OCCUPANCY DETECTION AND REPORTING

You may have many reasons for installing a system which reports the presence or absence of trains in various sections of track: displaying train locations to operators, progressive cab control, trackside signals, grade crossing gates, sound effects, etc. There are several different techniques available for doing this. If we overlook any particular technique, it is not to give offense.

The next few sections discuss some of the options available for detecting trains, along with their strengths and weaknesses. As you will see, the BD8 Block Occupancy Detector Board successfully overcomes every limitation normally present in occupancy detector systems.

OPTICALLY BASED SYSTEMS

An optically based detector uses a light source and light detector, arranged so that the train physically prevents the light from reaching the detector when it is present. These systems have the benefit of being totally isolated from train running power. They can therefore be used on any layout, no matter how it is wired.

But they also are limited in that the light source and detector have to be positioned so that the train blocks the light. This means that they only detect at a single point. To detect trains throughout a block, many detectors, with overlapping fields of view would be required. Also, depending on the specific design, the visual systems can be sensitive to ambient light or the stability of the voltage powering the system. And empty log cars can be very difficult to detect in any case since they have little cross-sectional area available to block the light beam.

Sometimes people attempt to provide detection for long or serpentine blocks by using detectors at the beginning and end of a block, a technique called gate detection. This can work well if trains always move through the block. But what if a train enters, and then backs out? Or stops completely, say between operating sessions? Or what if the train separates? Or can leave or enter through multiple paths? These are all factors which tend to limit the use optical detectors for signaling detection purposes.

But this doesn't mean that optical detectors have no place in train detection. Use them to take advantage of their real feature: the detection of a train at a single point. Use them to help to spot cars at difficult to see locations, to position hopper cars exactly right for the rotary car dumper to work perfectly, etc.

MAGNETICALLY BASED SYSTEMS

Magnetically operated detectors typically use a reed switch along with a permanent magnet to perform detection. The reed switch is activated by the presence of a magnetic field. Increase the magnetic field by bringing a magnet close and the contacts switch; remove the magnet and the contacts open.

Normally, the reed switch will be buried in the ballast between two ties, and magnets will be attached to the bottoms of engines and perhaps cars. As the train passes, the magnet briefly activates the reed switch. Its contacts can be wired to logic or latching relays to show that a train has passed.

But this is just another version of a gate detection system. All of the comments about optically based detection apply to magnetic detection. Except that optical detection seems to be better at detecting exact position than magnetic detection. And magnetic detection is not at all suited for detection over a 'field of view' of any meaningful size.

SWITCHED ELECTRICAL CONTACTS

Several different techniques have been used over the years to detect trains using switched contacts of one kind or another. Sometimes they are activated by the weight of the train passing over the electrical switch, and sometimes by the wheels physically contacting a wire mounted next to the rails.

Most of these systems are simple and cost effective. The weight operated switches seem to present a significant installation challenge. Getting them set up to operate with a relatively heavy engine (O scale) is fairly simple; getting them adjusted to operate with a relatively light car (HO or N scale) is great way to induce a headache!

But all of these techniques operate in a gate mode, just like the optical or magnetic detectors, only detecting trains at a point. This really isn't occupancy detection, but such switching techniques can serve some useful purposes. And, like all electrical contacts, switch operation can be impaired by dirt and oxidation. Even reed relays, which not sensitive to dirt, should have special metal alloy contacts if they are to function properly in low current logic circuits.

CURRENT DETECTION SYSTEMS

Current detection systems operate by detecting the current which passes through a motor or lights when a train is running in a block. There are many flavors of such systems, and they generally have similar advantages and disadvantages.

A major benefit of current detection systems is that they can easily detect train current, regardless of how long a block may be, or how the track is routed. And, as long as some current is present, they will respond properly to trains that stop, or enter and back out of a block. But they will generally introduce some drop in the voltage which actually reaches the motor. How significant this is depends on the specific technique used.

The next few paragraphs will describe the principle techniques of which we are aware.

RELAY CURRENT SENSING

Specially designed relays have been used to detect the current flowing in a track circuit. In this case, the current passes through the relay coil along with the train motor and lights. Relays have several benefits. First, they are difficult to burn out, and, since the coil is electrically isolated from the switched contacts, properly designed relays can be used with any form of DC or even AC train control. And contacts can easily be designed to switch many amps, and thus control any type of load current.

