Signal Control for One Direction Only 402
Introduction 402
Elementary Principles 402
Apparatus 402
Code Transmitters 402
Code-Responsive Track Relay 404
Master Decoding Transformer 405
Decoding Unit 406
Multiple-Indication Signaling Without Signal-Control Line Wires 407
For a Proceed Aspect 408
For a Stop Aspect 408
For an Approach Aspect 409
Approach Lighting 410
Series Circuit 410
Inverse Code 411
Decoding Inverse Code 413
Cut Sections 413
For Direct Code Only 413
For Both Direct and Inverse Codes 413
Coded Track Circuit for Cab Signals 416
Trakode® - Signal Control for Both Directions 416
Introduction 416
Operation 416
APB Signal Control 418
Typical Siding 420
Transmission 422
Pulse Timing 422
Decoding 423
Independent Coding 423
Control of Three-Position Signals 424

Intermediate Signal Location . 424
Repeating Cut Sections 428
Electronic Track Circuits 429
Track Circuit Characteristics at Audio Frequencies 429
Frequency-Shift Overlay Track Circuits 430
FSO Track Circuits for Highway Crossing Warning 430
Series FSO for Switch Lock Release 432
High-Frequency Track Circuits
WEE-Z® Bond 436
Application of High-Frequency Track Circuits
Coded Frequency-Shift Overlay Communication Circuits 436
Equipment and Application 436



The coded track circuit is not essentially different from the steady-energy track circuit. It is not so much a new kind of track circuit as it is an improvement upon the steady-energy type, since it retains all the advantages and has, in addition, other desirable attributes. Among these additional advantages are : improved shunting sensitivity, improved broken-rail detection, increased average track- circuit length, improved protection against the effects of foreign currents or failure of insulated joints, and elimination of cut sections. As will be explained later, “inverse” and “follow-up” codes to effect approach locking, approach lighting, annunciation, etc., may also be applied without adding more line wires.
Because of the improved shunting sensitivity of the coded track circuit and the high efficiency of the code-responsive track relay, it is possible to increase the length of a coded track circuit well beyond the length that would operate satisfactorily with a steady-energy track circuit- assuming similar conditions for each. Thus coded track circuits do not need as many cut sections as steady-energy track circuits.
The same factors that contribute to improved shunting sensitivity also contribute to improved broken-rail detection. The shunting sensitivity of the coded track circuit is affected directly by the relay’s pickup value rather than by its release value. Furthermore, the code-responsive track relay requires more current (intermittently) than the steady-energy track relay. Therefore, the leakage resistance around a broken rail does not have to be as high to effect a stop indication with a coded track circuit as it would with a steady-energy track circuit.
Steady foreign current of sufficient strength to affect a code-responsive track relay would only tend to hold it inoperative, either picked up or released, depending on the polarity of the foreign current, and would thus ensure that a stop aspect be displayed.
By interrupting or “coding” the current to the rails at different rates in accordance with the condition of the next track circuit in advance, and by providing appropriate apparatus at the relay end of the track circuit that will respond selectively to the code being received, multiple proceed aspects may be controlled through the rails, thus eliminat in

signal-control line wires. This and other capabilities of the coded track circuit will be explained more fully in the following pages.
Elementary Principles
Figure 401 illustrates an elementary coded track circuit. Code transmitter CT at the battery end of the circuit interrupts the energy to the rails 75 times per minute so that the energy in the rails pulses in alternate “on” and “off” periods.
At the other end of the track circuit, code- responsive track relay TR alternately picks up and releases in response to these impulses of energy and controls the flow of local energy to the master decoding transformer.
The signal circuits are controlled by HR, a steady-energy repeater of the code-responsive track relay. HR is so controlled through the contacts of TR and by the master transformer that it will release unless TR is coding.
When a train enters such a track circuit, the value of the train shunt need only be sufficient to reduce the energy in the coils of TR relay to a point below its pickup value to cause track-detector relay HR to drop.
In the steady-energy d-c track circuit, the value of the train shunt must be sufficient to reduce the energy in the track relay to a point below its release value. Inasmuch as it takes less energy to hold a relay in its picked-up position than it does to cause a relay to pick up from its released position, the coded track circuit has a correspondingly higher shunting sensitivity. Furthermore, if for any reason the code-responsive track relay should remain in either the energized or the de-energized position, HR would drop.


Code Transmitters

The code transmitter (CT) performs the important function of controlling the current supplied to the rails so that the rails will be intermittently energized, with the “on” and “off” periods of approximately uniform length. The rate at which these periods occur determines the “code” as 75, 120, 180, etc., per minute.
The plug-in, oscillating code transmitter is shown in Figure 402. It may be used with a repeater relay, depending upon the requirements of the circuit.


Referring to the simplified operational diagram of the oscillating code transmitter, Figure 403, when energy is first applied to the coil, magnetism builds up in the pole pieces and causes the armature to rotate counterclockwise. At a given pdint in this counterclockwise swing, the upper cam allows its contact to open thus interrupting current to the coil and removing the magnetism. The armature, however, continues to move counterclockwise until its momentum is overcome by the increasing resistance of the spring.
The mechanical energy stored in the spring by the counterclockwise movement now moves the armature clockwise, imparting sufficient momentum to the armature to carry it past the dead-center point from which it started. As the armature passes the dead-center point, the upper cam again acts to close the coil circuit. The armature continues to rotate clockwise until the magnetic attraction and the spring overcome its momentum and start it back again
- and so the oscillations continue. The recurrent pulses of magnetic attraction supply the energy to start the oscillations and to do the work of closing the contacts and overcoming the slight friction of the mechanism.






.lv I
Figure 401. Simplified operational diagram of coded track circuit using one code rate and line wires.

Figure 402. Type B oscillating code transmitter.




The action of the coiled spring and rotating
mass may be compared to the action of the balance
wheel and hair spring in a watch. They constitute
a rotary pendulum with a “rate” which is established
by the physical properties of the spring and
the rotating mass. Thus a given weight with a given
spring will develop and maintain a certain frequency
of oscillation. The rotating mass of the
75-code transmitter, for example, has more non-
magnetic weight discs attached to it than that of
the 180-code transmitter. Because of the inherent
frequency characteristic of each transmitter, the
code frequency is held within very close limits,
regardless of normal voltage variations. The oscillating
code transmitter is ruggedly designed to give
years of continuous service with a minimum of
A solid-state code transmitter, Figure 404, is
also available, which has, by electronic means,
provisions for generating as many as seven different
code rates. A separate code rate printed
circuit board is plugged into the transmitter for
each code rate desired. Presently, circuit boards
are available for 75, 120, 180, and 270 code rates. Each board can drive up to four separate relays.
Code-Responsive Track Relay
At the relay end of the track circuit, a relay is provided that will respond quickly so that it will follow the coded pulses of current supplied to the rails by the coding apparatus at the highest code rate that may be used in the system. This relay, Figure 405, has two dependent front-back contacts.
The simplified operational diagrams, Figures 406 and 407, illustrate the theory of operation of the code-responsive track relay. The middle leg of the core is a permanent magnet, with polarities as marked. In the de-energized position, Figure 406, a magnetic circuit is completed through the lower leg of the core and one half of the armature. Very little of the magnetism of the permanent magnets is diverted through the upper leg of the core and the other half of the armature because of the greater reluctance (magnetic resistance) of the wider air gap there.
When current of the proper polarity flows through the windings, Figure 407, the current flowing in the upper coil produces magnetism that combines with that of the permanent magnet, while the current flowing through the lower coil produces
Figure 403. Simplified operational diagram of oscillat- magnetism opposing that of the permanent mag tin code transmitter. net. As a result, the intensity of the magnetic flux

Figure 404. Solid state code transmitter.





across the lower air gap decreases, but the magnetic flux across the upper air gap increases to such an intensity that the armature is attracted to it, and the front contacts are closed.
When current ceases to flow in the coils, the bias spring returns the armature to the de-energized position, Figure 406, and the back contacts are held closed by the spring and by magnetic attraction of the permanent magnet.
Thus it may be seen that the code-responsive track relay is in effect a polar biased relay, that is, it will close its front contacts only when current of the correct polarity is flowing through its coils. When the relay is de-energized or when current of wrong polarity flows through the coils, the back contacts will be held closed.

