TRACK CIRCUITS, NON-CODED
Fundamentals of D-C Track Circuits .. .. 102
Circuit Adjustment .. .. ..103
Percent Release of Track Relay ... ...104
Track Batteries 104
Relay Resistance .104
Rail Resistance 104
Relay Operation ..105
Train Shunt ..106
Galvanic Action ....107
Storage-Battery Effect .. 108
Half-Wave Rectified Circuit .....108
A-C Track Circuits and Relays .. ...109
Electrified Railroads ...109
Non-Electrified Railroads ...109
Historical Background 109
Vane Type Relays ...110
Principle of Operation, Double-Element Vane Relay ...110
A-C Track Circuits .. ..112
Track Circuit Apparatus Used with A-C Track Circuits 112
A-C Track Circuit with A-C Relay for Non-Electrified Railroad . 113
Double-Rail A-C Track Circuit for D-C Propulsion ...113
Single-Rail A-C Track Circuit for D-C Propulsion 113
A-C Track Circuits for A-C Propulsion ..115
Phase-Selective A-C Track Circuit .115
Magnetic-Stick Relay .. ..115
Principles of Operation .....116
Coded Carrier Track Circuit ...117
Track Circuit Data and Calculations .. ...120
FUNDAMENTALS OF D-C TRACK CIRCUITS
Perhaps no single invention in the history of the development of railway transportation has contributed more toward safety and dispatch in that field than the track circuit. By this invention, simple in itself, the foundation was obtained for the development of practically every one of the intricate systems of railway block signaling in use today wherein the train is, under all conditions, continuously active in maintaining its own protection.
This paragraph is quoted from the Third Annual Report of the Block Signal and Train Control Commission, dated November 22, 1910, thirty-eight years after Dr. William Robinson patented the closed circuit track circuit. The track circuit is still fundamental to most of our signaling systems. There are many different arrangements of the track circuit in use today, but they are alike in their basic principle of operation.
A track circuit in its simplest form is an insulated section of track with a relay on one end and a battery, or some other source of energy, on the other end.
Figure 101 shows an elementary track circuit. It consists of:
1. A source of energy - in this example, a battery.
2. A limiting resistance, so called because it limits the current from the battery.
3. Rails and rail bonding, both offering resistance.
Figure 101. Simplified diagram of a track circuit, unoccupied.
Figure 102. Simplified diagram of a track circuit, occupied.
4. Ties and ballast, both offering a path for current leakage from rail to rail. This path has resistance, referred to as ballast resistance.
5. Relay series resistance (resistance placed in series with the relay).
6. A track relay.
The arrows show the direction of current flow. Starting from the positive post of the battery, current flows through the limiting resistance, the one rail, through the relay winding, the relay series resistance, and back through the other rail to the negative post of the battery. With the relay thus energized, it closes a contact to light the lamp (or to control a signal mechamism to its proceed aspect).
As the wheels and axles of a train move onto the track circuit, Figure 102, they provide a path from rail to rail through which the battery current flows, thus robbing the relay of its current and causing it to open the contact through which energy was feeding to the lamp behind the green roundel and to close the contact to cause the lamp behind the red roundel to light.
Underlying this simple concept are many factors that make the track circuit a complicated problem. Figure 103 shows the track circuitas a network of resistances: rail resistance, ballast resistance, limiting resistance, and relay resistance. In an average track circuit, the resistance of the rails may vary from 0.015 to 0.05 ohm per 1000 feet of track. The ballast resistance may vary from one to hundreds of ohms per 1000 feet of track. A 4000-foot circuit may have, for example, a total rail resistance of 0.1 ohm and a total minimum ballast
Figure 103. A track circuit is a network of resistances.
resistance of 0.25 ohm in wet weather. In dry weather, or in zero weather, this ballast resistance may increase to hundreds of ohms.
The maximum distance over which a track circuit can operate properly is dependent on several factors, principally on ballast leakage. Methods of lengthening the detecting area are described in Section 200, Block Signaling, Single-Direction Running.
Because there is a limiting resistance in the battery feed to the track, the track voltage and, in turn, the relay voltage vary with whatever leakage path exists across the rails. This leakage path consists of the various conducting paths through the ballast and ties and, when the track is occupied, it also consists of the wheels and axles of the train. Thus we have a widely varying resistance across the rails and a widely varying voltage across the relay. A familiar analogy is the varying water pressure evident at the upstairs shower when someone is opening and closing faucets downstairs.
Figure 104, shows how relay current varies as the resistance across the rails is varied. To simplify our example, we have chosen a relay that picks up at 100 milliamperes and drops away at 50 milliamperes. By picks up we mean the relay is sufficiently energized to attract its armature and close its contacts with full pressure. By drops away we mean the relay releases its armature so that full pressure is exerted on contacts that are normally closed when the relay is not energized. These are also spoken of as front (closed in relay energized position) and back (closed in relay de-energized position) contacts. In signalmans language we would say this relay had 100 mills pickup and 50 mills drop away - a 50 percent release relay.
Looking at curve A in Figure 104, we see that as we move from a shunt of zero resistance across the rails to a leakage of 0.25 ohm, the relay current increases to 100 mills, pickup of this relay. In other words, this circuit will not operate unless the ballast resistance is 0.25 ohm or more. Now, moving backward along curve A toward zero resistance, we find the current of this 50 percent release relay below 50 mills at 0.1 ohm. This is well above the 0.06-ohm shunt required by the Federal Railroad Administration to ensure the proper sensitivity of track-circuit shunting.
This picture is varied by many things: percent release of the track relay, relay resistance change with temperature, off and on charge voltage of the track battery, type of battery, rail resistance changes with temperature etc.
Figure 104. Relay current variation with varying resistance across the rails.
Obviously, however, the end to be obtained is not to get the best shunt with the highest allowable resistance but to get the best shunt when a train enters the track circuit.
Curve A in Figure 104, shows the conditions when we have a 4-ohm relay without added relay series resistance. Curve B shows the conditions when we add 4 ohms of series resistance to our 4-ohm relay. Incidentally, we have to reduce the limiting resistance in the track battery feed to do this, otherwise our relay will not pick up on the given 0.25-ohm minimum ballast.