The main limitation of this technique is the issue of balancing the sensitivity of the relay with the voltage dropped across the coil. As in most areas of engineering, there are conflicting factors which must be considered. With a relay, a magnetic field of a certain strength must be generated to switch the relay. To get sensitivity to low currents, many turns of wire are required in the relay coil. The more turns there are, the higher the resistance of the coil and the larger the voltage drop; using heavier wire reduces the voltage drop but increases the physical size and cost of the relay.

In years past, when motors routinely drew an amp or more of current, a workable balance in the design of the relay could be reached. But today's can motors, which draw only 20% as much current are much more difficult to detect. And when the train stops, so does the current. What is there to detect?

Relay contacts present a further limitation. They are subject to dirt and oxidation which can prevent the contacts from closing properly. Also, when used to switch the very low currents which are involved in logic circuits, typically a milliamp or less, the contacts should be gold plated to resist oxidation. Using gold increases the cost significantly; not using gold decreases the reliability significantly. Another one of those pesky engineering compromise situations!

Considering today's motor technology, relays may not work at all. In fact, we haven't seen or heard of anyone used relays for train detection and also using modern can motors in the engines.

TRANSISTOR CURRENT SENSING

The transistor was first applied to train detection, to our knowledge, by Linn Westcott in the mid 1950's. He designed the Twin-T circuit, and later the Twin-T with booster transistor, for train detection.

His circuits were very effective, and in wide use. They had the advantage of high sensitivity, yet resulted in a low and almost constant voltage drop in the track circuit. They could easily detect milliamp sized currents. By equipping your wheel sets with a resistor of some relatively large value, these circuits could detect a lone car parked in a block - just as the prototype does!

For practical reasons, the layout will usually be wired using common rail wiring, with the Twin-T circuit installed in the path from the common rail to the layout common return. You can avoid the need for common rail wiring by providing a separate, isolated power supply for each detector. This, of course, adds a lot to the cost.

The major limitation of the Twin-T circuit is the fact that the full train running current passes through what is normally a low current path through a transistor. Thus, relatively large and expensive power transistors had to be selected so that the base to emitter path could handle at least three to five amps. And these transistors had to be mounted on a heat sink of some fair size.

The only other limitation of which we are aware is the fact that these circuits operate very quickly, and can respond easily to momentary breaks in current due to dirty rail or wheels. This effect can be reduced somewhat by installing a capacitor in the right place.

This form of detection is still adequate, but it is no longer cost effective to use such high current transistors to carry the train running current.

DIODE CURRENT SENSING

The next form of current detector uses diodes in the common rail path to the block. Very low cost diodes, capable of handling 3 amps or more continuously, while giving a voltage drop of less than 1 volt, are readily available. This has the same effect as the power transistors used in the Twin-T circuit, at a much reduced cost. Sensitivity can be just as high or even higher.

The only thing remaining is to sense the voltage drop across the diode using an operational amplifier or comparator of some kind. Today, such integrated circuits are readily available at low cost. Once the voltage has been detected, it can be conditioned to drive relays, light emitting diodes (LED) or logic circuits.

Techniques of this type result in reduced costs when compared to Twin-T and similar transistor based circuits. They also operate very quickly, and will respond to breaks in current flow due to dirty rail or wheels. As with Twin-T circuits, this can be controlled with a capacitor or other filtering in the right place.

Generally, two diodes are used, connected in parallel with opposite polarity. In this way, a positive voltage is generated when the train current is flowing in one direction, and a negative voltage is generated when the current flows in the opposite direction. This usually requires that the detector board have both positive and negative operating voltages applied, adding cost in the form of an extra power supply.

The BD8 Block Occupancy Detector Board uses an advanced form of diode detection, and a special technique that eliminates the need for dual power supplies to power the board. Special circuitry and signal processing techniques are provided which minimize the sensitivity to wheel and rail dirt.

ISOLATED CURRENT SENSING

There are several forms of isolated current sensing in use. An isolated detector operates much like a relay: a circuit element is wired in series with the block, and an electrically isolated signal is picked off and used to indicate train presence. One such technique uses four high current diodes in series with the track circuit, and two optical isolators. Depending on the direction of current flow, one of the two optical isolators activates.

Since the current sensing diodes are isolated from the rest of the detection circuitry by the optical isolators, this form of detector can be used even on layouts which do not use common rail wiring. The system is fairly sensitive, and appears to have, as its primary weakness, a series voltage drop of about 2 volts because there are two series diodes in the track current path. Noise filtering circuitry must be provided to minimize the effects of wheel and rail dirt. And additional signal conditioning must be provided to allow such a detector to operate LED's, relays or bulbs.