TECHNICAL FIELD

This invention relates to a timing and scoring system and more particularly to such a system for use with multiple moving stations traveling on either a closed or open cir¬ cuit.

BACKGROUND ART

Many events, both competitive and non-competitive, involve multiple moving stations traveling on either a closed circuit or open circuit. For much of the remainder of the description of the invention herein, the point of view will be taken of multiple race cars traveling on a closed circuit in a competitive event. Via this point of view, descriptions herein will be more clear for the reader. However, the taking of this point of view in no way re¬ stricts the applicability of the invention to other situa¬ tions which do not directly involve automobiles or closed circuit courses or competitive events. Readers familiar with the art will be able to unambiguously extrapolate the descriptions herein to cover other situations. Furthermore, references to a single fixed reference line that is crossed repeatedly on a closed course do not detract from the viability of having multiple fixed reference lines on either a closed or open course, crossed only once or a multiple of times.

While developed for use in racing cars using radio frequency technology, this invention has application for both competitive and non-competitive events involving moving stations that are not restricted simply to automobiles and via communication means other than radio frequency.

During a car race event on a closed course, each time a car crosses a fixed reference line (the "timing" line or "start/finish" line) the identity of the car and its exact time of crossing are noted. This fundamental information may be used to produce compiled information which is highly desirable by essentially everyone at the event. Compiled information includes among other things the lap time, number

of laps completed, track position, and race position. Lap time refers to the time interval between crossings of the timing line, and may be generalized to segment times between different fixed reference lines. Frequently, the best (shortest) lap time achieved by an individual car is of interest. Laps completed refers to the number of times that the timing line has been crossed since the beginning of the event, typically not counting the initial crossing. Track position refers to the relative order of cars distributed around the track and serves as an indicator as to the order in which cars might be expected to cross the timing line again. Race position refers to the track position informa¬ tion reorganized by number of laps completed, such that the car in first position has managed to complete its superior number of laps earlier than any other car and is likely to be declared the "race winner".

Race officials, crew personnel and sponsors, specta¬ tors, and the like may be considered to be in a "fixed" location and collectively associated with the fixed station (base station) at the reference line. For brevity, all of those persons will be implied when only Officials are mentioned explicitly. Race car Drivers, on the other hand, are moving and individually associated with their own moving station (race car) . While both Officials and Drivers typically want access to this compiled information live during a racing event, traditionally they have been required to employ completely separate systems to gain access to it. Officials have traditionally used both manual and automated "Timing and Scoring" systems, while Drivers have tradi¬ tionally used "On-Board Lap Timing" systems. This invention provides for the needs of both Officials and Drivers simul¬ taneously at a cost less than or equal to the greater of the two traditional methods.

Traditional Timing and Scoring Systems

Traditionally, Officials use a "Timing and Scoring" system which notes the identity and exact time when each car crosses the timing line. From this information, a central computer is normally used to compile either qualifying or

race results. Qualifying results present a list of cars ordered by the best lap time achieved by each car. Race results present a list of cars ordered by race position as of the completion of the event.

Automated timing and scoring systems must at a minimum 1) recognize the identity of a car crossing the timing line, and 2) determine the exact time when this crossing occurred. In order to identify the car, minimal one-way communication from the moving car to a fixed base station is required, this communication including information to uniquely identi¬ fy the car. Typically (but not always) , the exact time when the crossing occurs is deduced from the exact time when the car's identity is determined, this deduction being reason¬ ably correct by virtue of system design. Therefore, both pieces of information required by the automated timing and scoring system are typically based on the minimal one-way communication from the moving car to the base station. Even when additional communication or other methods are used for timing, the minimal one-way communication is still a re¬ quirement for car identification.

In a primitive system using strictly one-way communica¬ tion, a moving station may continuously broadcast its identity regardless of track location. A highly directional or localized receiver at the base station is required in order to restrict its attention to cars actually within the vicinity of the timing line. Problems with such systems include substantial power consumption by each car's trans¬ mitter plus severe unintentional interference problems caused by many transmitters broadcasting simultaneously.

Many traditional timing and scoring systems use a form of bi-directional communication to overcome these problems. For the purpose of the description herein, this class of systems is referred to as "interrogation/response" systems, the on-board component in each car is referred to as a "transponder", and the fixed component at the reference line is referred to as the "base station". In these systems, as a car carrying a transponder nears the timing line, it comes within reception range of an interrogation signal fixed at

the locale of the timing line. In response to this inter¬ rogation signal, the transponder transmits information which allows (at a minimum) the unique identification of the transponder.

Interrogation/response systems take many forms. Many use a magnetic field as the interrogation signal, where this field is established by a loop buried in the track surface. The transponder may react to this field in one of two generally practiced ways in order to uniquely identify itself. 1) It may passively alter or reflect the interroga¬ tion field, or 2) it may actively transmit an RF signal. Other interrogation/response systems use explicit RF commun¬ ications in both directions. In all cases, the nature of the two-way communication is that of an interrogation eliciting a response which may be uniquely identified.