Master Decoding Transformer
The master transformer, Figure 408, (see also Figure 401) performs three functions:
1. It provides a link between the track relay and the track-detector relay, and it acts in such a manner as to provide energy to pick up track- detector relay HR only when the track relay is coding.

2. It supplies an a-c output which is mechanically rectified to furnish pulsating direct current to track-detector relay HR.
3. It supplies alternating current to the decoding units (described later) at the code frequency being received.
Figure 409, an operational diagram of the master transformer, shows the conditions that exist as the code-responsive track relay is alternately energized and de-energized in response to the code. Figure 409a shows the circuit at the instant the track relay front contacts are made. The d-c energy applied through the front contact of the track relay to the upper half of the master transformer primary winding creates a rising magnetic flux in the trans Figur

406. Simplified operational diagram of code- responsive track relay. Relay shown in position it assumes when track circuit is de.ener. gized.

i_I I


Figure 407. Simplified operational diagram of coderesponsive track relay. Relay shown in position it assumes when track circuit is energized.

Figure 405. Code-responsive track relay.




— — —






former core. This results in a rising voltage in the
secondary winding. The upper secondary winding
is connected to the coils of the track-detector relay
(HR) through another front contact of the track
relay. This causes the HR relay to become energ
At the end of the “on” period of the code, the
track relay drops. Figure 409b shows the circuit at
the instant the track relay back contacts are made.
When the track relay breaks its front contact, the
flux in the master transformer starts to decrease,
reversing direction, but after it has passed through
a zero point. The reversal of the magnetic flux in
the primary side of the transformer causes the flux
to reverse in the secondary side of the transformer,
resulting in a change of polarity of the output voltage.
However, the second contact of the track
relay connects the HR relay, through a back contact,
to the lower portion of the transformer coil.
Thus the polarity of the output voltage remains unchanged.
Since the HR relay is a slow release relay,
it will stay picked up during the coding cycle.
If the track relay should cease to code because
of track circuit occupancy, or any other reason, the
flux in the primary of the transformer would build
up as shown in Figure 409c. Since the transformer
can only have an output when the flux is changing,
Figure 409. Simplified operational diagrams of master
there would be no output, and the HR relay would drop. A similar situation would exist if the track relay ceased coding in the energized position.
Decoding Unit
All the apparatus necessary for the development of the circuit shown in Figure 401 has thus far been described. However, line wires would be required if more than one proceed aspect is desired. We will now discuss the additional apparatus that makes possible the control of multiple proceed aspects without using any signal line wires.
To secure multiple proceed indications, different codes are applied to the rails at the battery end, the code selection being governed by the position of the next signal in advance.
As has already been shown, the master transformer supplies an a-c output whose frequency corresponds with that of the code being transmitted. The decoding unit, Figure 410, which is a tuned device, is connected to the master transformer to provide a means of identifying the code being transmitted, as the fundamental characteristic of a decoding unit is that it will pass energy to pick up the relay connected to it when the input energy is supplied at the frequency for which the

a. Track circuit momentarily energized.

b. Track circuit momentarily deenergized.

c. Track circuit shunted by train.

Figure 408. Master decoding transformer.






decoding unit was designed and tuned, such as 120 or 180 cycles per minute.
As shown in the simplified operational diagram, Figure 411, the decoding unit contains a transformer, a capacitor, and a full-wave rectifier. The capacitor is connected in series between the secondary of the master transformer and the primary of the reactive transformer in the decoder. The rectifier is connected across the secondary coil of the decoder transformer to convert the a-c output to direct current to operate a d-c relay.
In order to understand why the 180-code decoder, for example, will pick up its associated relay when it is supplied by the master transformer with 180-cycle-per-minute alternating current and yet will not pick up its relay in response to 75- or 120- cycle-per-min,ute alternating current, it is necessary to consider briefly the fundamentals of series resonance.

The circuit within a decoding unit includes a coil having an iron core (the primary of the decoding-unit transformer), which is an inductance, connected in series with a capacitor. When the master decoding transformer is operating at the specific frequency to which a given decoding unit has been adjusted, the master decoding transformer will supply to that decoding unit more current than at any other frequency.

If the frequency of the master decoding transformer output is below that for which the decoder circuit is tuned, the greatest opposition to the flow of current through the circuit is the reactance of the capacitor, since capacitive reactance increases as frequency decreases. If the frequency of the master tranformer output is too high, the current flow is restricted by the reactance of the coil, since inductive reactance increases as the frequency increases.
At a certain frequency between these high and low points the inductive reactance will be equal to the capacitive reactance. For the 180-code decoder, for example, this frequency is 180 cycles per minute. This is the resonant frequency of the circuit. Since the inductive reactance in the circuit produces a positive effect, and the capacitive reactance produces a negative effect, they cancel each other when they become equal in value. This leaves the resistance of the circuit as the only oppostion to the flow of current. Under this condition, the current through the decoding unit is at its maximum and of sufficient value so that the secondary output of the decoding-unit transformer will pick up its associated relay as shown by the graph in Figure 412.
Reference to the graph will show how even a small deviation from the 180-code rate will sharply decrease the output of the decoding unit so that it does not pick up its associated relay. When the master decoding transformer output is 75 or 120 cycles per minute, for example, the capacitive reactance of the 180-code decoding unit is far too high to permit sufficient current to flow through it and produce an output that will pick up its relay.
Multiple-Indication Signaling Without Signal-Control Line Wires
Figure 413, shows a basic circuit for two-block, three-indication signaling without signal-control line wires. Note that an additional code transmitter has been added to the circuit. There is ordinarily one more code transmitter than there are decoding units because any of the codes is detected by the operation of HR directly from the master transformer.
A code-transmitter repeater has also been added to eliminate the necessity of carrying track current through the selecting circuit.

Figure 410. Decoding unit.





Figure 411. Simplified operational diagram showing decoding operation.


For a Proceed Aspect

i-• c Wit

track circuit 4T and the next track circuit in advance of 4T unoccupied, signal 4 is in the proceed position, and track-detector relay HR at signal 4 has its front contacts closed to feed current to the 180-code transmitter. The 180-code transmitter operates to interrupt local current to code-transmitter repeater CTPR at 180 pulses per minute. CTPR, in turn, interrupts the current fed to the rails (2T) 180 times per minute.
Code-responsive track relay TR at signal 2 follows the 180 rail code and codes local current to the master decoding transformer, alternately energizing the transformer primary windings and thus producing an alternating voltage which, when mechanically rectified by a second contact on TR, provides direct current to pick up slow-acting track-detector relay HR.
The master decoding transformer also supplies alternating current at the 180 frequency to the 180-code decoding unit.
The 180-code decoding unit, which is tuned for 180 cycles per minute (3 cycles per second), passes the 180-code energy to pick up decoding relay DR. With HR and DR up, signal 2 displays a proceed aspect.
For a Stop Aspect
With track circuit 2T occupied and track circuit 4T and the track circuit in advance of 4T unoöcupied, the 180-code transmitter continues to operate as in the preceding instance, and code- transmitter repeater CTPR continues to interrupt. local current to the rails (2T) at 180 impulses per minute.
















Figure 412. Circuit diagram of decoding unit and graph of decoder output current in relation to frequency of input current from master decoding transformer.