In comparing the two curves we note:
1. Curve A shows relay current below drop- away at 0.1 ohm, while curve B is still well above dropaway. In fact, on curve B the dropaway point is barely reached at the AAR 0.06-ohm shunt. The first reaction is that curve B is not as good as curve A, that our shunt resistance must be lower. Note now, for future reference, that when we added the 4 ohms in series with the 4-ohm relay, we doubled the interrail voltage at the release value of the relay - a point to remember when we consider ionization voltage as will be discussed under Train Shunt.
2. At infinity ballast resistance, the relay current in curve B rises to only 0.3 ampere as against 0.55 ampere for curve A. This lower current at dry ballast gives faster shunting as we shall note under Relay Operation.
Percent Release of Track Relay
Suppose we substitute a track relay with a percent release approaching 100, that is, we move the relay dropaway line up close to the relay pickup line. Now our circuit will release with a quite high resistance shunt, just a little less than 0.25 ohm. The answer to a signalmans prayer? Not quite. It is possible to build a very high release relay but it requires special adjustments, special core material. Such a high percent release relay will not maintain. its operating values consistently over a period of years. We expect a relay to stay in service for many, many years with little or no attention. By working to a 65 percent release, we can insure maximum stability in service.
Another factor affecting the choice of relay release percentage is what we might call the sub- low minimum ballast condition of a track circuit. This condition usually occurs at the start of a shower after a long dry period. The ballast particles are coated with fine dust. At the first wetting, the dust particles are merged to form innumerable rail- to-rail paths of unusually low resistance. As the rain continues, this material washes away, and the rail-to-rail resistance rises again. A low percent release relay will be far more likely to remain picked up through such a sub-low ballast condition. Curves in Figure 104 show that, with a 65 percent release relay, the ballast could drop much lower than 0.25 ohm, that is to 0.150 ohm, and not cause the relay to drop away and suddenly place a signal at stop in the face of an approaching train without prior warning. If we had a high percent release relay, we should have to adjust the circuit for that much lower ballast resistance. The result would be an approach to the same resistance shunt, and no advantage would be gained.
One other natural phenomenon also works against the high release relay. At nominal low ballast conditions and when there have been no trains over the circuit for some time, it has been noted that ballast resistance tends to drop markedly and may even reduce the relay current below its pickup value. Under this condition, if we should momentarily shunt the relay, it would not pick up again. Fortunately, the favorable part of this phenomenon is that when a train does run over such a circuit it disturbs the rail and ballast contact and increases the ballast resistance so that the relay picks up readily when the train gets off the circuit. Here again, a high percent release relay would be of no benefit.
There are several types of cells used in batteries for d-c track circuits: lead acid secondary cells, nickel cadmium secondary cells, and primary cells. Each of these differs in its nominal voltage and in its high and low voltage points. Temperature also affects cell voltages. (These characteristics are further discussed in Section 1500, Power Supply.)
As shown in Figure 104, obviously we must adjust our track circuit so that the relay will pick up at one quarter ohm ballast when our track cells are at their minimum voltage. When the cells are at their higher voltage, the circuit will operate at a lower ballast - and the shunt will have to be lower to cause the relay to drop away.
The resistance of the track relay will increase with the temperature. We must allow for plus or minus 20 percent in relay resistance as the temperature inside the relay housing varies from 150 degrees above to 30 degrees below zero. Thus when we adjusted our 4-ohm relay to operate at one quarter ohm minimum ballast resistance, we must bear in mind that it could become 4.8 ohms, with possible failure to pick up; or, in cold weather, it could become a 3.2-ohm relay, and the shunt resistance would have to be relatively lower.
On a hot day, the rails are expanded; their ends are forced together with great pressure. In zero weather, the rails contract; they pull hard against the splice-plate bolts. With either of these two conditions, rail resistance is nearly like that of solid rail, approximately 0.015 ohm per 1000 feet of track, depending, of course, on the size of the rail. In moderate temperature, the joints are not so tight. The bonds must carry the current.
Even with relatively good bonding, rail resistance of 0.03 ohm per 1000 feet of track is not unusual. Here again we must make allowances, and shunting may vary substantially. Continuous welded rail minimizes such resistance changes.
A relay, as outlined in Figure 105, is an electromagnet. Contacts attached to (and insulated from) the hinged armature make contact with the back contacts when the relay is de-energized and with the front contacts when energized.
Since the armature is hinged so that it may open or close the magnetic circuit, the flux within the magnetic structure of the relay will vary with the armature position as well as with the current through the coils. Figure 106 shows how the flux density is varied by these two factors, current and armature air gap. Figure 106 is not an exact representation of the values of any particular relay. It is simply offered as a general example of this phase of relay operation.
Note that as we increase the current through the relay to 0.065 ampere, the magnetic lines of force (flux) increase to 11,000. At this point the armature is attracted to the pole pieces (the relay picks up). This reduces the reluctance of the magnetic path by reducing the air gap, and the flux increases to 22,000 lines with no change in current. As we further increase the current to 0.3 ampere, we reach the saturation point, that is, a point where further increase of current will no longer increase the flux significantly. Here the flux has increased to 38,000 lines.
Figure 105. Diagram of relay showing how flux is varied by armature position.
Now, as we reduce the current from saturation to 0.065 ampere, curve B, we note that the flux does not decrease proportionately. We must reduce the current to 0.054 ampere to reach the, same 22,000 lines. This difference of current is caused by residual magnetism in the cores. To get the relay to drop away, we must further reduce the current, and in turn the flux, to 17,000 lines. This difference of 5000 lines is due to the contact friction and contact spring load curve not agreeing with torque exerted by the magnetism.
It is not our purpose here to go into more details of a relay than are necessary to show the percent release of the relay (ratio of pickup to drop- away, etc.) as these factors affect the track circuit.
We all know that when current increases in a conductor the expanding flux from that conductor in passing through a parallel conductor generates a counter or opposing voltage in the parallel conductor. This generated voltage is proportional to the rate at which the flux passes through the conductor. In a relay winding where all the turns are parallel, the expanding flux of every turn passing through all the other turns generates a counter voltage in each. Therefore, when we apply voltage to a track relay, as when a train moves off a circuit, the current and flux do not increase instantly. They increase at a rate that generates a counter voltage equal to the applied voltage. This slow pickup is desirable. On the other hand, when a train moves on to a circuit and reduces the voltage, tending to reduce the current and flux, the flux from each turn recedes back through all the other turns at a rate that will just generate a counter voltage that tends to maintain the current and, in turn, the flux. Therefore, the relay is slow in dropping away - which is not desirable.