The major obstacle to successful operation of an inter¬ rogation/response timing and scoring system involves recog¬ nizing the individual identities of multiple cars crossing the timing line nearly simultaneously. The reception aperture used by the base station to receive transponder responses is necessarily finite (non-zero in time) . As a result, there will be a certain window of opportunity for two or more transponder responses to overlap and possibly interfere with each other. This window corresponds to a distance margin between cars at a given speed as well as a time margin between cars. Unless this possibility of interference is directly addressed and eliminated, a timing and scoring system can not be successful.

Traditional interrogation/response systems use either frequency or time multiplexing to avoid interference. In frequency multiplexing, each transponder is assigned a unique radio frequency on which to respond. The base station must be capable of simultaneously receiving multiple responses on different frequencies. This solution is costly and complex. The transponder design must be such that each one can be configured to a different frequency. The base station design must be such that each frequency (or at least many of them) may be received simultaneously.

In time multiplexing, costly collision avoidance tech¬ niques must be used to insure that two or more transponders in simultaneous need of responding may do so and be reliably detected by the base station. Depending on the method of collision avoidance, the time available for a single trans¬ ponder response (and thus the minimum data rate allowable) may be restricted to the point of adversely increasing system cost or decreasing system reliability. Active colli¬ sion avoidance schemes require the cost of an additional receiver inside the transponder so that it may listen to its own broadcast to check if interference occurred. Otherwise, randomness factors may be used, which themselves increase transponder cost and only probabilistically increase system success.

Traditional On-Board Lap Timing Systems

Traditionally, Drivers are very interested in having an on-board display which indicates their most recent lap time as well as their number of laps completed. Typically, the display device also includes memory so that this same infor¬ mation may be recalled, printed, or downloaded to a personal computer afterwards. Many such lap timing systems exist on the market and most of them use infrared technology. An infrared transmitter is placed at the side of the track, typically near the timing line, and its beam is directed across the track surface. A corresponding infrared receiver is mounted in the car and aimed appropriately. As the car travels around the track, it will suddenly cross into the infrared beam of the transmitter. The receiver will detect this beam and provide a trigger to a timing and computing device. The timing and computing device compiles the information at hand in order to produce a lap time and a lap count. This compiled information is then displayed on the on-board display and possibly stored into memory.

It is important to note in this context that the communication involved in a traditional lap timing system is one-way, from the transmitter at the side of the track to the receiver on board the car.

A significant problem with traditional infrared on-

board lap timing systems is missed laps due to the direct line-of-sight requirement for infrared. If a second car obstructs the infrared beam from the side of the track, the first car's receiver will not detect the beam and will miss the lap. An inconvenience with these systems is that each Driver is responsible for placing their own infrared trans¬ mitter at the side of the track, if there isn't already one of the same brand and encoding present. This leads to missed sessions due to lack of time or attention in setting up a transmitter, lost transmitters after an event due to theft and forgetfulness, as well as an unsightly and possi¬ bly dangerous plethora of small tripods and transmitters crowded around the start/finish line.

DISCLOSURE OF INVENTION

An object of the invention is to provide a timing and scoring system which is inexpensive, even when thousands of unique transponder ID's are required, through the use of a fixed and shared frequency scheme, yet very reliable through the use of redundant lap time information contained in each transponder response, based on the concept that the trans¬ ponder is able to self-determine its lap times using strict¬ ly the signal from the base station.

It is a further object of the invention to increase the reliability of timing and scoring system designs through the use of redundant lap time information contained in each transponder response, even when means other than a fixed and shared frequency scheme are used.

It is a further object of the invention to provide a timing and scoring system which is both inexpensive and portable, using a portable yagi antenna instead of a loop antenna cut in the surface of the course.

It is a further object of the invention to provide a timing and scoring system which makes available to officials and others in a fixed location not only lap times, but also segment times.

It is also an object of the invention to provide an on-board display of lap count and lap time for view by each driver, cost effectively using hardware already required for

the purpose of performing timing and scoring for officials, without adding additional communication capabilities to this hardware, without necessarily taking advantage of the other objects of the invention, being applicable to traditional timing and scoring systems which already operate on an interrogation/response basis by virtue of adding self- determination of lap times to an existing transponder design, these traditional timing and scoring systems includ¬ ing but not being limited to systems using a loop antenna cut in the surface of the course.

It is a further object of the invention to provide segment times on the on-board display by encoding different data on the signal produced by different base stations.

It is a further object of the invention to provide additional information on the on-board display, including race and flag status as well as individualized messages directed to a single or select group of on-board displays via the unique ID of the associated transponder by encoding additional information on the signal produced by one or more base stations.

It is a further object of the invention to provide a memory within the on-board display for the purpose of allowing after-the-fact recall, printing, and downloading of the information gathered while on the course.

Additional objects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objects and advantages are realized and attained by means of the instru¬ mentalities and combinations particularly pointed out in the appended claims.

For a plurality of moving stations traveling on a closed or open course and crossing one or more fixed refer¬ ence lines, this invention simultaneously provides a combin¬ ation of centralized and distributed information. It also uses redundancy to solve a traditional interference problem when multiple moving stations cross the same reference line nearly simultaneously. Performance of either or both of

these tasks is performed with reduced system cost and complexity.