The coded current cannot reach code-responsive track relay TR to operate it because of the train shunt. With TR no longer following the code, pulsing current is no longer fed to the master decoding transformer, and it produces no output either to pick up track-detector relay HR orto pick up DR through the decoding unit. With both HR and DR released, signal 2 displays a stop aspect.
For an Approach Aspect
With 4T occupied and 2T unoccupied, 4T track- detector relay HR and signal’ 4 repeater relay HDGPR are released and close their back contacts. This interrupts the local current to the 180-code

transmitter and applies it to the 75-code transmitter. The 75-code transmitter operates to cause code-transmitter repeater CTPR to interrupt the current fed to the rails (2T) 75 times per minute.
Code-responsive track relay TR follows this 75-code and codes local current to the master decoding translormer at the same rate, thus energizing HR. The master decoding transformer output to the 180-code decoder is now alternating at 75 cycles per minute. The 180-code decoder will not pass current to pick up its associated relay, DR, when energized at other than 180 frequency. Thus DR will not pick up. With DR released and







75 120 180




Figure 413. Simplified circuit for 2-block 3-indication coded track circuit without signal control line wires. Contacts are shown in the positions they assume when tracks 2T and 4T are unoccupied.

HR picked up, signal 2 displays an approach aspect. When the train clears track circuit 4T, HR and HDGPR will again be picked up, and the circuit will return to the condition as described under For a Proceed Aspect.”
Approach Lighting
It is common practice to light signals only when a train is approaching, in order to save energy, lamps, etc. This approach lighting may be secured in

various ways with coded track circuits. One method is shown in Figure 414, where the approach relay is connected in series with contacts of the CTPR relay. Approach lighting is effective approximately 4000 feet in advance of the train, the distance being dependent upon ballast resistance, adjustment of resistors, and other track conditions.
Series Circuit
Referring to Figure 414, series approach lighting operates by increasing energy to AR as the




[a. __


Figure 414. Simplified series approach-lighting Circuit. Contacts shown coding. Track 2T is unoccupied.





train nears. Since it is economical in current consumption, it is well suited for primary-battery installations.
With track circuit 2T unoccupied, current from the positive side of the track battery flows through two parallel branches: one through approach relay AR and resistor A, the other through resistor B. The branches then join, and the current flows through the coding contacts, out to the rails, and through code-responsive track relay TR. Resistors A and B are so adjusted that most of the current flows through resistor B, and the current flowing through AR and resistor A is not great enough to operate
When a train enters track circuit 2T, the low resistance of the train shunt permits an increased flow of current from the track battery, and sufficient current then flows through AR and resistor A to make AR repeat the code to hold up slow- release relay APR. The lighting circuit to signal 4 is then closed through the front contact of APR.
Inverse Code
The series approach-lighting circuit just described does not provide for full-block approach lighting under all conditions. This can be accomplished without the use of line wires by sending an “inverse code” from the relay end of a track circuit to the battery end. The current impulses for an

inverse code are transmitted through the rails during the off” periods of the direct code; see Figure 415.
Figure 416 illustrates the fundamentals of a typical inverse-code circuit. The contacts are shown in the positions they assume during the transmission of a direct-code current impulse. At the leaving end, code-transmitter repeater CTP is picked up, and current flows to the rails in series with approach relay AR, which is a magnetic stick relay. It is magnetically held in its last-operated
& 80 COO€
Figure 415. Inverse code, shown in red is transmitted during the “off’ periods of the direct code.




position. With positive current entering its negative terminal, it assumes its reversed position.
Flowing through the rails to the entering end, the current passes through the back contact of an impulse relay, TPA, to pick up code-responsive track relay TR, which is a polar-biased relay as described previously. When TR picks up, a direct- code impulse is applied to the master transformer and to the decoding unit corresponding to the code being received. (For purposes of simplification these units are not shown. Operation would be same as has been described previously).
Thus far the path of a direct-code impulse has been described. The following deals with the path of an inverse-code impulse as shown in Figure 417. This is the same circuit as shown in Figure 416 except that the contacts are shown in the positions they assume during the transmission of an inverse- code current impulse.
To return to the leaving end, CTPR closes its back contacts, connecting the negative terminal of AR to the negative track lead. The positive terminal of AR is regularly connected to the positive track lead.
The current of the direct-code impulse ceases to flow through TR, and TR drops, applying local current through its back contacts to both TPB and impulse relay TPA. TPA picks up faster than TPB,

which is slightly slow pickup, thus closing its contacts before TPB can pick up and break the circuit to it. During the interval that TPA is up, a short impulse of inverse-code current is applied to the rails. The duration of this inverse-code current impulse is governed by the time TPA remains in its energized position.
Passing through the rails to the leaving end, the inverse-code current flows through magnetic stick relay AR at its positive terminal, and the relay assumes its normal position. Local current then passes through the front contacts of AR and picks up APR. APR is sufficiently slow in releasing so that the recurring current impulses received through front contacts of AR will hold it up.
A train entering track circuit 2T prevents TR from again picking up. With TR back contacts steadily closed, TPB remains up and TPA remains down, thus stopping transmission of the inverse- code current impulses. The direct code current impulses now pass through the circuit completed by the train shunt, and AR is driven to its reversed or “dropped” position. Thus current is cut off entirely from APR, and it drops to close an approach-lighting circuit to signal 4. Note that this system gives full-block indication, as the inverse code ceases immediately upon entry of a train into track circuit 2T.

Figure 416. Simplified typical inverse-code circuit. Circuit is shown with contacts positioned at the instant of a directcode pulse.




Decoding Inverse Code
When inverse code is used to control vital circuits such as approach locking, annunciation, indication, etc., it is generally considered preferable to decode the inverse code, using an arrangement such as is shown in Figure 418, where APR is controlled by current derived from a master decoding transformer designed for use with inverse code. This system requires that AR be continuously pulsing in order to hold APR with its front contacts closed.
Figure 418 represents the circuit as coding, that is, responding to the recurring current impulses of direct and inverse code. The approximate relative spacing and durations of the inverse code current impulses and 180-rate direct-code current impulses are illustrated in the code pattern diagram above the circuit.
Cut Sections
In installations where signals are spaced so far apart that the distances between signals are greater than the practical operating length of a coded track circuit, it obviously becomes necessary to introduce a code-repeating cut section. It should be noted here that where one cut section would be

required for a coded installation, two or even more cut sections would probably be required for steady- energy track circuiting of the same installation.
For Direct Code Only
Figure 419 shows a cut section for use in installations with direct code only. The polarities of circuits A2T and B2T are as shown to ensure that restrictive signals will be displayed if both insulated joints at the cut section should become defective. In such a case, the track battery at the cut will retain TR in the energized position, no code will be transmitted to A2T, and signal 2 will display its most restrictive aspect.
For Both Direct and Inverse Codes
Figure 420 shows the fundamentals of a cut section for an installation using both direct and inverse codes. Operation is as follows: when a direct-code current impulse is applied to the rails of track section B2T at signal 4, code-responsive track relay TR is picked up through back contacts of TPA. With TR up, a direct-code current impulse is relayed to A2T through front contacts 2 of TR and through the coil of AR. AR isa magnetic stick relay. It is magnetically held in its last-operated position. With positive current entering its negative terminal, it is driven down and closes its back contacts, applying current to the master decoding transformer which, in turn, energizes APR in the manner

Figure 417. Simplified typical inverse-code circuit. Circuit is shown with contacts positioned at the instant of an inversecode impulse.





Figure 418. Portion of simplified typical inverse-code circuit showing method of decoding inverse code. Circuit is shown with contacts coding, track circuit unoccupied.

described in the section concerning the master transformer.
When the direct-code impulse ceases and B2T is de-energized, TR drops and TPA is then energized through the front contact of APR, TR back contact 1, and the back contact of TPB. (TPB picks up less quickly than TPA.) With TPA up, an inverse- code current impulse is applied
tc, B2T through the

front contacts of TPA. The inverse-code current impulse is terminated when TPB picks up and drops TPA.
Although the inverse-code impulses transmitted to B2T are originated at the cut section, they can be transmitted only when AR is being alternately picked up by an inverse-code impulse from A2T and driven down by the relayed direct-





Figure 419. Simplified typical cut-section circuit for direct code only.