Figure 106. Curves showing how flux varies with both current and armature position.
If the relay current is just at pickup (ballast resistance at its minimum), the relay shunt is quite fast, for the flux need only recede from 22,000 lines to 17,000 lines. If the relay current were at saturation, with 38,000 lines, it would take considerably more time to drop to 17,000 lines.
This condition is undesirable for a short train moving from one track circuit to another. The track relay of the first circuit could pick up before the other dropped away, thus releasing route locking, for example, in an interlocking.
To avoid this condition, we put resistance in series with the relay. If we put 4 ohms in series with a 4-ohm relay, the flux will drop twice as fast because it must drop fast enough to maintain the current through 8 ohms in place of 4 ohms; and if we inserted 12 ohms it would drop 4 times as fast. Furthermore, adding this series resistance affects the circuit so that the relay current does not rise as high at dry ballast and therefore drops more quickly to the release point of the relay. However, if we refer to Figure 104, we note that when we add relay series resistance, we change the shape of the curve as from A to B and find that the ohmic shunt is poorer.
Figure 107 shows a slide type resistor, arranged as shown, with a battery and an ammeter of such low resistance that its effect on the circuit may be
Figure 107. Curves showing how effectiveness of shunt varies with its position in the circuit.
ignored. As shuttle X is moved from one end of the resistor to the other, the current curves plotted from the ammeter readings will be symmetrical on either side of center as shown. Repeating this procedure, using values of X from 5 to 0.06 ohm, will produce the curves shown, each labeled with the X value used.
Now let us make this slide resistor represent an unbalanced circuit, that is, a circuit where the relay resistance is much higher than the limiting resistance. Let section A, 1/2 ohm, represent the limiting resistance; section B, 1/2 ohm, the rail resistance; section C, 4 ohms, the relay; and shuttle X the ballast and shunt resistance. With this arrangement, we find the following:
1. With the travel of shuttle X restricted to section B, shuttle X (the shunt) is twice as effective on the relay end than it is on the feed end of the circuit.
2. This variation of the effectiveness of the shunt varies with the location of the rail resistance (section B) in the rest of the circuit resistance.
3. If the limiting resistance (section A) equalled the relay resistance (section C), the rail resistance (section B) could be of the same value, and yet you would have the same shunt effectiveness all along the rails.
Where two conductors are in contact, a certain voltage, regardless of how small, is required to cause current to flow across the boundary. We have this contact boundary in train shunting, and in the majority of cases the voltage required is negligible. However, in many cases the rail is covered with dirt rolled into the surface or with rust from lack of use. Depending on such insulating or semi-insulating materials as may be between wheels and rails, more or less voltage will be required to break down the film so to speak and to establish a current path. This is termed the ionization voltage. The action is comparable to what occurs in a zener diode.
When sufficient current is established to maintain ionization, a further attempt to raise the voltage is futile, as the ionized path simply passes more current, and the added voltage is dissipated in the series resistance. In other words, with a given weight of train moving on a circuit having rails rusted to a given degree, the voltage across the rails will be of a certain value, regardless of the limiting resistance or battery voltage (assuming the voltage is at least at the ionization level).
Unusually high ionization voltage is also found in rails where no traction or braking is applied. For example, there may be a station where all trains stop by simply shutting off the power, the grade being a substitute for applying the brakes. Where this occurs, the rail surfaces tend to become glazed, as there is neither traction nor braking friction to keep them clean.
Figure 108 shows the test circuit used and the results of tests made on a side track having very rusty rails due to infrequent use. The circuit was very short and the weather dry so that there was practically no ballast leakage. A potentiometer was used so that the voltage applied to the rails could be increased from zero to whatever was needed. One truck of a car was pushed out onto the circuit, then the voltage across the rails was increased until it dropped back to practically zero.
The curves show the result. The dots on the ends of the curves show where the current flow was great enough to decompose the rust, leaving a metal-to-metal contact with practically zero resistance and a voltage and current proportional to
Figure 108. Test circuit and graph showing ionization voltage for test conducted on rusty rail.
the potentiometer setting. Inspection with a magnifying glass after the test showed bright metal spots at the contact areas.
After each test, the car was moved a few inches to a new spot. The irregularity of the wheels and the rail, and the degree of rust, account for the variation in the curves.
The most trouble is found in little-used connections to sidings and between switches in crossovers. In such cases, we have found ionization to require more than 0.7 volt. An interrail voltage of 0.7 would put 175 mills through a 4-ohm track relay, a relay that picks up at around 65 mills. Obviously we would have to add resistance in series with the relay, at least 12 ohms, to ensure relay dropaway with this interrail voltage.
At interlockings where the track circuits are short, usually enough relay series resistance can be added. However, in some cases an additional cell of battery must also be added so that required limiting resistance can be maintained. In many cases of side track connections in long track circuits, enough relay series resistance cannot be added, and therefore a separate track circuit must be installed.
Under some conditions of atmospheric pollution, contaminants may be deposited in the ballast. Then, when one rail has better contact with the ballast than the other rail, there is a possibility of galvanic action - of the rails acting like electrodes of a chemical cell, producing a track voltage independently of the track battery.
The rail making better contact with the ballast is positive, and the rail making poorer contact, as through the ties, is negative. One theory is that the ions in the ballast attack both rails, carrying the iron into the ballast and in doing so leave their extra electrons on the rails, tending to make both negative. The rail that is in contact with the ballast does not become as negative because the electrons can flow back through the ballast contact. The other rail, where this ballast contact is not present, becomes more negative, and the electrons will flow through the relay across the rails to get back to the ballast.