A base station is located in the vicinity of each reference line and includes a means for transmitting a highly directional interrogation signal along the reference line such that moving stations cross through this signal. Each moving station carries a transponder which receives and recognizes this interrogation signal. While in traditional systems the base station is responsible for determining the exact time when the moving station's transponder crosses the fixed reference line, this invention gives the transponder this responsibility. This self-determination may use one of several methods already established for base stations, in¬ cluding but not limited to methods based on signal strength pattern. A fixed delay (possibly zero) after this self- determination, the transponder responds by sending a signal back to the base station. It also stores into its own memory information about when the interrogation occurred. The signal transmitted back to the base station includes a unique code identifying the transponder. It may also include redundant information about when a number of pre¬ vious interrogations occurred. The base station receives the response from each transponder and uses the timing of the response to determine exactly when the identified moving station crossed the local reference line. Any fixed delay in the transponder responses will be common to all trans¬ ponders and will have no adverse effect on the information gathered by the base station.

The invention provides for an optional display on board each moving station. This display provides information about when each reference line is crossed. Specifically, a lap count and recent lap time may be displayed. The ability to provide this information on board the moving station depends on system capabilities already required of an interrogation/response system: bi-directional communica¬ tion. However, since this invention makes the transponder responsible for self-determining its time of crossing, this knowledge is now on board the moving station and

conveniently available for display. If additional informa¬ tion is encoded with the base station's interrogation signal, it may be displayed as well.

This invention also provides for a redundancy method. If the response from two or more transponders overlap in time and therefore interfere with each other, all or part of one or more of the responses may not be successfully re¬ ceived by the base station. These laps would be missed without the method of this invention. However, with redun¬ dant lap time information encoded in each transponder's response, successful receipt by the base station of sub¬ sequent responses will allow reconstruction of the missed laps. The amount of redundancy in each response required to achieve a target reliability is a function of the duration of the response, the nature of possible interference, and the relative motion and interaction of the moving stations. (See Fig. 9) If each transponder response contains a log of the N most recent lap times (possibly excluding the current one) , then the base station will have N+l opportunities to gather the necessary information for a single crossing of the timing line. The first opportunity is upon receipt of the original transponder response. The N additional oppor¬ tunities are by inspecting the redundant log of prior crossings contained in responses received subsequently. Statistically, cars will very frequently interfere with each other. At the same time, they will very frequently not interfere. Through a judicious selection of the redundancy factor N, the probability of actually missing a lap may be reduced to an arbitrarily small number. In practice, the system's maximum data rate is restricted by cost and engin¬ eering factors. Even though increasing N will increase the duration of a transponder's response and thus increase the window of opportunity for interference, the resulting decrease in the probability of missing a crossing is over¬ whelming. (See Fig. 9) As a result, failure rates below one in a million may be theoretically achieved. As a final note, it would be impossible to reconstruct information from an interference that occurred on the very last lap.

However, races are always followed by cool down laps and then by the cars leaving the track. A simple restriction that cars exit the track single file (past the same base station or an auxiliary one) insures that successful data collection will occur.

Each of these features, the on-board display and the redundant response encoding, depends on the transponder's self-determination of when it crosses each reference line. These two features are independent from one another and may be implemented individually or together.

Summary of Cost and Complexity

This invention successfully merges the capabilities of two traditional systems: timing and scoring systems and on-board lap timing systems. By merging these capabilities the information highly desired by two disparate classes of people. Officials in a fixed location and Drivers aboard moving stations, may be provided at significantly reduced cost and complexity.

When reduced to its most simple components, the tradi¬ tional timing and scoring system consists of two "black boxes" , a track-side base station and an on-board trans¬ ponder. Meanwhile, the traditional on-board lap timing system consists of three "black boxes", a track-side infra¬ red transmitter, an on-board infrared receiver, and an on-board display. Since the actual cost of many electronic devices has more to do with their packaging than their function, counting them up as "black boxes" is a valid method of roughly estimating not only the cost but also the complexity of a system. In this invention, the track-side base station and the track-side infrared transmitter merge to become a single "black box". Meanwhile, the on-board transponder and the on-board infrared receiver merge to become a single "black box" as well. Finally, the on-board display remains its own single "black box". As a result, five "black boxes" have been reduced to three, resulting in significant cost and complexity savings.

Nevertheless, "black boxes" may sometimes have signifi¬ cantly different costs or complexity. In the case of this

invention, however, this is an advantage. The interroga¬ tion/response class of timing and scoring systems already requires two-way communication capability in both the base station and the transponder. The one-way communication need of the on-board lap timing system may be superimposed on the timing and scoring system with essentially zero overhead. Specifically, the transponder simply makes a note of each time it was interrogated, effectively enabling the on-board lap timing function. Traditionally, timing and scoring components cost significantly more than on-board timing components. In this invention, these components are merged with zero overhead beyond the traditional timing and scoring system cost. Furthermore, the redundancy method of this invention allows the cost and complexity of the transponder to be further lowered, making it closer to the low level of traditional on-board lap timing receivers.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory but are not restrictive of the invention.

BRIEF DESCRIPTION OF THE DRAWINGS

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate a preferred embodiment of the invention and, together with the description, serve to explain the principles of the inven¬ tion.