Figure 420. Simplified typical cut-section circuit for both direct and inverse codes.







code impulses from B2T, since APR will hold its front contact closed only when AR is coding. Thus, whenever the inverse-code current impulses cease to flow in A2T, none will be transmitted to B2T.
Coded Track Circuit for Cab Signals
Figure 421 shows a simplified circuit for coded track circuits designed to operate cab signals on installations where there are no wayside signals. (For engine-borne equipment, see Section 600, ‘Cab Signals.”)
This is practically the same as the elementary circuit for 2-block, 3-indication signaling shown in Figure 413, except that the codes are selected by HR only, and alternating instead of direct current is fed to the track circuits. The alternating current coded by the code transmitter is fed to the primary of a track transformer. The pulses of alternating current produced in the secondary winding are fed to the rails and operate the engine-borne cab signaling equipment in accordance with the code frequency. 180-code causes a proceed aspect to be displayed; 75-code caution; and no code, a stop aspect.
At the relay end of the track circuit, the a-c code impulses are rectified to operate the d-c code- responsive track relay.
Another form of coded track circuit control is known as GRS Trakode. This system employs track circuits in which the energy is polarized, and pulsates at a relatively slow rate. These track circuits provide a means of controlling wayside signals in both directions without the use of line wires. Hence the system may be applied to control absolute permissive block signaling or the automatic signals in centralized traffic control territory.
Trakode can also be used to control approach locking, approach indication, and electric switch locking on tracks signaled for double direction running. It may be used with all types of signals and also with intermittent inductive train control.

Because the track energy is pulsating, the track circuits can usually be considerably longer than conventional direct current track circuits - for reasons same as given for rate coded track circuits. The slow rate of pulsation lengthens the service life of the code-responsive relays.
To help understand the basic principles of Trakode, consider a polar track circuit having a polar-biased track relay. This relay is equipped with two armatures, one responding to each polarity of current. By pole changing the current at the transmitting end of the track circuit, two pieces of information can be sent to the relay end. This could be used to control an H ora D relay for controlling a signal. By pulsating this current so that the track circuit is alternately energized for a fraction of a second and then de-energized for about two seconds, more can be accomplished as follows:
1. Track circuits can be made longer.
2. By changing polarity, positive and negative code characters are available, but by pulsating the energy, two more characters may be transmitted. This is done by pole changing the current in the middle of an energized period, so that one rail is first negative and then positive in respect to the other. The order may be reversed so that positive is first. Hence these code characters are negative-positive and positive-negative respectively. At the receiving end of the track circuit, decoding units detect which of the four characters is received.
3. In the long de-energized period, the same information may be transmitted back from the relay end of the track circuit. Therefore, any one of the four characters may be used to control an opposing signal: positive, negative, negative-positive, and positive-negative.
A pulse transmitted from a head block (end of siding) is received at the first intermediate signal, or repeating cut location, and is repeated to the next track circuit as soon as it is received. The pulse is thus repeated through a succession of track circuits until it reaches the next head block.
However, at the intermediate signals, the character of the pulse about to be repeated may be changed from that being received, to indicate the position of the intermediate signal.
When this changed pulse is received at the next head block, it conditions the circuits so that a




Figure 421. Simplified circuit for coded a-c track circuit for cab signals without wayside signals.

similar pulse is transmitted and repeated back to the first head block, during the de-energized period between pulses.
It must be remembered that for normal Trakode operation, only one pulse is being transmitted through a block at a time. A pulse is sent in one direction, and after it has been received at the other end, another pulse is sent back. This is termed “dependent” coding, since the transmission circuits at the end of the block are conditioned by the pulse being received so that they will transmit a pulse back as soon as reception is ended.

When a train enters a black, it prevents the transmission of pulses to the opposite end and so ends the dependent coding. It also puts all opposing signals in the block to stop. The apparatus at the opposite end of the block then transmits pulses at a fixed rate and with polarity characteristics to give the proper signal indications to the train. Such transmission is termed “independent” coding.
When Trakode is used in APB signaling, the tumbledown time of the opposing signals is not dependent upon the cascaded times of the signal mechanisms or their slow repeaters. Since the











entrance of a train will release all opposing track relays almost simultaneously, the opposing signals will go to stop at the same time, regardless of the number of intermediate signals in the block.
APB Signal Control
Figure 422 shows a typical APB signaling layout with control limits for opposing and following moves. This particular arrangement of controls is shown here only for the purpose of demonstrating Trakode and may be simpler or more complex than shown. For example, the opposing stop control of entering signal 21, shown in the upper diagram, does not have to reach all the way to the opposing leaving end.
Figure 423 shows a typical track diagram with the signaling arranged for APB operation. The diagrams show successively how the signals are controlled by the polarized Trakode pulses as a train passes from station B to station C. The signal aspects are controlled by the presence or absence of coded energy in the track circuits and by the polarities of the codes as follows:
Aspects Code
red 0
Diagram A shows the normal condition with no train present. All the signals are normally clear for both directions of traffic. All of the pulses for

controlling the signals to green are shown as positive and are transmitted, in the direction of the arrows, alternately from station B to station C, and from C to B. The dependent coding with the repeating action from track circuit to track circuit is normally present in each section, except the detector track circuit at each siding end.
When B transmits a particular pulse to C and it is received, C in turn sends a pulse back to B with the same repeat operation. The polarity of the pulses need not be the same.
Diagram B shows a train at station B, with the block ahead of it clear to C. The dependent westbound coding behind the train has been removed, thus controlling signal 12 to the red aspect. This causes negative pulses to be transmitted in the next track circuit to the rear so that signal 10 is now at yellow, for following moves. Opposing signals 13 and 15 have likewise been controlled to red and yellow respectively.
In Diagram C, the train has moved onto the detector track circuit. The overlap has been arranged so that this removes all eastbound transmission right up to station C, and causes a tumble- down of westbound signals 13, 15, 17, 19, and 21 to red. Signal 21 at red causes negative pulses to be transmitted to signal 23, controlling it to yellow. Station C, after waiting the normal time for pulses to arrive from B, changes to independent coding and continues to send positive pulses toward B, keeping the eastbound signals cleared for the train to proceed toward C.
In Diagram D, the train has accepted signal 14, cutting off westbound positive pulses and placing 14 at red. Signal 12 remains at red rather than changing to a yellow aspect because the train is still in the overlap limits of signal 12 as shown in Figure 422. Note that independent positive pulses are transmitted toward the train from signal 13. The train nearing signal 16 has actuated the approach circuit which picks up the directional stick relay and changes to negative the pulses feeding from signal l6to the train.
In Diagram E, the train has passed signal 16, leaving it at red. Signal 16 transmits negative pulses to control signal 14 to yellow for a following move, and 14 transmits positive pulses to control signal 12 to green. Note that although positive pulses are being transmitted from signal 13 to signal 15, signal 15 does not clear because of the directional stick circuit, which also prevents transmission toward signal 17. The approach-actuated stick circuit at signal 18 changes the positive pulses feeding from signal 18 toward the train to negative pulses.



g 921 23
I2— 14’ 16’—. i8-. 20’-l_
22”k z0Tc

t2—.. 4. I6’— 18’—. ZO’—..
aa.+k. 24—.
—— -
Solid line indicates limit of Stop control.
.———— Dotted line indicotes limit of Approach control.
Figure 422. Simple APB system, showing signal controls.


