keep the racer on the track by adding some weight. I eventually was running with a half-pound chunk of lead duct taped to the top of the shell. Needless to say, adding a half pound of dead weight was not an ideal solution.
Jack Rice, who started racing at Cal Stewart in the early ‘90’s, provided a simple solution. He added a forked guide to the front of his racer that straddled the center rail. This kept the racer pointed in the correct direction. I copied his idea, and that eliminated the lead weight.
Part 4 More Suction, Nurse!
Although I had lots of power and fairly good top speed, I still had a lot of wheelspin at the start. I needed a way to increase the traction between the drive wheels and the track. I knew from previous experience that adding weight wasn’t a good solution. If only I could increase the force of gravity….
I decided to try using magnetism as a substitute for gravity. I purchased some small rare-earth magnets that are very powerful for their size. I soon found that size matters, but even more important is the placement of the magnets. A small magnet very
CAUTION!! Powerful magnets can easily pinch your skin. Use care whenever a magnet is near a steel component or another magnet.
CAUTION!! Rare earth magnets can shatter and forcefully eject debris if they collide together. Wear eye protection.
close to the rail is just as effective as a much larger magnet working over a large gap. On the other hand, if a magnet actually touches a rail, the magnet locks tightly to the rail and acts like a dragging brake.
The magnets need to be very close to the rail, but still provide enough clearance to not hit any bumps in the track or flared ends at rail joints. All three rails can be used, but the center rail is a bit higher than the outer rails because of the insulation inserts that separate the center rail from the ties.
Testing the effectiveness of the magnets was a bit of a challenge. My first attempt was to simply try to lift the engine off a piece of steel track. This told me that I was getting some magnetic “grab”, but I couldn’t compare the effectiveness of various magnet placements.
My more scientific method was to attach a spring scale to a piece of inverted track and calibrate the scale to zero out the weight of the track. I then placed the inverted engine to be tested on the piece of track from below, holding the engine in one hand.
I first checked the weight of the engine on the scale, and then I pulled downward on the engine until I overcame the magnetic attraction. The spring scale readings allowed me to compare the weight of the engine to the amount of added magnetic downforce.
For the test shown above, the track weighed 1 Newton, the engine weighed 5 Newtons (1.1 pounds) and the magnetic force was 8 Newtons (1.8 pounds), more than the weight of the engine. From a traction standpoint, this 1 pound engine would have the traction equivalent to a 3 pound engine. I was picking up almost 2 pounds of downforce using magnets that weighed less than 1 ounce! This was much better than adding lead weights!
Note that the amount of steel in the track is one of the factors determining the amount of magnetic attraction. Lionel tinplate O gauge track has much more steel than O27 track. Nickel silver and stainless steel tracks don’t provide any magnetic attraction. Fortunately, our racing track is classic tinplate O gauge.
My testing on the relatively short straight sections of my layout showed a great improvement in initial acceleration. There was no wheel spin, and the engine looked like it was shot from a cannon. The Cal Stewart races gave the same result, with very impressive runs.
My racing technology had reached a development plateau. My biggest problem now was that the poor little 12 volt motor wasn’t lasting very long. The high RPMs and high surge currents meant replacing the motor every racing season or two, but the big boys at the drag strips have similar problems.
Part 5 My World is Turned Upside-Down
At this point my racing efforts got redirected into a new quest. With the help of Jon Zahornacky, I had developed a “super SC-2” capable of controlling up to 128 switches. My layout currently has 58 switches, and when I added up the cost of enough Lionel SC-2 boxes to do the job, I decided that I should design my own unit. Jon was kind enough to offer to assist me, and we collaborated on the design project to develop a Massive Switch Controller (MSC). The resulting SwitchMaster MSC met all of the design objectives.
I wanted to introduce SwitchMaster at the 200 TTOS/TCA Cal Stewart Train Show in Pasadena, but I felt that I needed something beside the new SwitchMaster hardware to attract traffic to my display table.
I had spent a lot of time with racing trains dangling upside-down for test purposes, but it never crossed my mind that maybe these trains could also run upside-down. Since I didn’t know of any reason why the idea wouldn’t work, I decided to apply power to an inverted track. Much to my awe, the engine was quite content to run forward and reverse while inverted.
I decided to build an inverted display using PVC pipe that would have a straight track on which the inverted engine could shuttle back and forth. The display would be high enough above the table that it would be easily visible to shoppers wandering down my aisle at Cal Stewart.