Figure 1 is a diagrammatic plan view of a representa¬ tive section of course 1 and a block diagram view of an assortment of system types. Three reference lines are shown 2, 3, 4 with three base stations 20, 21, 22 and one host computer 30. Examples of different types of base station antennas 50, 51, 52 and different configurations of moving stations 10-15 are included. Various transponders 30-35 and on-board displays 42-45 are illustrated;

Figure 2 is a diagrammatic plan view of a section of course 1 and representing traditional timing and scoring communications. It includes a loop antenna 52 buried in the

course surface near a reference line 4, a nearby base station 22 and host computer 30, plus two moving stations 14, 15, each carrying a transponder 34, 35, respectively;

Figure 3 is a diagrammatic plan view of a section of course 1 and representing traditional infrared lap timing communication. It includes an infrared transmitter 23 with pattern 24 near a reference line 3, plus a moving station 12 carrying an infrared receiver 32 and an on-board display 42;

Figure 4 is a diagrammatic plan view of a section of course 1 and representing the invention's method of commun¬ ication. It includes a host computer 30 and a base station 21 with an antenna 51 having a directional pattern 56 directed along a reference line 3, plus a moving station 12 carrying a transponder 32 with pattern 36 and an on-board display 42;

Figure 5 is a diagrammatic plan view of a section of course 1 along with a directional antenna 51 producing a pattern 56 along a reference line 3. The detailed signal field strength along the length of the course is plotted 56b. Threshold levels 57, 59 and a peak 58 are indicated;

Figure 6 is a diagrammatic perspective view of a transponder 32 with a half-dipole antenna 37. The antenna pattern is indicated at 36b, c, d to be somewhat doughnut shaped in Figure 7 and Figure 8 wherein Figure 7 is a side view of transponder 32 and antenna 37 and Figure 8 is an end view of transponder 32 and antenna 37;

Figure 9 is an example timing diagram of data Wl to be FSK modulated W2 by a base station 21 and then sent to a transponder 32 where it is received and demodulated into a facsimile W3 of the original data;

Figure 10 is an example timing diagram of data W4 to be represented by a "missing pulse" scheme and FSK modulated W5 by a transponder 32 and then sent to a base station 21 where it is received and demodulated, passed through a first monostable multivibrator to produce an interim waveform W6 and then passed through a second monostable multivibrator to produce an eight microsecond delayed facsimile W7 of the

original data. Note that the two monostable multivibrators form data shaper 77 of Figure 13.

Figure 11 is a block diagram of a transponder;

Figure 12 is a block diagram of an on-board display. Included are an example lap time report and printer cable specifications;

Figure 13 is a block diagram of a base station;

Figures 14 and 15 illustrate the electrical schematic for the preferred embodiment of a transponder;

Figures 16 and 17 illustrate the electrical schematic for the preferred embodiment of an on-board display; and

Figures 18, 19, and 20 illustrate the electrical schematic for the preferred embodiment of a base station. BEST MODE FOR CARRYING OUT INVENTION AND INDUSTRIAL APPLICABILITY

This invention may be implemented in each of the configurations indicated in Fig. 1, including but not limited to the use of a highly directional antenna 50, 51 or a buried loop antenna 52, with or without on-board displays 42-45 and with or without redundancy in the ID packet.

The preferred embodiment uses the configuration indi¬ cated in Fig. 4. A base station 21 is located near a selected reference line 3 across the course 1. The base station antenna 51 is a yagi type antenna with a highly directional antenna pattern 56. The base station continu¬ ously transmits a 926.5 MHz FSK modulated signal 56 across course 1 via antenna 51. This FSK modulated signal contains repeating encoded data identifying the base station and is called a "ping". (See the section on Encoding herein) . In general, each base station is configured to repeatedly transmit a single encoded 0 nibble, indicating that it is a primary station. Additional secondary base stations may transmit other encoded nibbles or sequences of nibbles.

A moving station 12 is depicted on course 1 in the vicinity of reference line 3. This moving station carries both a transponder 32 and optionally an attached on-board display 42. Transponder 32 is continuously listening for a

base station ping 56. When it detects a signal whose strength is above a required threshold (57 of Fig. 5) , it begins looking for FSK modulated encoded data on that signal. If the signal is legitimate, then the encoded data will be successfully decoded and possibly error corrected to indicate base station 21 as either primary or secondary.

Meanwhile, the signal strength of detected ping 56 is monitored until it reaches a peak (58 of Fig. 5) or other¬ wise fails to increase, for example, by more than 5% in a 25ms window. This behavior constitutes the self-determina¬ tion of the exact time of crossing. At 75ms after the exact time of crossing, transponder 32 transmits an "ID packet" at 915.8 MHz, consisting of FSK modulated data at 62,500 bits per second. The laterally oriented doughnut shape of the transponder's antenna pattern 36 (detailed in Figs. 7 and 8) compliments the base station's antenna pattern 56 and assists in preventing the ping from being detected pre¬ maturely, especially when other moving stations ahead of this one reflect that ping. At the same time, it insures that base station antenna 51 can properly receive the transponder's signal 36.