+— +—, -+ [_-+--, + + I +— F -_÷ +
3 lb I 21 23
I r1i ‘1 I I I ‘1 ‘i
IC 12 ‘ I-r lb 18 20 I—___ 22 24
—f 4 + 14- + - 14— + 4 + 4 + 14 + —l 14- + —l 14— + )4 +
+ +—f-+*{o1 j__- .4 + + ,f-__+-—,{ j-+, —+ +
I l lb 17 21
I I I I’’I •i’I
0 Fcç 12 74 I_t:__ 16 Ir 8 20 I—— 22 24
14 --—F-o-+ —+ 14 + 14 + 14 +— +- f*—+ j +
+ 4 +— F-’ I—o I a o o—4 f—o-- I—- +
l lb 23
r”i bI’I “j i’i I
12 14 lb 8 20 22 24
I --H f-o--j 14_-+ 14 + 14 + 14 +—j G÷-1 #—-—-I +
+ . +—, [—+ F÷+o—I—--o I o o-H —oH f—- . +
-4-j1 r J_15 I—fl? J.47 r i__f 23
i”i i’1 I”I ‘i
70 l2[_r_ l4j- 10hE. l8j 20[r_ 22 I—•—t:__ ?‘
14 --H F-H I-o---14— _+ 14 +—_ +- —+ 14 +
+ “I +— .-+. F-_+ I a I o c—H [—a-I F—- +
a. -i-H -a-H 1 JH2J 23
i’ I I ñ ‘i
70 12 14 16 p-i— IS 20 22 24
+ — + -H + - 14— - —j o —14-o-- --14-- + —1 + 14— + 14
+ i +—, I-+.I ÷ . + + o—H Ho-I f• +
13 19
,:1 r 15 —‘—H 17 _L1 1 r 23
r’i it I I I”I I’I I
ioj— 121.r 4j*. 6[ 1Sf]. 20[T 22r 24f-_
+ [4 +—f I.+—1 f4._+ [4 - 4 ----_I_o_f4.+-j I4--- [4—+ [4 +
+ I +— 1-÷+I }—+ . + f + .4 +—.4 F-OH f—- f +
_ 15 17 __ 23
12 74 16
I•]Z__ 18 20 22 24
I -H F-H I4.—+ [4 + 14 + 4 -—--1 Fo-F-I [4—+ [4 +
Figure 423. Trakode APB signaling.

In Diagram F, the train has passed signal 18, leaving it at red. Signal 18 transmits negative pulses to control signal 16 to yellow, which in turn transmits positive pulses to control signal 14 to green. Independent positive pulses feed toward the train from signal 20.
In Diagram G, the train has moved past signal 20, which thereupon goes to red and allows the negative code in the track circuit to the rear to release the directional stick relay associated with signal 18. The repeat of the pulses through all the track circuits in the block allows dependent coding to be again established upon the reception of the first pulse at signal 19.
Typical Siding
The block diagram in Figure 424 shows the function organization of Trakode when applied to a typical siding. Each siding end has a detector track circuit which serves a two-fold purpose. It provides additional overlap against simultaneous entry and also isolates the two Trakode circuits, thereby protecting the system from improper operation that could result from the breakdown of insulated joints. The signaling is arranged for APB operation in both directions. Signals are three-position, except those governing entrance to double track. These are two-position in this instance for reasons of simplicity, although Trakode is capable of controlling a third aspect if desired. For the same reasons, only two code characters, positive and negative, will be used in the circuit descriptions to follow, although four characters are available.

Briefly, the transmission procedure involves the following steps:
1. The polarity of the pulses to be transmitted is determined by the position of a particular signal.
2. The length of each pulse, and therefore the rate of transmission, is predetermined.
3. The pulse, positive or negative, is generated and applied to the rails from the battery.
4. A time delay is provided between transmission and reception of successive pulses. This is to prevent a possible induced voltage from the rails (rail kick) from operating the track relay inadvertently.
At the opposite end of the track circuit, reception of the pulses takes place thus:
1. The track relay responds to the polarity of each pulse by operating one of its two armatures as indicated by POS. and NEG.
2. The pulse activates a code detecting unit which operates one of the control relays.
3. Either the H or D signal control relay picks up, according to the pulse polarity, and remains up as long as pulses of the same polarity are received.
4. The H relay responds to negative pulses and controls the signal to the yellow aspect, and the D responds to positive pulses for a green aspect.
The simplified circuits in Figures 425 A and B are for a typical siding as shown in the function diagram. It is assumed that conditions are normal, that is, the switches are lined normal for a move straight through on the main track and the signals show the green aspect. Therefore, positive pulses

Figure 424. Function diagram of Trakode.



Nra 3
SIG 12

- ____
Figure 425B. Simplified Trakode circuits for a siding, shown just as West TD picks up and just before West T responds to the pulse from East.




are being dependently transmitted in both directions, and received to clear the two-position signals 12 and 13.
Before describing the operation to clear signals
12 and 13, it may be convenient to list the relays
involved, and their functions:
- Track relay, for reception of the pulses. This is a quick-acting polar-biased relay (Type VTB) having eitherone or two armatures, depending on the code requirements. One armature of a double-armsture Type VTB relay responds to positive pulses and the other to negative.
- Transfer delay relay, which provides a time delay between the transmission of a pulse and the connecting of the track relay to the rails for receiving a pulse. This delay allows the induced rail potential to dissipate through the track ballast and bleeder resistor before the track relay is connected to the rails, so it will not be falsely energized.
- Code pulse relay, which combines with the PL relay to generate the pulses, with polarity as determined by the position of the signal. This is a quick-acting polar- biased relay (Type VTB) with double armatures, if required. When two armatures are used, one armature controls positive pulses and the other negative.
- Pulse length relay, which determines the length of each pulse with the aid of the CP relay. It also determines the length of the off time between pulses with the aid of a capacitor-resistor unit. This is a Type CR code-responsive relay.
- Home and distant signal control relays. Although H relays are not shown in Figures 425 A and B, they are included here for future reference in Figure 427. These are biased-neutral relays which determine the aspect of the signal, the H relay controlling it to yellow, and the D relay controlling it to green.
- OS track repeater relay, which repeats the position of the track relay in a detector track circuit, that is, it is normally picked up unless the detector track is occupied.
The control of signal 13 depends upon the aspect of signal 11, which is shown clear. Figure 425A shows a pulse being transmitted from West to East to control signal 13 to clear. In this instance, the CP relay has a single armature, since only

pulses of one polarity need be transmitted to control a two-position signal, the stop position being absence of code.
The circuit to energize the CP relay is traced from positive energy through:
1. Front contact of OSTP relay, to check that the detector track is unoccupied.
2. Back contact of the track relay T to check that reception is not taking place.
3. Back contact of the pulse length relay for reasons given below.
4. Front contacts of liD and 11H relays in multiple, which determines that a pulse is to be transmitted.
The closing of the CP front contact releases the TD relay and applies energy to PL, which is slow acting and does not pick up until TD has been down long enough to apply a pulse to the rails. When the PL relay picks up, it releases the CP relay, which determines the length of time a pulse is applied to. the rails. Figure 425A shows circuit conditions at the instant TD drops and before PL picks up.
When PL picks up, CP drops and then TD picks up, Figure 4258. The track relay at West is again connected to the rails, ready to receive a pulse from East.
Pulse Timing
The time between pulses is determined by the length of time the PL relay remains energized. It is maintained in this position by the discharge of the capacitor in a local circuit that includes a PL stick contact and a back contact of the track relay. The track relay contact also conditions the PL relay to be ready for transmission after a pulse has been received. Before the capacitor charge is dissipated, the track relay receives a pulse and picks up. (This happens in dependent coding only. If the pulse were not received, the charge would dissipate entirely and independent coding would commence.) The PL relay thereupon releases, allowing the capacitor to recharge. The capacitor and limiting resistor are in parallel with a relatively high resistance, which is adjusted to give the desired time constant.
The release of the PL relay again energizes the CP, which thereupon releases the TD and closes its own front contacts to apply energy to the rails. At the same time, the pickup of CP completes a circuit energizing the PL relay again.
The capacitor charge keeps the PL picked up in the time between the pulses
- the off time. The code pulses are made rather short in duration (a fraction of a second) whereas the time between


successive pulses is much longer (about two seconds) than the pulses themselves.
The intermittent pickup and release of the CP relay is continued as long as the detector track is unoccupied, reception is taking place at the proper time, and signal 11 is at either the green or the yellow aspect.
At the East end of the siding, Figure 425A, the pulses to control signal 13 are received by the track relay T, through front contacts of the TD relay. Each time the T relay picks up, it connects capacitor A in code detecting unit CDU to energy and charges it. Each release of the T relay between transmitted pulses closes a circuit that permits the capacitor to discharge through the winding of 13D relay.
The path of the current under both conditions is illustrated in Figure 426. In the upper diagram the arrows show the direction of current flow when the track relay picks up. The 5000 mfd. capacitor A