I added a magnetic reed switch at each end of the track to trigger a reversing circuit that would continuously shuttle the train back and forth.
The display was an outstanding success at Cal Stewart, but my effort was a marketing flop. Nobody was interested in SwitchMaster. They were all captivated by the sight of a train running upside down. I guess I should have taken Marketing 101.
The reactions of the spectators varied for quiet awe to raucous laughter. Everyone had questions, and most people thought that this was some kind of a really high tech concept. I kept my mouth shut that it was just a few magnets in the proper places.
Part 6 Going to the Wall
Like Don Quixote, I was started on a grand (and maybe futile) quest. The TTOS Southwestern Division was having an open house the next spring, and they invited me to display my inverted shuttle. I wasn’t satisfied to rest on my laurels. I started developing a new Novelty Layout that would one-up the shuttle.
I built a simple loop of track on a carpeted base. I began testing with the base on the floor, but I progressively lifted on side of the base up into the air to see what angle of operation I could achieve. I kept lifting and lifting until the base was completely vertical! I had a kinetic wall sculpture with the train sticking out from the wall, running around the layout – straight up one side, across the top, straight down and then across the bottom. It was a bit taxing on the poor little engine, but by now I was using Ready Made Toys BEEPS which had low gearing.
Since people felt that this should be difficult, I decided to humor them. I fastened a vibrator coil from an electric massager to the back of the layout base. The vibrations transferred to the sounding board base as a low hum, sounding very powerful. I also kludged together parts from a computer power supply and a TV picture tube driver to represent a Gravitonics Antigravity Generator. I even added a sign:
Gravitonics Antigravity Generator in Use
May disrupt cardiac pacemakers
The illusion was convincing. I remember one fellow who asked me what would happen if the power to the house failed. I replied “Gee, I didn’t consider that. Let’s see what happens.” I reached over and switched off the vibrator coil…. Nothing came crashing down. I shrugged my shoulders and the fellow went away happy.
A couple of people came up and asked if this was what they had read about in a recent magazine. I assured them there was no magazine coverage on this, but I did get the details of the article. It was an issue of Model Railroader.
On the way home, I stopped off and purchased a copy of the magazine. Sure enough, there was a product review about a new product that would run inverted trains on the ceiling. The product was for HO, and the review was very enthusiastic. I then noticed that it was the April issue – April Fool!! Apparently the folks who mentioned the article hadn’t caught the joke.
I wrote a letter to the editor of Model Railroader. I told him I had two good laughs from the article. The first laugh was regarding the product they reviewed. The second laugh was on them because I had already been publicly demonstrating inverted operation for six months! I didn’t get a reply.
Part 7 You’ve Got to be Flippin’ Crazy!!
The next Cal Stewart was now only 6 months away, and it was time for something new. If I could run trains both upside-down and vertical, what about everything in between? I began construction of a display that would feature not only moving trains, but also moving track. I constructed a rotating rotisserie-like oval that had separate loops of track on the two faces of the oval.
The design had several difficult aspects, including the motor and pivots for the oval, slip rings to feed power to the moving track, and a special train controller.
The job of the controller was to keep the layout balanced so that the load on the rotator motor would be moderate. With engines running on both faces of the loop, it would be possible that both engines would be on the same side of the loop simultaneously. I wanted to keep them on opposite sides to preserve some level of balance. I designed a digital controller that used four magnetic reed switches at the two axis point at the ends of each of the loops. The logic would detect the first train to arrive at the axis point, and it would hold that train until the train running on the other face also arrived at the axis point. It would then release the stopped train to continue around for another half circle, stopping whichever train arrived first at the other end.
Construction was not easy. The frame for the track loops had to match the geometry of the track exactly. The frame for supporting the loop and motor needed to be simple and portable. The controller was best implemented with a custom printed circuit board that required both circuit design and printed circuit board layout and fabrication.
Fortunately, fellow TTOS member Tom Meleck, a very good professional set and production designer, volunteered to help me create a very nice display. He sent me sketches for the display and designs for signage on the display. With his help, I was able to produce a very nice display.
The display was very popular at Cal Stewart. I didn’t have a good site from a foot traffic standpoint, but I was kept busy with visitors.