A ballast-resistance network with a rail broken is shown by Figure 109. The battery end of the broken rail to ground and on to the other rail forms a voltage divider, with the relay end of the broken rail to ground in series with the relay. At near minimum ballast resistance, there is little voltage to spare, and the relay will drop. On the other hand, at near infinity ballast resistance, the resistance of
Figure 109. Possible leakage paths with broken rail (top) and with broken rail and galvanic action (bottom).
the relay end of the broken rail to ground is too high to allow the relay to operate. However, there is a critical range of ballast resistance where the track voltage can affect the relay if the circuit is not properly proportioned or adjusted.
The track battery charges the rails, oxidizes the positive rail, and deoxidizes the negative rail (analogy is found in the iron negative plate of a nickel-iron battery); and this charge may operate the relay after the rail is broken. The track battery polarity also affects the result, as it either adds to the galvanic and storage-battery action or, if reversed, it tends to neutralize the galvanic action. However, with a train on the battery end, the battery is eliminated, as shown in lower diagram, Figure 109.
Half-Wave Rectified Circuit
In this arrangement, Figure 110, the track circuit is fed from an a-c source through a step-down transformer and a limiting resistor. At the relay end, a half-wave rectifier converts the a.c. to pulsating d.c. for the d-c track relay. Note that the rectifier can be installed at the feed end if desired, but then must have capacity to carry the current which flows when the track is shunted.
Figure 110. Half-wave rectified track circuit.
Advantages of the half-wave rectified circuit are higher voltage for ionization of rail film, and less variation in relay current with changing ballast resistance.
Figure 111 shows how the track voltage peaks during the half-cycles which feed the relay. The peaks are 2.82 times higher than the rail voltage for a d-c track circuit with the same average voltage. This provides more reliable ionization of the wheel- to-rail contact.
Average current through the relay depends on two factors. During the on half-cycles, relay current increases with increasing ballast resistance since inter-rail voltage increases. During off half-cycles, decay of the magnetic field in the relay coil maintains relay current via external paths, which include the ballast. When ballast resistance increases this current decreases, thus tending to offset the increase of relay current during the on half-cycle.
IONIZATION VOLTAGE OF ROLLING SHUNT
Figure 111. Wave peaks and ionization voltage with half- wave track circuit.
Figure 112 compares the regulation of d-c and half-wave rectified track circuits. As ballast resistance increases, relay current for the d-c circuit (dotted line) increases continuously toward a maximum. In the half-wave circuit, relay current (solid
Figure 112. Regulation in half-wave rectified track circuit.
line) actually decreases at high ballast and is never higher than for the d-c circuit. The half-wave circuit can therefore detect a higher resistance track shunt at high ballast, while providing equivalent sensitivity at low ballast.
If ballast resistance approaches infinity, the off half-cycle current in the relay might decrease so much that average relay current would drop below relay pick-up. To pervent this, a bleeder resistor is connected across the relay. This ensures that there is always a shunt across the relay coil to maintain average current above relay pick-up, even at infinity ballast.
A-C TRACK CIRCUITS AND RELAYS
In signal terminology, a-c track circuits are energized at power (relatively low) frequencies, such as 60 Hz or 100 Hz. These circuits are very similar in arrangement to d-c track circuits, although they use a-c operated relays, utilize transformers where batteries would be used in d-c circuits, and often use reactors where resistors would be used in d-c circuits.
High-frequency (also called audio-frequency) track circuits are a more recent development. These also utilize a-c energy, but at frequencies ranging from about 600 Hz to over 10 kHz. Equipment for high-frequency circuits is largely electronic, and their characteristics differ substantially from d-c and a-c track circuits. High-frequency track circuits are discussed in Section 400, Track Circuits, Coded and Electronic.
On rapid transit systems and electrified railroads the rails typically serve as return conductors for propulsion current, which may be d.c. or a.c. Track circuits on such roads must function in the presence of propulsion currents which can be hundreds of times greater than the track-circuit current flowing in the same rail. D-c track circuits cannot be relied upon to operate properly in this environment.
A-c track circuits and relays were developed essentially concurrently with the growth of electric propulsion. By utilizing induction effects associated with a.c. but not d.c., a-c track relays were immune to propulsion d.c. By using track circuit energy of a frequency different from that of propulsion a.c., a-c track circuits were developed which operated reliably in the presence of the propulsion a.c.
Early track circuits for use with a-c propulsion utilized a-c track relays which were inherently responsive to track-circuit energy, but which rejected propulsion energy. More recently, a-c track circuits have been developed using circuit arrangements external to the relay to discriminate between track circuit a-c energy and propulsion a-c energy. This permits rectification of a.c. for use with a d-c track relay.
On occasion, a-c track circuits are used on non- electrified railroads. For example, if the track is subject to foreign current, as from a parallel rapid- transit system, a-c track circuits can be used to obtain reliable track circuit operation. Should high- voltage d-c power transmission become widely used, stray currents from the power system might affect d-c track circuits sufficiently to require the use of a-c track circuits.
A-c track circuits are also used for specialized purposes as in classification yards. In such applications the reason for use of a.c. may be the improved track shunting provided by the high peak voltage of the a-c wave, or the economy of using commercial power as the source of track energy. Ability to reject foreign current is not usually a consideration. Track a.c. is converted directly to d.c. by a rectifier and used to operate a d-c relay.
The first electrified railroads employed d-c propulsion. A-c track circuits using a single-element a-c vane relay were developed for such roads. Operation of this relay was based on the reaction of eddy currents in a non-magnetic vane, and currents induced in copper ferrules (shading coils) on the poles of an a-c electromagnet. D.c. would not produce induced currents, hence the relay was immune to d-c propulsion effects.
The single-element vane relay needed appreciable power for operation, all ol which was obtained from the track. Since losses caused by ballast leakage could be high, track circuits were typically restricted to maximum lengths of 1000 feet or less, to keep input power to the track at practical levels.
Subsequently, two-element a-c track relays were introduced. These took advantage of the fact that relay operation could be obtained by energizing the relay from two sources: the track and a local (line) a-c supply. Such relays required much less power from the track. Local a.c., from a signal-frequency line, could furnish the bulk of the energy required by the relay. The corresponding reduction in track energy requirements made a-c track circuits as long as 5000 feet practical. Failure of the local supply did not impair safety, since the relay could not operate unless both elements were energized.