Transponder 32 further reacts to the ping by making a note of its microprocessor (91 of Fig. 11) internal clock. If base station 21 is a primary one sending out an encoded 0 nibble, then transponder 32 will subtract the value of the clock from a previously noted value in order to calculate a lap time. This lap time is in turn stored in the trans¬ ponder's microprocessor memory (90 of Fig. 11) . The current value of the clock then becomes the noted value for next time a ping is detected. Note that a minimum lap time of three seconds is enforced in the preferred embodiment. The ping signal strength must fall below the threshold (59 of Fig. 5) for at least three seconds and rise back above it (57 of Fig. 5) in order for a new ping to be recognized. This avoids multiple triggering in situations where the base station antenna pattern 56 is significantly disturbed by fixed and/or moving reflections, or when a moving station

stops near the reference line.

In general, transponder 32 will transmit an ID packet when successfully detecting either a primary 21 or a second¬ ary base station's ping. However, transponder 32 will only increment its internal lap count and update its ID packet if the ping is from a primary base station 21. In this way, primary base stations 21 are intended to be located at true reference lines 3, while secondary base stations are in¬ tended merely for additional data collection services. The use of secondary base stations can further decrease the probability of completely missing a lap.

When transponder 32 transmits its ID packet, base station 21 receives this packet, decoding and possibly correcting the received data. Either base station 21 or host computer 30 next analyzes this packet along with prior packets received from the same transponder 32 in order to construct a log of exact lap times and lap counts. This information is in turn the basis for more sophisticated timing and scoring reports.

If two or more transponders 32 et al. transmit ID packets at the same time, then interference will result and base station 21 may not successfully decode and recognize the packet. In this case, a subsequent ID packet must be received from each transponder 32 not recognized, and the redundancy in that packet must be used to reconstruct the missing lap. Note that the ability to provide redundant lap time information in the packet does not depend on the use of an on-board display and may be implemented without a dis¬ play.

Finally, the moving station also carries an on-board display 42. The invention's on-board display 42 does not require the redundancy described for the transponder ID packets and may be implemented without this redundancy. Instead, it only requires the self-determination by trans¬ ponder 32 of when it crossed reference line 3. At the same time that transponder 32 determines that the peak (58 of Fig. 5) of a primary base station ping 56 has been detected,

transponder 32 sends a signal along a wire to on-board display 42. The on-board display reacts by displaying a lap time and other useful information. On-board display 42 of the preferred embodiment stores at least ninety-nine laps instead of the eight logged by the transponder itself. After the fact, on-board display 42 allows those laps to be recalled on the display, printed to a printer, or downloaded to a personal computer. While the preferred embodiment does not implement this feature, additional information may be encoded on the base station ping for the purpose of display on selected or all on-board displays. Also, while transponder 32 and on-board display 42 may be built as a single unit to further save on cost, sharing their processors and enclosures and many other components, the preferred embodiment implements them as completely separate units connected by a signal wire and a common.

Encoding

The preferred embodiment uses a nibble encoding scheme. For each and every eight-bit byte received, a four-bit nibble of data is included. The additional bits may be used to detect up to two bit errors or correct up to one bit error. Encoding and decoding may be performed with a trivial table lookup when using and eight-bit micropro¬ cessor. The simplicity of this scheme along with its high degree of error detection and correction ability, allow superior communication performance in adverse conditions with less expensive equipment.

Nevertheless, the preferred embodiment requires that synchronization bytes be received error-free. These bytes are outside of the encoding scheme and would otherwise appear to be bad data bytes that can not be corrected. This allows the packet's data content to be relatively trans¬ parent to its packaging. Table 1 is an encode table. The "IN" column contains all hexadecimal nibble values from zero (0) to sixteen (0x10) . The "OUT" column containε the resultant encoded eight-bit byte. The first four bits of this byte equal the original "IN" nibble. The remaining

four bits are selected in order to enable single-bit-error- correction and double-bit-error-detection. In order to encode a nibble, locate the desired nibble in the "IN" column and encode it by selecting the number to the right of it in the "OUT" column.

Table 2 is a decode table. The "IN" column contains all hexadecimal byte values from zero (0) to two hundred and fifty-five (OxFF) . The "OUT" column contains a lookup result consisting of a high order flag nibble and a low order data nibble. Flag nibble bits are as follows. Bit 0x80 indicates that the input byte can not be corrected, too many bit errors have occurred. Bit 0x40 indicates that the input byte may be a "SYNC1" byte (0x00) , which is otherwise outside of the nibble encoding scheme. Bit 0x20 likewise indicates that the input byte may be a "SYNC2" byte (0x03) , which is also outside of the nibble encoding scheme. Bit 0x10 indicates that there were no bit errors at all in the input byte.

Decoding proceeds as follows.

Example 1: If the byte 0x27 is received, this byte is used to address the table of "OUT" values, seen in Table 2 next to the 0x27 "IN" value. The resultant "OUT" value is "0x80", indicating that 0x27 can not be corrected.

Example 2: If the byte 0x08 is received, the resultant "OUT" value is 0x41. Since the 0x40 bit is set, the 0x08 byte received might be a "SYNCl" byte (0x00), although it must have had a bit error because the 0x10 bit is not set in the "OUT" value of 0x41. If the "SYNCl" byte is being sought with no bit errors, then the received 0x41 will not be accepted. On the other hand, if data is being expected, the 0x41 "OUT" value corresponds to the nibble 0x1. In this case, a bit error still must have occurred because the 0x10 bit is not yet.