One visitor said that they had been putting up upside-down Christmas trees at their house for many years. That had all kind of ornaments, including live fish (with aerating tubes from the attic) and hamsters in cages. They had always felt that the tree deserved a train layout, but the layout would need to be inverted on the ceiling to fit in. He asked if I could build a ceiling layout. I said “Yes” and instructed him what kind of train set he needed to buy from the Cal Stewart tables.
An hour later he returned with an appropriate train set in hand. I took the set home and modified the train components to run upside down. I later found out that they did not put up the train that Christmas, and I have since lost track of the fellow.
Part 8 Santa, Am I Seeing Things?
After being challenged to provide a train for an upside-down Christmas tree, I decided to build my own upside-down tree and layout – sans the live fish. I built the train layout on a 4’ x 4’ sheet of ¼” Masonite. I used my old artificial tree, but I had to modify the base of the center pole so that it would not pull apart when the tree was upside down. I suspended the tree and layout with a single ¼” stud screwed into a ceiling joist.
After I had the tree and train up for a while, I called the local newspaper, and they ran a story with photo showing the setup. I eventually packed up the Christmas tree and put it away in the attic.
A few years later in 2009 I was once again trying to figure out what to take to Cal Stewart. I decided to display the upside down tree and train, but I had a serious problem figuring out how to support the display when there is no usable ceiling. (Cal Stewart is in a large convention center that has high ceilings.) I decided to build a gallows-like support that would hold the layout about 8’ off the ground.
I needed a rugged base to support the rigging and tree. I chose the old blue-carpeted “face of the wall” layout from Part 6 above because it had a heavy 2×4 frame and ¾” plywood sheet. As an added bonus, the base provided a second loop of track so that I could run trains above and below.
I added one enhancement over my living room version. I had purchased a rotating base for a Christmas tree a few years before during an after-Christmas close-out sale. I devised a way to mount the base to the layout and to couple the base of the tree to the rotator. The rotator included an internal slip ring assembly that feeds power to the tree lights. I also added some rings to securely hold the “branches” of the tree to the trunk. I also engineered a few extra gadgets that allowed me to transport and assemble the display without any help.
Part 8 Getting Loopy
Back in 2005 I had experimented with the concept of a vertical loop. I formed long pieces of O gauge track into an inside loop that was 5 ½’ in diameter. Bending tinplate track is no easy task, and I had to make special forms to use for the bending. Joining the track segments together with a smooth joint was yet another problem.
I added an outside hoop of ½” conduit pipe for support. Two stationary side posts with V-shaped ball bearing supports held the loop vertical. The bottom of the loop sat on a pulley driven by a gear reduction motor that provided rotation of the loop in either direction.
My initial testing was stymied by the vertical curvature of the track. Most locomotives do not have a lot of clearance at the front and rear. Even my BEEPs were dragging at both ends. I did get the loop to rotate, and I made a few trial runs, but I put the loop aside as needing substantial engineering before it was ready for use.
In 2010 I decided to revisit the vertical loop and use it for Cal Stewart. I fabricated new side supports and mounted the loop on the banquet table that I had previously used for the fancy tumbling display. Since the vertical loop sitting on the banquet table was too tall for the normal ceiling in a house, I had to fold up the legs of the table and substitute some concrete blocks for the legs.
I modified a BEEP chassis by seriously chamfering both ends to provide the required clearance. The BEEP shell would have required even more cutting, but I was unwilling to sacrifice a shell to this project. I decided to run just the chassis, but I added a small stuffed snowman as a passenger.
At first the Cal Stewart folks were only interested in displaying the inverted Christmas tree again. I told them that they would need to provide space for both the tree and the loop if they wanted the tree. As it turned out, there was plenty of space. The two novelty layouts were near the center of the hall, adjacent to a huge conventional Christmas tree that complemented my tree.
My displays got a lot of foot traffic, and they seemed to appeal to all ages of viewers. I think the adults had more of an appreciation for the difficulty involved, but everyone enjoyed the novelty.
Train racing has taken a back seat during most of the Novelty Layout years. For 2011 I hope to concentrate my effort on racing once again. At the current time I don’t have any Novelty Layout concepts hatching in the wings, but all it takes is one stroke of inspiration to start me on my next idea….
You will find videos of some of the Novelty Layouts mentioned above in other post on the Trainfacts.com menu item “Novelty Layouts”. I hope you enjoyed this history and those videos.