One early design of two-element a-c relay employed the principle of the induction motor for operation. A rotor was located in a magnetic field generated by the two windings. When the track and local windings were properly phased the magnetic field rotated, causing the rotor to turn in the same direction. The rotor drove a pinion gear and sector, which in turn drove a linkage to operate the contacts. At the end of travel the rotor stalled, the stalled torque holding the contacts closed. When the track was shunted, the magnetic field of the track element disappeared. Since the magnetic field of the local element did not rotate there was no torque, and a counterweight reversed the rotor, opening the contacts.
Single-element and two-element a-c relays of the types described could respond to a relatively wide range of frequencies. When 25-Hz a-c propulsion was introduced it was necessary to prevent track relay response to propulsion a.c. To achieve this, frequency-selective a-c relays were developed. Such relays would operate on 60-Hz energy but not on 25-Hz. Typical of these were the static-frequency and centrifugal-frequency relays.
The static-frequency relay was a rotor-type single-element a-c relay, generally similar in construction to the two-element rotor relay previously described. Since all power was obtained from the track circuit, it was suitable only for relatively short
Figure 113. Vane relay, GRS Type B Size 2.
track circuits. A frequency-responsive bridge, included in the relay, phase-shifted a portion of the track energy to produce the out-of-phase component necessary for a rotating magnetic field. The required amount of phase shift occurred with 60-Hz input but not with 25-Hz, hence propulsion current was rejected.
The centrifugal-frequency relay was a two- element frequency-responsive relay which could be used on longer track circuits than was practical with the static-frequency relay. Advantage was taken of the fact that an induction motor rotates at a speed proportional to the frequency of the driving current. In this relay the rotor rotated freely to actuate a centrifugal linkage. When the rotational speed was high enough, the centrifugal mechanism operated the relay contacts. Required speed was achieved only if the relay was driven by 60-Hz energy. When the track circuit was shunted, the rotor stopped and the contacts opened. The contacts were also open if the rotor ran at a speed lower than the 60-Hz speed.
Vane Type Relay
The modern relay for a-c track circuits where there is d-c propulsion is the GRS Type B2 two-element vane relay, Figure 113. Since it is a two-element relay, it has high efficiency, permitting relatively long track circuits. A laminated magnetic structure minimizes hysteresis and eddy current losses. An aluminum vane, which rotates through a limited arc in a vertical plane, operates the contacts through a crank linkage. The vane assembly includes a counterweight which provides gravity- actuated return to the contact-open position when either relay element is de-energized.
Principle of Operation, Double-Element Vane Relay
The conditions which produce operation of a vane relay are not as readily visualized as those of a d-c relay. The following simplified description covers the underlying principles.
Figure 114 demonstrates the physical effect on which operation of the vane relay is based. A circular copper disc with a shaft through the center is supported so that it is free to rotate. If a permanent magnet is moved in a circular path around the disc shaft in such a way that the magnetic lines of force cut through and across the disc, the disc will start to rotate, at a slightly slower speed and in the same direction, as the permanent magnet.
Figure 115 is an enlarged view of a section of the disc showing what takes place. When the permanent magnet is rotated, its magnetic lines of
Figure 114. Theory of operation of vane relay.
Figure 115. Eddy currents in disc and resultant magnetic forces.
force move through the disc and produce voltage between points on the disc. Eddy currents circulate between these points.
The eddy currents in turn produce magnetic lines of force which react with the field of the permanent magnet to drag the disc in the direction of magnet movement. Figure 115 shows the disc in a horizontal plane; obviously the effect would be the same if the disc and magnet motion were in the vertical plane.
The two-element vane relay produces an effect similar to the moving permanent magnet by means of two fixed electromagnetic structures, one on either side of a vertically mounted vane.
Figure 116 shows in simplified form how the two a-c electromagnets interact to produce the equivalent of a moving magnetic field. The wave forms, bottom of the illustration, show the currents flowing through the local winding and the track winding, with the latter lagging by one quarter of the total cycle or 90 degrees. Reference lines A, B, C, D, A1 indicate selected instants in a cycle. A1 corresponds to A, but for the next cycle. The diagrams above the wave forms show current flow in the two sets of coils at those instants. If we concentrate on one particular line of flux in the magnetic structures, shown by the dotted line, we see
Figure 116. Theory of operation of double-element vane relay.
that its path shifts as the magnetizing effect of the relay windings changes with the coil currents. Where the flux line crosses the air gap and disc, N and S magnetic poles are indicated. It will be seen that the N-S configuration associated with the flux line moves from left to right across the pole faces of the electromagnets. The result is equivalent to that of moving a permanent magnet: the vane follows the moving magnetic poles. The action is of course more complex, since there are innumerable lines and a multitude of equivalent moving magnets. But all move in the same direction, thus all tend to drag the vane the same way.
If we reverse the connections to either the local coil or track coil, the phase relationship is shifted one-half cycle, and the local current would then lag the track by one-quarter cycle. If a diagram similar to Figure 116 were drawn for this arrangement, the instantaneous flux patterns in the relay would be in the sequence DCBA, the mirror image of ABCD. The vane would then be dragged from right to left.
The same principle applies in a vane relay. If the local and track windings are not properly phased, the vane tends to move in the direction of the back contacts. By energizing adjacent track circuits with a.c. of opposing instantaneous polarity, this characteristic of the two-element vane relay provides protection against improper operation from breakdown of insulated joints.
A-C TRACK CIRCUITS
Three general types of a-c track circuits are in use in North America.
1. Track circuits in which a.c., either steady energy or coded, is rectified to operate a d-c relay. These may be regarded as hybrid d-c circuits and do not require additional discussion.
2. A-c track circuits using a-c track relays.
3. A-c track circuits using phase-selective and coded carrier methods of assuring immunity from propulsion current. These use d-c track relays. Track circuits of this type are used with a-c propulsion since the vane relay is not frequency selective.
Energy used in a-c track circuits is typically 60-Hz for non-electrified railroads, and where d-c or 25-Hz propulsion a.c. is used. With 50-Hz propulsion (used outside the United States), 83-1/3 Hz is a typical track circuit frequency. For 60-Hz propulsion, a-c track circuits commonly use 100-Hz energy. In general, the frequency for a-c track cirCu its is selected to be distinct from the fundamental and harmonics of the propulsion frequency.