Example 3: If the byte 0xC9 is received, the resultant "OUT" byte is OxlC. Since the 0x10 bit is set, no bit errors occurred at all and a perfect OxC nibble was received.

Table 1 - Nibble Encode Lookup Table

Fig. 9 indicates the method used by the preferred embodiment to further encode data for the purpose of FSK modulation and RF transmission.

Base station 21 transmits data at 2400 baud using a traditional asynchronous format, including a start bit, eight data bits, and a stop bit. These are sent by base station 21 in direct FSK modulated form and then received and demodulated by transponder 32. At 2400 baud, each data bit takes approximately 417 milliseconds. Fig. 9 illus¬ trates the timing for the transmission of a 0x0 nibble which has been encoded to be a 0x0f byte and then transmitted least significant bit first. Waveform Wl illustrates the data to be sent. Waveform W2 illustrates the actual FSK modulation, which appears identical to Wl . Waveform W3 illustrates the data received by transponder 32, which is of course a facsimile of the data sent.

Transponder 32 transmits data at 62,500 bps using a "missing pulse" encoding of an otherwise traditional asyn¬ chronous format. For purposes of comparison with the more traditional base station 21 transmission, Fig.10 illustrates the same OxOf data byte being transmitted by transponder 32. The logical data appears the same as that for the base station. However, the FSK modulation is in the form of one or two pulses per data bit time of sixteen microseconds. For a logical "1" data bit, two pulses are transmitted, each lasting four microseconds and separated by four micro¬ seconds. For a logical "0" data bit, the second pulse is omitted. Data shaper 77 in base station 21 reconstructs the original data, time delayed by eight microseconds, through a two-stage monostable multivibrator circuit. Operation of this circuit is detailed in the second about base station 21.

lable 3 - hanspondei ID Packet Contents (bytes i.ibble encoded unless otherwise noted)

Table 4 - Equations

Deliiiitions

P _t2 = success iate ior identifying c us belou ledundane) is applied

I' R2 ⁼ success rale for identifying c us ilk ι leiluiidancy is applied

ECR2 = failure rate, I P<_ _R2

W =the tune window dunng which to consider lntcileience

T = the durali m of an ID packet

U = the cntical duiation of an ID μacku needed l i identification

R = redundancy factor

Assumptions = 78ms for cais traveling .H I ) MI'H through a 15 toot window I = 944 ms ior the pieieπed embodiments 59 bytes at 02,500 bps K = 9 for the prelened embodiment U = I 6ms for the piclened embodiment

Equations

Equation I Equation 2 lot cuiienl assumptions Equation 3 Equation 4 p _CR , = () <Ji)q97 lor cuiienl assumptions Equation 5 1/350,0' 7 loi cuiienl assumptions Equation 6 P _{C 2} -0737 lor R- 10 and 1 = 1024ms Equation 7 P _LP2 = 099999S loi K-I and I = 1024ms Equation 8 I _fR2 = 1/642,101 lorR=I a-ιd r= 1024ms Fquulion 9 P _c2 (l-2l)ΛV) lor cntic.il identificntion Equal ion I0 > . loi cm rent cntical assumptions Equation 11 r,. _R2 = 1/3,037,000,000,000 loi cuirent cntical assumptions

ID Packet Redundancy and Reconstructing Laps Table 3 indicates the contents of an ID packet. Each byte is the result of nibble encoding and therefore contains only four data bits. The packet includes a unique sixteen- bit transponder ID number, a lap number, and a number of prior lap times, eight of them in the preferred embodiment. Additional synchronization and checksum information is also included.

The lap number is used by base station 21 and/or host computer 30 to reconstruct any transponder ID packets that may have been missed for any reason. This lap number is an arbitrary sequencing and may not correspond to number of race laps completed. For example, if packets are received for laps fifty-two and fifty-four for a given transponder, then it is obvious that lap fifty-three was missed. Since the lap time for lap fifty-three is included in the redun¬ dant information of the ID packet for lap fifty-four, the exact time at which lap fifty-three occurred can be accur¬ ately reconstructed by adding this redundant lap time to the known exact time of lap fifty-two. Further intelligence and resulting robustness may be added to base station 21 and/or host computer 30 software by diligently comparing all of the redundant lap time information from each successive ID packet.

If more than eight laps are missed, the reconstruction can not occur successfully and information is truly lost. The probability of this happening may be shown statistically to be very small. The lap number within the ID packet is allowed to repeatedly cycle from zero to two hundred fifty- five and then back to zero again without any significant probability of accidentally reconstructing a small number of laps (one to eight) when actually a large number were missed (two hundred fifty-six to two hundred sixty three) . Also, the lap time stored in the ID packet is limited to a maximum value of 1,048,576ms (over seventeen minutes) , which may be indicated by twenty bits in five nibbles. If the trans¬ ponder goes longer than this between detecting primary base station pings, then the usefulness of the redundancy

information for that lap is lost.