Figure 426. Decoding operation.

is charged by a circuit through the track relay front contacts. The previously charged capacitor B, which is 1000 mfd., is discharged through the D relay winding and keeps the relay picked up during the time a pulse is received. The rectifier prevents capacitor B from discharging through a front contact of T relay.
In the lower diagram, the arrows show the direction of current flow in an interval between successive pulses wlien the T relay is released. Capacitor A discharges through the relay winding, the polarity of current being such that the biased- neutral relay is energized to keep it picked up. A parallel circuit charges capacitor B at the same time.
Thus, in the short interval when a pulse is received, the relay is energized by capacitor B and in the longer interval between pulses, is energized by capacitor A. If coding ceases, the capacitors will not be charged and the relay will release. If steady energy is received from the rails inadvertently, the track relay will pick up and stay up. However, capacitor A will not be permitted to discharge, and the D relay will release when the charge of capacitor B is dissipated. This tends to minimize the influence of any storage-battery effect in the track.
The D and H signal control relays used in such decoding circuits are biased-neutral to help protect against the possible effects of breakdown of the capacitor.
Independent Coding
The description thus far has dealt with dependent coding, that is, the track relay must pick up and release the PL relay so the capacitor may be charged for the next transmission, which will start when the track relay releases. However, if a pulse is not received in the predetermined length of time, independent transmission takes place.
Referring to Figure 425 A or B, assume an eastbound train occupies the detector track at the west end of the siding. The OSP relay releases and prevents further operation of the CP relay. Hence transmission stops.
At the other end of the track circuit, the track relay remains released, so the pulse length relay PL is held picked up until the capacitor charge is dissipated. Its release closes the pickup circuit for the CP relay. The transmission operation proceeds as described before, except that the track relay is inoperative and the discharge time of the capacitor determines when the next pulse is to be transmitted.




At the west end of the siding, when the train leaves the detector track the OSTP relay picks up and allows signal 11 to clear. When the train leaves the detector track at the east end of the siding, the H relay is energized, which de-energizes the traffic directional stick relay and allows the CP relay to pick up, to resume coding.
When traffic conditions return to normal, transmission at both ends is quickly brought into step for dependent coding. To facilitate this, the rate of independent coding at one end is different from the rate at the other end, for example, 33 pulses per minute as compared to 29 pulses per minute. The faster rate has an opportunity to be received at the other end and return the system to dependent coding.
Control of Three-Position Signals
The circuits in Figure 427 are a continuation of those for the siding. They will serve to show how both positive and negative pulses are transmitted and received to control three-position signals. For

example, information about signal 13 is transmitted eastward to control intermediate signal 15, and signal 14 is controlled by transmitting information west concerning signal 16.
In this circuit, the CP relay and the track relay have double armatures, one responding to positive pulses and the other to negative. Otherwise the operation is much the same as described previously.
In transmission to control signal 15, the circuit to energize the CP is the same as before except that closing the front contact of signal control relay 130 will operate the positive armature of CP relay, and the release of 13D causes the negative armature to operate. Therefore, when 13 signal is green, a positive pulse is transmitted to control 15 signal to green, and when 13 is red, a negative pulse is transmitted to control 15 to yellow. Occupancy of the detector track circuit will, of course, release OSTP relay, place all opposing signals at stop, and start independent coding as shown in Diagram C, Figure 423.
In the decoding operation to control signal 14, the aspect of signal 16 determines whether a positive or negative pulse will be received, and therefore which armature of the track relay will operate. If the pulses are positive, signal control relay 140 picks up and remains up in the same manner described previously, and signal 14 will be green. If negative, 14H relay is energized, and signal 14 is yellow.
Intermediate Signal Location
Figure 428A shows the circuits for a double intermediate signal location. In addition to the relays discussed so far, others are required for direction control purposes:
- Series approach relay, which repeats the code when a train is approaching, and thereby picks up the proper directional stick relay. This relay has two windings, one for each direction.
- Directional stick relay, with prefix denoting east or west. In picking up, this relay records the direction of a train passing an intermediate signal location.
With the signals clear as shown, positive pulses are being transmitted dependently. A positive pulse received by the track relay WT, for example, applies energy to the P winding of the ECP relay, so that the pulse is repeated into the adjacent track circuit. The reception of a negative pulse will usually re AR


Figure 427. Circuits to control 3-position signals.





suit in transmission of a positive pulse. The impulse is always passed on, but its character may be changed, since this depends upon the position of the signal. The ECP pickup circuit to relay the pulse is taken through back contacts of directional stick relays ES and WS to check that there is no traffic, through back contacts of the ET to check that reception is not taking place, and through the positive (or negative) front contacts of WT to the P (positive) winding of ECP.

This operation is a simple repeat of the pulse received; there is no pulse timing involved in dependent coding. In like manner, the reception of a pulse by ET will cause WCP to transmit through a circuit similar to that described.
When a train enters the block between sidings, the tumbledown feature of Trakode places all opposing signals in the block at stop. Assume an eastbound train enters the block in which the intermediate signals W and E are located and signal W

Figure 428A. Simplified Trakode circuits for an intermediate location, dependent coding.





Figure 428B. Simplified Trakode circuits for an intermediate location, dependent coding. Eastbound train is on West track circuit. Shown just as ES picks up.

therefore goes to stop, Figure 428B. The track relay WT is shunted and the signal control relays WH and WD are de-energized.
As the train proceeds into the track circuit, with WCP (N) energized as shown above, the shunted current pulsing through the E coil of AR increases enough to cause AR to code. The coding operation of the E contacts of AR feeds energy to pick up directional stick relay ES which is then held picked up as the AR (E) contact operates,

because of the capacitor-resistor shunt making it slow release. This ES pickup circuit checks that the opposing directional stick WS is released and that EH or ED is up.
Figure 428B shows the circuit just as ES closes its front contacts. Note that the N winding of WCP is energized, and the P winding of WCP is deenergized. Figure 428C shows the train across both circuits, the N contacts of WCP closed, and a circuit established to pick up PL. When PL picks up,





it cuts off the positive feed to WCP, ending the rail pulse, and opening the pickup circuit to PL. However, PL is slow release and stays up until its capacitor charge is dissipated, at which time it drops, WCP (N) is again energized, and independent coding continues, as previously described. The energy on the track circuit is, of course, shunted out by the train.
As the train entered East circuit, the signal control relay ED released placing signal E at stop. This

established a stick circuit for the ES relay through back contacts of the ED and EH relays and E contact of the AR relay.
When the train is past the location, Figure 428D, the circuit operates to place the signal in the rear at yellow for a following move. With the ES picked up, a circuit is completed to energize the negative coil of WCP relay. The WCP operates to transmit negative pulses westward to control the signal in approach to signal E to yellow. The alternate blue

Figure 428C. Simplified Trakode circuits for an intermediate location, dependent coding. Eastbound train on both circuits, PL up, WCP about to drop.