Track Circuit Apparatus Used with A-C Track Circuits
A-c track circuits are energized by track transformers, usually with a primary rated for a standard commercial input, such as 120 volts, 60 Hz. Secondary windings are tapped to permit a choice of track voltage. Typical is the GRS Type UT transformer, Figure 1509 (Section 1500), designed for flexibility of application with two independent 120-volt primaries tapped (20-30-60-10) and two independent 18-volt secondaries tapped (3-3-3-3-4-1-1). Each secondary can therefore provide a range of 18 volts in 1-volt increments. The flexibility available makes it possible to use the same transformer for auxiliary loads, such as signal lamps, while simultaneously feeding the track circuit. For track transformers designed to AAR specifications, volt-ampere ratings at 60 Hz are fully applicable up to 100 Hz and with only slight reduction down to 50 Hz.
Adjustable impedances (reactors or resistors) are used between track transformers and track. They are used to adjust the current to the track circuit, to limit the current when the track circuit is shunted by a train, and to aid in obtaining favorable phase relation between currents in the track and local elements. Non-adjustable reactors or resistors may be used in series with the local element of two-element track relays to improve the phase relation between currents in the local and track phases of the relay. For simplicity, illustrative track circuit diagrams accompanying this discussion do not show all arrangements actually used.
A balancing impedance, an impedance of special design, is used across the track phase of two-element track relays on single-rail track circuits on d-c electrified railroads in order to prevent magnetization of the relay magnetic circuit if any d.c. should enter the relay track winding. Application of the balancing impedance is discussed later in this section.
Fuses or Fusetrons (time-delay fuses) are used between track and transformer secondary winding and track relay track element on single-rail track circuits to protect the apparatus if excessive direct current should continue to flow.
Impedance (reactor) bonds are large reactors connected between the rails at each end of doublerail track circuits on both d-c electrified and a-c electrified railroads. Their purpose is to permit flow of the propulsion current from one track circult to the one adjoining while at the same time retaining the a.c. for the operation of the track relay within the limits of its particular track circuit.
The size of impedance bonds is determined by the amount of propulsion return current the windings must conduct continuously, or for certain specified periods, from each rail of the track circuit. The windings are made from copper strap or large cross-section wire, depending on propulsion current levels. Referring to Figure 118, the propulsion current flows from each rail into the winding and out through the midpoint connection to the midpoint of the adjoining bond and hence out in equal amount to each rail of the adjoining track circuit, continuing in like manner through successive track circuits.
If the propulsion current is the same in each rail, the equal propulsion currents which flow in opposite directions through the two halves of the bonds windings neutralize each other, and no magnetic flux is produced in the magnetic circuit of the bond. If there is a difference between the propulsion currents in the two rails, the unbalanced current in one half of the bond will produce a flux in the magnetic circuit, causing some reduction in the impedance of the bond. The decrease will depend on the amount of unbalanced propulsion current.
Since an impedance bond is connected between the rails at each end of the track circuit, some track-circuit a.c. will flow through it from rail to rail. Therefore, this current will be drawn from the track transformer in addition to that required for the track element of the track relay and the current drain through the track ballast. Hence the impedance of the impedance bonds is made as high as the conditions under which they are used will permit.
Propulsion voltages are typically higher in a-c than d-c propulsion systems. For given propulsion power, higher propulsion voltages require less current. Thus impedance bonds for a-c propulsion are usually smaller than those for d-c propulsion. With typical d-c propulsion systems fed at about 600 volts, rail currents on the order of 1000 amperes per rail may be expected. A-c propulsion systems may operate at much higher voltages: 11 kilovolts, 25 kilovolts, even 50 kilovolts. Rail current in such systems would not normally exceed a few hundred amperes.
Unbalanced return currents in the two rails is not as likely to cause magnetic saturation in an a-c bond as a similar condition in a d-c bond. Excess a-c current through one half of the bond winding produces an autotransformer effect which tends to increase the current in the other winding, thus restoring balance. This self-balancing tendency permits the use of more efficient core design, further reducing the size and weight of a-c impedance bonds.
A-C Track Circuit with A-C Relay for Non-Electrified Railroad
Figure 117 shows a basic a-c track circuit with a two-element a-c relay for use on a non-electrified railroad. The similarity to a conventional d-c track circuit is obvious. Adjustable reactors could be substituted for the tapped and adjustable resistors shown in the transformer and relay connections to the rails. The tapped resistor serves not only as a current limiter but also is used to adjust the phase of the track energy for the most favorable relationship between the track and local currents in the two-element track relay.
Double-Rail A-C Track Circuit for. D-C Propulsion
Figure 118, shows a typical double-rail a-c track circuit for use with d-c propulsion. The tapped resistor is both a current limiting device and a means for obtaining the most favorable phase relationship between track and local currents in the two-element track relay.
Single-Rail A-C Track Circuit for D-C Propulsion
This a-c track circuit, Figure 119, utilizes one continuously bonded rail for propulsion current return with the other rail divided into track circuits by insulated joints. Examination of the circuit shows that the signal rail is not actually isolated from propulsion currents, since the circuit made up of the track transformer secondary with its track leads, the signal rail, and the relay with its track leads, is in parallel with a section of the power rail. When d-c propulsion currents flow in the power rail, voltage differentials develop which cause d.c. to flow in the parallel signal-rail circuit. The magnitude of this d-c voltage and the resultant d.c. in
Figure 117. Track circuit with 2-element, a-c relay for non-electrified operation.
Figure 118. Track circuit with a-c track relay and impedance bonds for d-c propulsion.
the signal circuit depend on the amount of current flowing, in the power rail, the resistance per 1000 feet of the power rail, and the length of the track circuit.
As shown in Figure 119, a balancing impedance, made up of a reactor and a resistor, is used to neutralize the effects of this d.c., which could saturate the relay magnetic structure and interfere with its operation. The reactance of X in the balancing impedance is quite high: 24 ohms at 5 volts
Figure 119. Single-rail track circuit for d-c. propulsion.
60 Hz, and 10.6 ohms at 5 volts 25 Hz. The resistance of X, 0.67 ohm, is the same as that of resistor R in the balancing impedance. Due to the high impedance of reactance X, practically all the ac. for operation of the relay flows through resistor R and winding Cl of the relay track element, operating the relay.