If two transponders transmit ID packets overlapping in time, then the first part of the earlier packet and the last part of the later packet may still be properly received by the base station. The ID is included very close to the front of the packet in order to increase the probability that the earlier transponder will be identified. Further¬ more, the ID is repeated in a packet footer in order to increase the probability that the later transponder will be identified. The two synchronization bytes that are used to identify the beginning of a packet are also repeated to introduce the packet footer. In this way, if the base station receives a complete packet, it does so by first detecting the leading synchronization bytes. The packet footer and its embedded synchronization bytes are then considered as simply embedded data. However, if the leading portion of the packet was corrupted by interference, the base station will detect the repeat copies of the synchroni¬ zation bytes and follow through decoding the packet until it reaches the flag byte indicating that it has actually decoded a packet footer and not a complete packet. Through this organization, substantial overlap and interference may still occur between two ID packets without losing the ability to identify the sending transponders. Only if the overlap is excessive or if a third ID packet interferes does it become necessary to resort to reconstruction from redun¬ dancy. Since cars are much more frequently two-abreast than three or more, the situation of only two ID packets inter¬ fering is much more likely. In this situation, the amount of redundancy included and the actual length of each packet may be significantly increased without increasing the probability of failing to identify both transponders. This in turn makes possible the almost arbitrary reduction in the probability of completely missing a lap.

Statistics

Of concern is the probability that multiple trans¬ ponders 32 will transmit their ID packets overlapping in time, such that one or more of the transponders can not be

identified by base station 21. This may occur when two or more cars cross reference line 3 at nearly the same time, presumably side-by-side. If and when cars are missed due to this circumstance, the redundancy encoded in each ID packet must be used to reconstruct the missing lap whenever the missed cars are successfully identified on a subsequent lap. In the preferred embodiment, a car may be missed on up to eight consecutive laps without losing the ability to recon¬ struct all of its laps.

The probability of completely missing a car is calcu¬ lated, using the equations listed in Table 4. For simpli¬ city, these equations initially assume only the probability when two cars are likely to be side-by-side, as indicated by the numeral "2" in the suffix.

The table begins with a few definitions. Variables for the probability of successfully identifying a car, without and with the use of redundancy, are defined.

At times, actual numbers are applied to the equations. Whenever this is the case, the following assumptions are made. The probabilities for cars being side-by-side are considered of a fifteen foot length of course. For example, cars travelling at 130MPH will travel fifteen feet in a time window of approximately 78ms. For cars travelling at other speeds, the time window may be linearly prorated.

Equation 1 indicates the raw probability for missing a car. Given the time at which a first car may transmit an ID packet of duration T, the second car, if by its side and thus transmitting during the time window W of concern, may transmit such that its ID packet overlaps with the first car's ID packet. This overlap may vary from just barely overlapping on the front, through complete overlap, to just barely overlapping on the end. This variation allows a range of 2T in timing for the second ID packet over the window W. The probability of successfully receiving both packets in their entirety is unity less the probability of overlap. This result is indicated in Equation 1. Note that "edge effects" when the second car is outside of the window are ignored. This makes the probability of Equation 1 more

conservative. Also, the identity of a car is not necessar¬ ily lost completely whenever arbitrary overlap occurs. Omission of this fact makes the probability of Equation 1 further conservative.

Equation 2 indicates that, under the assumptions, successful identification of both of two cars side-by-side will occur (conservatively estimated) in about seventy-six percent of the cases. By itself, this probability is unacceptably low. However, once redundancy is added, it becomes greatly improved.

Equation 3 incorporates the redundancy by asserting success if only one out of R redundant packets is received. Note that unity less the probability of failure equals the probability of success, and vice versa. Note also that the redundancy factor R equals one more than the number of redundant lap times included in the packet. Therefore R equals nine for the preferred embodiment.

Equations 4 and 5 indicate a very good success rate under current assumptions.

Equations 6, 7, and 8 indicate that the probability of failure is cut in half by increasing R by one, in spite of the increase in the failure rate ignoring redundancy. Thiε pattern of improved performance increases dramatically as R is further increased, limited to R reasonably smaller than W.

Whenever the probability of successfully identifying a car on a particular lap is adjusted for the fact that partial overlaps do not cause a failure, this probability increases. For the preferred embodiment, overlap which still allows the first or last eight bytes of the packet to be received will allow successful identification of both cars. To be conservative, a time of ten bytes is used in the equations, corresponding to the time U of 1.6ms.

Equation 9 updates Equation 1 for this critical identi¬ fication.

Equations 10 and 11 show that the probability for failure of the preferred embodiment is essentially zero.

Transponder

Fig. 11 indicates the preferred embodiment of trans¬ ponder 32. In general, the transponder is a microprocessor controlled half-duplex FSK transmitter and receiver. The transponder normally listens for a "ping" signal at 926.5 MHz. After recognizing one, it transmits an "ID packet" at 915.8 MHz.

Processor 91 normally instructs the PLL-based FSK transmitter 93 to output a low power level and instructs antenna switch 94 to route any incoming ping signal from the low-gain, half-dipole, doughnut-directional antenna 37 to bandpass filter 98. The ping, FSK modulated on a 926.5 MHz carrier, is isolated from off-frequency noise by the filter and then forwarded to a mixer 97. At this time, no modula¬ tion is being sent to transmitter 93 so its 915.8 MHz output may be used as the local oscillator signal in mixer 97. This dual use of transmitter 93 keeps transponder cost at a minimum. The output of the mixer is 10.7 MHz intermediate frequency signal that is further band pass filtered by filter 100 and then forwarded to an FSK receiver 96. The baseband data from FSK receiver 96 is then shaped by data shaper 95 and forwarded back to microprocessor 91 where it is decoded in order to distinguish between primary base stations, secondary base stations, and illegitimate ping signals.