This is the original prototype of the MANCO rotisserie layout. In the background on the left is another rotating layout, but this one is built on a sheet of Masonite. When finished, this layout will have a summer scene on one side, and the same scene in snow-covered winter on the other side. The seasons will alternate as the layout rotates.
The blue layout in the background is a wall-mounted vertical layout that is equivalent to a picture hanging on a wall. The train runs on the face of this layout, protruding out into the room. This layout is quite demanding on the train since the train must run straight up and straight down in addition to the two sideways segments.
This is a video of the MANCO 2005 Cal Stewart display. This features both a rotisserie layout and both right-side-up and upside-down back and forth shuttles. The device with the flashing red light at the lower back is the Gravitonics Antigravity Generator. A warning sign states that the Gravitonics Antigravity Generator may disrupt cardiac pacemakers.
The concept drawing for this display, courtesy of TTOS member Tom Meleck is below.
The MANCO Autonomous Layout Mapper (ALaM) provides the capability of generating a 3-dimensional map of an 0-gauge 3-rail train layout. The device is rolled around the layout manually or with a pushing locomotive to sample all of the available trackage. Data from the ALaM is wirelessly transmitted to the host computer. The host computer renders the data into a table of X-Y coordinates describing the entire layout.
Since the ALaM senses actual curvature and slope, there are no restrictions on track radius or grade. Variable curvature such as easements into curves or flex-track custom curves will be accurately plotted.
The ALaM sense the rotations of independent left and right wheels to determine distance traveled and curvature. Each wheel sensor outputs two quadrature pulse streams that can be
First Breadboard of Mapper
decoded to provide distance and direction of each wheel. (A simple 8-pin chip provides the quadrature decoding and can output data as either Up-clock/Down-clock or Up/Down Direction Flag and pulse count.)
The pulse data is read, processed and formatted by a PIC microcontroller. The data processing determines the distance and curvature using the following relationships:
Distance traveled along centerline = (Left pulse total + right pulse total)/2
Curvature from Differential count = left count – right count
This data is formatted as a serial stream and transmitted from the ALaM to the host computer via a wireless link.
The host computer determines the absolute X and Y coordinates of the ALaM using the Distance and Curvature data and plots the data as a drawing of the layout. The coordinate data is also formatted for export as a file compatible with commercially available layout planning programs.
As an additional feature, the ALaM also contains an inclinometer that measures any uphill or downhill track segments. This allows the software to distinguish multiple levels of track and to determine if crossings are at the same level or overhead/under via tunnels and bridges.
A pair of orthogonal accelerometers, mounted off-axis by 45 degrees, detects the difference in gravitational pull angle as the track rises and falls. The accelerometer outputs are lowpass filtered and differenced, then A/D encoded for transmission to the host computer. In the host computer, the tilt angle is integrated with respect to the distance traveled to determine net height change.
1. Conceptualize Autonomous Layout Mapper (ALaM)
2. Breadboard dual-tachometer setup with gates – worked, but too complex
3. Add Inclinometer to concept
4. Build and test inclinometer – good output
5. Simulate data and build Excel spreadsheet that turns data into a layout drawing
6. Design GIZMO mechanics with dual tachometers and bi-directional readouts
7. Construct GIZMO
8. Construct operating chassis with GIZMO, power pickup trucks and circuit board provisions
9. Implement test circuitry with dual bi-directional biphase chips and 4-digit left and right (up-count only) readouts
10. Build new Excel spreadsheet that plots actual data points
July 7, 2006
11. Revise GIZMO design to provide better wheel contact for sensor wheels
12. Design Version 2 circuit board with PIC to provide 4 registers – 2 up- and 2 down-counters and RS232 or similar output and inclinometer with 12-bit A/D input to PIC that connects to computer via tether. Add backup battery/ultracap to sustain memory voltage during track voltage dropouts.
13. Build test circuit board with biphase chips, PIC, inclinometer, serial output
14. Write PIC microcode for registers, A/D and serial output
15. Write (C?) program to automatically plot data using techniques of Excel test in #9. Provide data export in a file that is compatible with track layout planning programs.
16. System testing
1. Count inputs from wheel sensors (two versions possible)
a. Four inputs from two quadrature decoders (http://www.lsicsi.com/pdfs/Data_Sheets/LS7183_LS7184.pdf ) that provides either
1. Up-count and down-count for each wheel, or
2. Up/down direction flag and non-directional pulse count
At 256 ticks per revolutions, a scale speed of 100 MPH would be about 1500 counts/second. Speed could probably be limited to 30 MPH if required (<500 cps)
2. One differential accelerometer analog input
a. Range nominal 0-5V
b. Analog gain determines volts/degree factor
c. +/- 10 degrees for 0-5V would yield .25V/degree
d. A 10-bit A/D would yield 20/1024 = .02 deg/count
e. Signal is heavily lowpass filtered to remove noise. No sample/hold required.