When there is any flow of d.c. from the track to the relay, it divides symmetrically between the equal-resistance d-c paths made up of R plus winding Cl, and X plus winding C. Since the d.c. flows in opposite directions in C and Cl, the opposed ampere-turns neutralize each other, and no net magnetic flux is produced in the magnetic circuit of the relay track element.
The balancing impedance has certain limitations. One is the heating effect of the d.c. on elements R and X of the impedance and on windings C and Cl of the relay. They will stand 5 amperes d.c. continuously and intermittent short period spurts of 10 amperes d.c.
A second limitation is the saturation effect on the laminated steel magnetic circuit of the reactor by the d.c. flowing through its coil. Flux induced in the magnetic circuit by the d.c. reduces the permeability of the magnetic circuit, with corresponding reduction in the impedance of the reactance section of the balancing impedance. This permits some increase in the a-c flow through the reactance section and its corresponding section C of the track element winding which opposes the a.c. in the Cl section of the track winding, thereby weakening the operation of the track relay.
Tests have shown that 5 amperes d.c. from the track to the balancing impedance and track relay will require an increase of about 8% in the a.c. from the track to maintain the same strength of operation as when no d.c. is present.
Single-rail track circuits are used extensively on rapid transit lines. They are also used at interlockings on railroads which use double-rail circuits elsewhere.
A-C Track Circuits for A-C Propulsion
As previously indicated, a-c propulsion complicates the problems associated with a-c track circuit design, since the track circuit must be immune to propulsion a.c. This requires that track circuit a-c. energy have characteristics that permit positive identification in the presence of propulsion a.c.
A typical element in solving the problem is the use of a-c track circuit energy with frequency different from the a-c propulsion frequency, with the track relay energizing circuits responding selectively to this energy. Track circuits using static- frequency and centrifugal-frequency a-c track relays, mentioned earlier in this section, represent solutions based on earlier technology for discriminating between frequencies. Such relays, however, were bulky and relatively costly. Moreover they were not adaptable for use as code-following track relays, a highly desirable feature. More recent technology retains the basic idea of using a distinctive a-c frequency in the track circuits, but superimposes other special characteristics, for example modulation of the a-c track energy, to add an additional distinguishing feature. By means of circuitry which requires not only that energy of track-circuit frequency be present, but also that it be modulated in a specified way, these newer designs result in efficient a-c track circuits which permit use of d-c track relays. At the same time these circuits provide excellent protection against track circuit malfunction which might arise from the effect of a broken rail or failure of an insulated joint.
A typical form of modulation is on-off coding of track energy at a relatively low rate, on the order of a few Hz. Similar coding may also serve to transmit information via the rails for signal control, as described in Section 400, Track Circuits, Coded and Electronic. At this point, however, we will consider coding only as an aid in identifying track circuit energy in the presence of propulsion currents.
Phase-Selective A-C Track Circuit
The phase-selective a-c track circuit is applicable to d-c or a-c propulsion systems. It may be readily coded for use with cab signal systems or for other coded track circuit functions. The coded a-c energy of the track circuit is rectified to produce coded d-c output which, in conjunction with rectified phase reference a.c., drives a d-c magnetic stick relay alternately to its normal and reverse positions. Proper code-following action of this code- responsive track relay energizes the track-detector relay, as described in Section 400 of this publication. Invasion of the track circuit by propulsion energy disrupts the operation of the code-responsive track relay. If the track relay is not coding, the track detector relay will not pick up.
The GRS Type B magnetic-stick relay, Figure 120, is an important element in the operation of the phase-selective track circuit. A magnetic-stick relay operates in response to a change in the direction of current flow through its coils. The armature is mounted in a vertical plane, and is pivoted so that it can rock toward either the upper or lower poles of the electromagnetic structure. No gravity or spring bias is applied. Two permanent magnets included in the relay magnetically polarize the armature. The poles induced on the armature are attracted to magnetic poles of opposite polarity
Figure 120. Magnetic-stick relay, GRS Type B Size 1.
and repelled by poles of the same polarity. The direction in which the armature will move thus depends on the polarity of the electromagnetic structure as determined by the direction of current in the relay winding. Since the armature is not mechanically biased, the permanent magnetic field retains the armature in the position to which it was lasted operated if the coils are de-energized. If the coils are alternately energized so that the electromagnetic structure reverses polarity, the armature rocks between its normal and reverse positions. Figure 121 illustrates this operation.
Principles of Operation
Figure 122A shows the general arrangement of a phase-selective track circuit.
Track-circuit frequency is selected to be distinct from propulsion-circuit frequency and from harmonics thereof. In European practice, for example, where 50-Hz propulsion is used, track-circuit frequency could be 83-1/3 Hz. With 60-Hz propulsion, track-circuit frequency typically would be 100 Hz.
For purposes of this discussion, it will be assumed that the track frequency is 100 Hz. The 100- Hz a.c. is obtained from a line circuit installed for the purpose, and is fed to the rails through a track transformer. Coding contact CPR (driven by code- generating circuitry described in Section 400) alternately energizes and de-energizes the transformer primary, producing on-off intervals of track energy. Resistor R1 limits the maximum current which the track circuit can draw.
At the relay end, a track transformer and a phase reference transformer feed a phase-selective
detector. The phase-selective detector unit includes a transformer with windings L2A and L2B, two bridge rectifiers CR1 and CR2, and filter L1-C1. The filter has low impedance at 100 Hz but substantial impedance at propulsion frequency, thus attenuating any propulsion-frequency energy which might reach the phase-selective detector. Note that the a-c feed to rectifier CR1 is obtained from the secondary of the phase reference transformer in series with the secondary of the track transformer through filter L1-C1. By choice of taps (not shown) on the track transformers, and by selective paralleling of the capacitors included in the filter, the voltages impressed on CR1 by the two secondaries are adjusted to be essentially equal and 180 degrees out of phase.