Figure 428D. Simplified Trakode circuits for an intermediate location, dependent coding. Eastbound train has cleared West circuit, and WCP is coding independently.

and green circuit paths show how this is done. This coding continues until the train clears the block to the east of the location at which time WCP is conditioned to transmit positive code.
With the train out of this block, negative pulses will be transmitted west to control signal E to yellow, and positive pulses will control the next following signal to green. Track relay ET responds to the negative pulses by operating the N contacts. These pulses are decoded in the usual manner, described

previously, so that signal control relay EH is picked up and stays up as long as coding continues. Thus signal E is placed at yellow. When EH picks up, it interrupts the circuit to ES. When ES drops, the code transmitted westward is changed to positive.
Repeating Cut Sections
Where signals are spaced so far apart that the distance between them is greater than the practical operating length of Trakode, it is necessary to



introduce one or more code-repeating cut sections between the signals.
Two track relays having double armatures are used at each cut section. Pulses are received by one track relay on condition that the other is not receiving pulses from the opposite direction. The relay receiving the pulse then operates its positive or negative armature in response to the polarity of the pulse, and in so doing, applies the same battery polarity to the rails in the adjacent track circuit.
With the advent of solid-state electronics it became possible to produce equipment which could efficiently generate power to feed track circuits at higher frequencies, in the audio-frequency range of the a-c spectrum up to 21 kHz. Similarly, solid- state receivers were developed which could detect and amplify for track relay operation the low levels of track energy available due to the high losses in track circuits using such frequencies.
These developments opened up new possibilities for track circuit design. By using frequency multiplexing, more than one track circuit could be

superimposed on a given track section. By utilizing rail impedance and losses in a constructive way as isolating elements, track circuit boundaries could be defined by the points of rail connection of transmitters and receivers without the need for insulated joints. This simplifies installation and reduces track maintenance requirements. It is also corn patible with the use of continuously welded rail and, where electric traction is employed, facilitates the utilization of the rails for propulsion current return.
Track Circuit Characteristics at Audio Frequencies
Track circuits operating in the audio-frequency range have characteristics which differ substantially from those developed earlier. The networks shown for d-c track circuits, Figures 103 and 104 (Section 100), are not complete enough for a-c track circuits, because they omit inductance and capacitance. Figure 429A shows more accurately the circuit elements actually present in a track circuit on single track. Figure 429B illustrates the additional factor of inductive coupling between tracks which occurs in multiple track installations. As track circuit frequency increases, these reactive components of track impedance become dominant factors.

Figure 429. Track impedance networks for high-frequency track circuits.






Inductive reactance increases with frequency and capacitive reactance decreases. So it is not surprising that track circuit characteristics at a-c frequencies are more complex than with d.c. But the situation is even more complicated than might appear. The reactive circuit elements shown symbolically in Figure 429 are not constant. Their values change in a complex way as ballast resistance changes and as different frequencies are applied.
It is not feasible to forecast the characteristics of track circuits at audio frequencies from simple models. GRS has developed computer programs, derived from actual experience, which provide reliable information on track-circuit performance over a wide range of frequencies and ballast conditions. It is necessary to use computer assistance of this type when theoretical analyses are required. For routine purposes, it is sufficient to use the information included in publications covering the application of electronic track circuits developed by GRS.
In general, track energy is attenuated more rapidly at higher frequencies. Thus, when longer track circuits are required, lower track frequencies are used. Where short track circuits are desired, higher frequencies are employed.
As the name implies, frequency-shift overlay (FSO) track circuits may be “overlaid” (i.e., superimposed) on track circuits installed for other purposes, without disturbing the function of the original circuits.
FSO circuits of different frequencies may also be superimposed on each other. Thus two FSO circuits may be overlapped to provide a track section in which both circuits respond simultaneously to occupancy in the overlapped section while responding independently to occupancy outside the overlapped section.
The detailed electronic technology of FSO transmitters, receivers, and related equipment is beyond the scope of this discussion. It is feasible, however, to present the general features of FSO track circuits without such details.
The basic units of FSO apparatus are the transmitter and receiver, shown in Figure 430. The internal functions of these units are indicated in Figure 431.
In the transmitter, the carrier signal generated by the oscillator is frequency shifted at a relatively

low audio-frequency rate by the modulator. The output is fed to the amplifier, where the power level is increased, and through the coupling unit to the track.
At the receiving end of the circuit, the signal feeds through the coupling unit and the input section (a carrier filter) to an integrated circuit amplifier. The amplifier raises the signal level sufficiently to drive the detector. With sidebands (frequency- shifting signal) present, but not otherwise, the detector produces output to operate the relay driver, the output of which in turn picks up the biased-neutral track relay.
The relay connection to local energy is opposite to that normally expected for biased relay operation. This requires the driver stage to produce an output voltage (“super-plus”) higher than that available from the local source, thus protecting against improper relay operation if wire crosses or other conditions occurred which might allow current from the local source to flow from local to corn mon.
FSO Track Circuits For Highway Crossing Warning
Rail-highway crossing warning systems and circuits are described in detail in Section 800, “Highway Crossing Warning.” This discussion will be confined to FSO track circuits as used to provide track occupancy information required by the warning system.
FSO track circuits operate on 18 frequency channels over the range 618 to 20,900 Hz, as detailed in Figure 432. The frequencies are selected to reduce the possibility of interference from harmonics of power line frequencies, or from the combined effects of power line harmonics and

Figure 430. Approach FSO transmitter and receiver.




Figure 431. Simplified block diagrams for FSO transmitter and receiver applied to track circuit.

other FSO equipment. Channels 10 through 24 are normally used for approach track circuits, channels 25 through 27 for island circuits.
Application of FSO approach track circuits is facilitated by inclusion of pluggable coupling units as elements of transmitters and receivers. The coupling unit may be used directly with the transmitter or receiver, or it may be installed independently in a remote case at the end of the track circuit, with connection to the associated transmitter or receiver via line wire. This permits installation of both transmitter and receiver for approach FSO track circuits in the same conveniently located wayside case, operating from a common energy source.

Line wire connections may be as long as 6000 feet for receivers and 3000 feet for transmitters. Line quality should be adequate to avoid interference problems with low-level receiver signals, or cross-talk into adjacent circuits from relatively high-level transmitter output.
Since FSO track circuits are not confined by
insulated joints, it is important that track energy at
a given frequency be attenuated by track losses to
a negligible level before it could reach another FSO
track circuit on the same channel. Figure 432 indicates the separations specified to ensure this.
The frequency indicated for each channel represents center value or average frequency. In actual operation, the frequency of a given channel

1 2V

____________________________ I














Minimum separations between same frequency
(a) 25,000 ft. (b) 20,000 ft. (c)
(d) 8,000 ft.
(e) 7,000 ft. (f)
Figure 432. FSO frequencies. Channels 10-24 for approach circuits, channels 2 5-27 for island and series overlay circuits.

shifts continuously between a value four percent higher and four percent lower than the nominal channel frequency. This shift occurs 13 times a second on even-numbered channels and 15 times a second on the odd-numbered channels. The frequency shift is an essential element of the track circuit. It provides protection against foreign audio- frequency currents which might appear in the rails, and against possible undesirable interaction of FSO circuits on adjacent channels.
Track circuit length feasible with FSO track circuits depends on frequency and ballast resistance. With ballast at 5 ohms per 1000 feet, approach track circuits as long as 6000 feet may be operated with channel 10 or 11. Feasible length decreases with frequency, reducing to 1400 feet for channel 24. The highest frequencies, channels 25, 26, and 27, are applied to island circuits ranging from 60 to 300 feet.
Figure 433 illustrates the use of two overlapped FSO track circuits for crossing protection. A northbound train first shunts the rails between transmitter fl and receiver fi, releasing the fl track relay. The train next shunts the rails between receiver f2 and transmitter f2, releasing the f2 track relay, with fi relay remaining down. When the shunt moves beyond receiver f 1, fl OTR picks up but f2 OTR remains down. Finally, with the shunt beyond the f2 transmitter, f2 OTR picks. A southbound train reverses the operation. This sequence, with the aid of auxiliary circuits described in Section 800, provides the information necessary for two- direction approach warning at the crossing.