3. RESET input
a. Contact closure to ground resets internal counters to zero after device is positioned at a physical reference point.
4. Serial data output
a. Data payload is 4 bytes of distance, 1 byte of curvature and 1.5 bytes of slope
b. Data rate – at least every 50 counts of distance 10-30 Hz? 80-bit word x 30 Hz = 2400 bits/second
c. Data format – self-synchronizing and self-clocking
1. Accumulate net count for each wheel
Using 12-bit counters would require counter rollover after 55.3”; 16-bit is 73.7’. Longer counters minimize chance for accumulated errors in the host computer’s data after RF transmission?
2. Process wheel data to yield centerline distance and curvature:
a. Distance traveled along centerline = (Left pulse total + right pulse total)/2
b. Curvature from Differential count = left count – right count
3. A/D conversion of inclinometer info – 10-bit minimum With the truncation and noise errors, a 12-bit A/D would have a better chance of actually delivering 10 bits accurately.
4. RESET pushbutton contact closure to ground
The uC could also be used to perform:
1. The quadrature decoding, and
2. The inclinometer signal differencing (requires 2 A/D inputs instead of 1)
a. 5V regulated supply derived from track AC voltage.
Halfwave rectified to maintain ground reference to AC track Common
b. Short-term (30 seconds?) ultracapacitor backup
LEDs in wheel sensors are the major current load – 2 x 10 mA
Test system with 4-digit LED readouts for each wheel
Chassis with power pickup
Closeup of the tachometer disks and sensors
Current Lionel TMCC products are aimed at new TMCC layouts that distribute the ‘intelligence’ around the layout in small bundles. Controllers such as the SC1 and SC2 can be placed near a cluster of switches. The disadvantage of this TMCC-only configuration is that there may be no manual controller/indicator at a central location.
Classic layouts, on the other hand, usually have a central control location that includes all the switch controller/indicators for the entire layout. This configuration permits a quick scan of the indicators and immediate operation on any one switch. This cluster of controls begs for a single TMCC interface that can control many switches by simply tapping into the wiring to the controller/indicators at the central location.
The SwitchMaster is just what the doctor ordered. The modular system consists of a central controller and one to four relay boards that can each control 32 switches.
The central controller utilizes a PIC microcontroller with software developed by The Electric RR Company to decode the serial RS-232 commands from a Lionel Command Base and/or a computer through two separate 9-pin serial port inputs. The PIC can control up to 128 switches (4 relay boards). In addition, an auxiliary EEROM provides enough non-volatile memory to store up to 32 Routes of up to 128 switches each.
To provide maximum reliability, each switch control line is controlled by a heavy-duty relay with 13 amp contacts. The relay provides a contact closure to the common line of the relay circuit, which is usually ground for most switch types. The contact closure is a short pulse, providing sufficient energy to move the switch, but avoiding any excessive currents to a jammed switch or a switch sitting under a parked train. The current SwitchMaster version is not intended to use with continuous-current devices such as Tortoise switch machines.
The interface to the relay board is via two 34-pin flat ribbon connectors per board. The wires from these connectors can be wired to existing terminal strips for the controllers if a convenient connection point exists. In the absence of terminal strips, the wires can be tapped onto the typical 3-wire cables of the controllers as shown below.
Up to four relay boards can be stacked, with interconnections between the boards provided by vertical risers between boards. One jumper on each board programs the relay board to the correct address block of 32 switches.
To avoid any grounding problems, power for the logic and relays is provided by an independent ‘wallwart’ power supply.
The following series of photos illustrates the installation of the first unit. The control panel already utilized flat ribbon cables to all the toggle switch/bicolor LED controller/indicators for 60 switches. These flat ribbon cables were tapped in groups of 32 switches plus 2 ground wires by simply crimping 34-pin headers onto the flat ribbon cables.
Jumpers of 34-wire flat ribbons connect the control panel to the relay boards.
The Command Base and computer share a common connector on the prototype, but this is expanded to 2 separate 9-pin connectors on the production version.