Figure 122B indicates circuit conditions during the energy-off period of the track code. A-c line energy flows from the phase reference transformer through the track transformer (also through L2A winding of the phase-selective detector in parallel) to rectifier CR1. D-c output of CR1 energizes the N coil of magnetic-stick relay TR, driving the TR contact to the N position. Since relay coil N makes up the predominant load in this circuit, nearly all the circuit voltage appears across the relay coil. With relatively little drop across other circuit elements in the path, such as the track transformer and winding L2A, there is negligible effect on circuits to which these windings are coupled.
Figure 122C shows circuit conditions when the track is energized. Since the phase reference voltage is essentially equal and opposite to the track transformer voltage, the circuit through rectifier CR1 is effectively turned off. However, winding L2A receives full track transformer voltage, inducing comparable output from winding L2B. Output of L28, rectified by CR2, energizes winding R of relay TR, driving the TR contact to the R position.
Alternate on-off coding of the track circuit thus causes TR continually to reverse its position. If track energy is lost for any reason, such as a track shunt, TR remains in the N position. If line energy is lost, TR remains in the R position. If foreign energy invades the phase-selective detector so that the phase and amplitude balance between the line and track energy is lost, TR also fails to follow the code.
Continual coding of TR is required to keep an associated track detector relay energized, as described in Section 400. Thus the track detector relay of the phase-selective track circuit releases with
Figure 121. Operation of Type B magnetic-stick relay.
train shunts or if anomalous circuit conditions develop the desired performance in a track circuit.
Coded Carrier Track Circuit
The phase-selective track circuit permits track energy to be coded on-off at a variety of selected rates for signal control purposes. The track circuit would thus be coded at different rates at different times as track and traffic conditions require. Protection against improper track circuit operation that might result from failure of insulated joints is obtained by relating the phase of the received track energy to the phase of the reference line. By staggering the phasing of adjacent track cirCu its, the desired protection is obtained.
If the signal system on an a-c propulsion railroad does not require rate codes for signal control, it is feasible to employ unvarying but different code rates in adjacent track circuits. This makes it possible to distinguish, on the basis of code rate, between legitimate track circuit energy and energy that might be present because of insulated joint failure. The coded carrier track circuit for a-c propulsion uses this principle.
The block diagram, Figure 123, shows the arrangement of a typical coded carrier track circuit. The transmitter and receiver are compact solid-state electronic units. A-c track energy is generated and modulated at code rate by the transmitter, which receives its input energy from a local d-c supply. The receiver processes energy from the
Figure 122A. Phase-selective a-c track circuit.
Figure 122B. Phase-selective a-c track circuit track de-energized.
Figure 122C. Phase-selective a-c track circuit track energized.
track circuit and produces a coded output appropriate to driving a conventional rate decoder and track relay as described in Section 400. Like the transmitter, the receiver is energized by local d-c supply. The code rate used in this circuit may be any of the available rates, such as 120, 180, 270 or 410 on-off pulses per minute. Adjacent track circuits would be coded at two of the other rates available. The same code would thus be repeated only at every fourth track circuit in the sequence.
For illustrative purposes, Figure 123 shows a type of impedance bond which functions also as a track transformer. An additional winding on the bonds magnetic structure provides this facility. However, circuit operation does not require such bonds. Equivalent operation can be obtained with conventional bonds and separate track transformers.
The carrier frequency, which is typically 83-1/3 Hz for 50-Hz propulsion energy or 100 Hz for 60-Hz propulsion, is generated independently by each transmitter. The system is not sensitive to minor variations in carrier frequency, nor to the phase of the carrier energy in the rails. Hence the independently generated carriers at each frequency need not be precisely identical. This eliminates the need for a signal-frequency power line along the railroad, a major advantage.
The transmitter incorporates two basic functioris: carrier generation and code generation. The carrier is modulated by the code, and fed to the rails at an appropriate power level through the transmit impedance bond.
At the receiver end of the circuit, the modulated carrier is transferred from the rails to the receiver through the receive bond. A filter in the receiver rejects off-frequency energy, such as might appear from propulsion current and its harmonic components. Track-circuit energy is amplified and demodulated to extract the code, which is checked in a level detector. The level detector verifies that the code pulses swing between pre-established voltage levels and have on-off time values characteristic of valid codes. Additional amplification then increases the code energy to a level appropriate to driving a conventional coded track rate decoder. With input of the appropriate code rate at the required level, the decoder produces d-c output to pickup the d-c neutral relay which serves as the track detector relay.
The preceding discussion has been confined to coded carrier track circuits in which differing code rates in adjacent track circuits provide protection against breakdown of insulated joints. As noted, this is not compatible with the use of coded cab signal and train control systems. It
Figure 123. Block diagram for coded carrier track circuit.
should be mentioned that a track circuit using essentially the same solid-state electronic apparatus has been developed by GRS which does permit coding for signal control. In this design, a 100-Hz signal-frequency line is employed, and a receiver used which is sensitive to the relative phase of the track energy and the line. Broken down joint protection is provided by staggering the phase of track energy in adjacent track circuits. Individual track circuits may be coded as required without regard to the code rate present in adjacent track circuits. The system is thus closely related to the phase-selective concept previously described, but uses electonic circuits in its operation.
TRACK CIRCUIT DATA AND CALCULATIONS
In the earlier discussion of d-c track circuits, it was pointed out that track circuits and the shunting of track circuits are not as simple as might appear from schematic diagrams. A-c track circuits introduce further considerations involving frequency discrimination and phase relationships which make for additional complexity. High-frequency track circuits, discussed later, involve circuit parameters which are negligible in d-c or a-c track circuits but which become of controlling importance at audio frequencies.
Nevertheless, given characteristic data on a track, it is possible to apply mathematical methods to assist in the design of track circuits for that track. The information needed is related to the transmission line characteristics of the track, including series rail resistance and inductance, also shunt leakage and capacitance. Measurements made on the track, in terms of input impedance with the track open-circuited and short-circuited at a remote point, are an aid in obtaining the necessary data.
GRS has developed a number of computer programs for use in designing track circuits, applicable to tracks having a variety of physical and electrical characteristics. Track circuit performance predicted by these programs has been substantiated by extensive observations on actual track circuits in commercial service.
The computer programs are used by GRS for signal system work in which GRS has responsibility for track circuit design. The programs are also used by GRS to provide consulting service to railroad signal engineers engaged in track circuit design.