A train shunt is always effective when it is between the transmitter and its receiver. Since the FSO circuits are not terminated by insulated joints, a train shunt is also “seen” by the circuit for a limited distance beyond the connection points. These pre-shunt and post-shunt distances vary with frequency, becoming negligible at higher frequencies. If a more precisely defined island segment is desired, three FSO track circuits are used, Figure 434. In this arrangement, one of the high-frequency channels 25-27 is typically used for island circuit 2T, while lower channels are used for approach circuits iT and 3T. In special circumstances, such as an installation requiring an unusually long island section, a channel lower than channel 25 may be used for the island circuit.
Series FSO Track Circuit for Switch Lock Release
Overlay technology is also employed to provide a short detector track circuit on the immediate approach to a switch, without the need for installing insulated joints. The circuit produces lock release with minimum delay for a train wishing to enter the turnout.
A series FSO track circuit is used with the track relay normally de-energized. The relay picks up to provide lock release when a train shunt completes the circuit between the rails. This is consistent with the signaling principle that a de-energized relay results in a restrictive condition, in this case a locked switch.

on same track:
10,000 ft.
6,000 ft.














































Figure 433. Two FSO track circuits overlapped to provide three indications of occupancy.





Figure 434. Island FSO track circuit with two approach FSO track circuits.







- —





An overlay transceiver, Figure 435, is used, operating on one of the channels 25 through 27 tabulated in Figure 432. The transmitter section of the transceiver generates the channel carrier and modulates it by frequency shift at 330 Hz. The receiver section extracts the modulation from the carrier and produces output to operate the track relay. If the carrier and modulation are not present, the relay remains de-energized.
Figure 436 shows the general organization of a series overlay installation. The transmitter section of the transceiver is connected to one rail of the main track and the receiver to the other rail, 75 feet in approach to the switch points. When a train shunt appears across the rails in the immediate vicinity of the track connections, overlay energy feeds through the shunt to the receiver. The track relay then picks up, permitting switch lock release.
Maximum detection range for pickup does not exceed 50 feet on either side of the rail connection. Note that the rail jumper connections and insulated joints shown for the turnout are required for the underlying block system track circuit, and are not part of the overlay installation.
Track frequencies in the audio range are not restricted to application in overlay circuits. They may be used also for basic track circuits in a block

signal system, and are usually referred to as high- frequency track circuits.
High-frequency track circuits are suitable for general application. They are particularly well

Figure 436. Series FSO track circuit for switch lock release.

Figure 435. Series frequency-shift overlay transceiver.


>____•,•_•j— TO SWITCH LOCK




adapted to electric rapid transit systems, where the maximum track circuit length of 2000 feet is adequate, and where the elimination of insulated joints simplifies the use of the rails for propulsion current return. With welded rail there is the additional benefit of eliminating breaks in the continuous rail structure.
A typical installation uses eight distinct track frequencies, Figure 437. These are on-off coded to transmit block information in essentially the same way as for other rate-coded track circuits. The specific carrier frequencies are selected for non-interfering compatibility with each other and with any a-c component of propulsion energy.
The basic arrangement of a high-frequency track circuit is shown in Figure 438. The transmitter generates energy at frequency fi which is carried over a twisted pair transmission line to a WEE-Z® bond (described later).
The bond applies fi to the rails, which transmit the energy to a similar WEE-Z bond at the other end of the track circuit. The receiving bond in turn couples fi to a line which carries it back to a re CARRIER

Figure 437. Typical carriers for a block high-frequency track circuits.

system using

ceiver. With fi energy received (coded as may be required), the track relay is energized. A train shunt between the WEE-Z bonds prevents energy from reaching the receiver and the track relay releases.
Line lengths up to 6800 feet may be used between WEE-Z bonds and associated transmitters and receivers. This makes it possible to centralize electronic equipment for many track circuits at a common location.



Figure 438. Basic high-frequency track circuit configuration.



L — — — a — a — a — eJ
























WEE-Z Bond
The WEE-Z bond, Figure 439, includes a two- turn ioop, formed from copper bar stock, which is connected from rail to rail. The loop is threaded through toroidal inductors tuned to parallel resonance by capacitors housed in the tuning unit.

assigned sequentially to one track and odd frequencies to the other.
Block 1 operates at fi with bond Z2 coupling fl to the track. Simultaneously bond Z2 serves as receiving coupler for f7 used for block 2. The twisted pair line circuit to Z2 thus carries both fl and f7.
Blocks 3 and 4 use f5 and f3 respectively. At block 5 the sequence of odd-numbered frequencies begins to repeat. In each case the WEE-Z bond at the block boundary is tuned to the two frequencies associated with the adjacent blocks on either side.
Similarly, the track circuits on the parallel track are assigned sequentially to f2, f4, f6 and f8, then repeat. At intervals determined by local conditions the tracks are crossbonded through WEE-Z bond center taps. If a power return takeoff were located in this section of the system, it would also be connected to a bond center tap.
Insulated joints are not necessary. The sequential assignment of track frequencies places at least three block lengths of rail and two low-impedance WEE-Z bond shunts (because tuned to other frequencies) between track circuits on the same frequency. This ensures that crosstalk between such circuits is essentially zero.

Figure 439. Typical WEE-Z bond installation.

Since the toroids are closely coupled to the two-turn loop, the impedance across the loop terminals is much higher at the resonance frequencies of the tuned toroids than at other frequencies. With the bond tuned for resonance at two frequencies it can act simultaneously as the receiving coupler for one track frequency and as the transmitting coupler for another. Additional tuned toroids may be included for other frequencies to be transmitted or received over the rails.
In addition to functioning as a coupling unit, the WEE-Z bond serves to equalize traction current in the rails since the loop acts as a short circuit between the rails at traction frequencies. A center- tap on the two-turn loop also provides a connection point for power return cables of the propulsion system, and for crossbonding of parallel tracks.
Application of High-Frequency Track Circuits
The multi-frequency capabilities of the WEE-Z bond are utilized in applying high-frequency track circuits over extended track sections as indicated in Figure 440. To eliminate the possibility of crosstalk, eight frequencies are used. In the two-track example shown, even-number frequencies are

The coded track circuits described earlier in this section provide two distinct functions: (1) detection of train occupancy; (2) block-to-block communication through the rails for purposes of signal control. Using the rails for communication eliminates line wires and their costs.
Line wires can also be eliminated by applying rate-coded frequency-shift overlay corn mu nication circuits to the rails, superimposed on d-c track circuits which provide occupancy detection.
Although the coded overlay communication circuits are carried in the rails, they are not required to meet sensitivity standards for train shunt detection, which remains the function of the underlying track circuit. Thus the overlay communication circuits are not track circuits in the usual sense. But they have much in common with FSO track circuits, thus it is appropriate that they be included with the discussions of electronic track circuits.
Equipment and Application
FSO communication circuits use the six audiofrequency channels 10 through 15 listed in Figure
432. Communication range for a transmitter/re-




ceiver pair depends on track characteristics and channel frequency. Maximum range with 5-ohm ballast is 8900 feet for channel 10 on 127-lb rail, decreasing to 5100 feet for channel 15 on 90-lb rail. Range may be extended by insertion of repeating transceivers between terminals. If FSO track circuits for highway crossing protection are installed on the same track carrying FSO communication circuits, channels different from the communication channels are used.
Figure 441 shows the equipment used for FSO communication. Figure 442 illustrates the arrangement to provide a desired communication link.
The transmitter unit includes code-rate generators and audio-frequency circuitry. The signal system relay logic selects the appropriate code and feeds it to the frequency-shifter, which causes the transmitter output to shift frequency at the code rate.

At the receiver, the incoming carrier is admitted through a filter, amplified, and demodulated to extract the code. The code is further amplified and fed to decoding units, one of which responds to energize a decoding relay.
If insulated joints intervene between transmitter and receiver, or if transmission distance requires, a transceiver is added as shown in Figure 442. This accepts coded signal at one frequency, extracts and reshapes the code pulses, then remodulates a carrier on a different channel for application to the track. The transceiver provides bridging around insulated joints, compensates for circuit losses, and by channel-to-channel translation protects against interference problems that might arise.

zi f3 fi
Z2 f1 f7
Z3 i
Z4 5
Figure 440. Frequency assignment sequence for high-frequency track circuits.



C. Transceiver D. Decoder
Figure 441. Equipment for FSO communication circuits.

A. Transmitter

B. Receiver











rXMm I1





Figure 442. Arrangement of FSO communication circuit.