Meanwhile, the received signal strength indicator from FSK receiver 96 is provided to microprocessor 91 so that it may determine the signal peak (58 of Fig. 5) and thus the exact crossing time over reference line (3 of Fig. 4) .

When the time comes to transmit an ID packet, micropro¬ cessor 91 switches antenna switch 94 to route the trans¬ mitter 93 signal directly out antenna 37. Meanwhile, microprocessor 91 increases the power level of transmitter 93 and begins providing modulation data. Fig. 10 details the 62,500 bps asynchronous "missing pulse" format used.

At this time, microprocessor 91 also provides a trigger signal to an optional on-board display 42. The trigger signal is time shifted and encoded to prevent use of the

trigger by other devices.

Memory 90 for storing redundant lap time information and ID packets is actually included as a part of micropro¬ cessor 91. Light emitting diode 92 is used as a power on indicator, a low battery indicator, and a diagnostic ping detect indicator. Power supply 99 is designed to produce a regulated five Volt supply while being connecting to a variety of sources, including a nine Volt battery, a twelve or eighteen Volt automotive system, or an AC power adapter with a DC output between seven and thirty volts.

Some of transponder 32 blocks of Fig. 11 are shown in detail in Fig. 14, and each transponder 32 block is func¬ tionally based on a single off-the-shelf RF component. Most of the other components in each block of transponder 32 are passive components acting in support of the central off- the-shelf component, as recommended by its manufacturer.

In PLL-based FSK transmitter 93, U5 is a Motorola MC1317D PLL chip. Crystal X2 with load capacitors Z36 and Z35 provide a 28.618750 MHz reference frequency, which may be shifted by turning on diode Zll and thus providing capacitor Z32 an AC path to ground. Diode Zll is turned on by dropping processor bus signal RC6 (TXDATA) to ground, a signal RC filtered by Z15 and Z34. The PLL loop filter connects to pins six and seven of U5. Z10, Z12, and Z31 form a lag-lead filter, the effect of which is amplified by transistor Ql , the base bias of which is in turn established by Z13 and Z14. Z13 and Z14 also act in combination with Z10 to set the gain of Ql . Control current then flows through Z9 and Z8, augmented by Z7, into control pin six of U5. The PLL high frequency oscillator consists of Z80 and Z79, and operates at exactly thirty-two times the crystal reference. Components Z30, Z37, Z38, and Z78 provide power supply decoupling. Current through Z16 enables U5. Current through DI and Z3 sets the PLL transmit power whenever the transponder is in transmit mode (processor bus signal RCl is high) . Current through D2 and Z6 sets the PLL transmit power whenever the transponder is in receive mode (processor bus signal RCO is high) . Components Z83 and Z84 provide a

DC path to power for the PLL outputs and perform impedance matching in combination with Z39, ZlO, and Z86. Although at the same frequency and interchangeable, it' s convenient for one PLL output to provide the RF transmit signal and the other output to provide the receive RF local oscillator signal.

In antenna switch 94, U4 is a Motorola MRFIC2003 antenna switch chip. Z42 and Z43 provide power supply decoupling. Z41 provides AC coupling to the antenna. Z18 and Z17 assist in pulling up the two control lines of U4 , since the processor can not pull these lines up to Vcc on its own. When processor bus signal RC4 , driving U4 pin seven, is high while simultaneously RC3 , driving U4 pin six, is low, the U4 connects the antenna to the receive path for receive mode. When the control lines RC4 and RC3 are reversed, U4 connects the antenna to the transmit path for transmit mode. Z44 provides AC coupling to RF bandpass filter 98.

In RF bandpass filter 98, Fl is a Motorola Ceramics KFF6141A bandpass filter centered at 926.5 MHz, the signal transmitted by base station 21 and received by transponder 32. This filter is also designed to provide a strong stop band at 905.1 MHz, protecting against this parasitic image frequency after mixing down to the 10.7 MHz IF frequency.

In mixer 97, U3 is a Motorola MRFIC2001 mixer chip. Z89 provides a DC path to power while Z47 and Z48 decouple the chip from the power supply. Z46 assists in power supply decoupling. Z90 provides a DC path to power for the U3 output signal, while Z50 and Z51 decouple it from the power supply. Z45 AC couples RF bandpass filter 98 to U3, while 249 is chosen to match impedance between U3 and IF Filter 100.

In IF filter 100, F2 is a Toko SK1O7M1-A0-00 filter centered at 10.7 MHz with a 230 kHz pass band width.

The remaining transponder 32 blocks of Fig. 11 are shown in detail in Fig. 15, and each transponder 32 block is functionally based on a single off-the-shelf RF component. Most of the other components in each block of transponder 32

are passive components acting in support of the central component, as recommended by its manufacturer.

In processor 91, U2 is a Microchip PIC16C73A micropro¬ cessor which is one-time-programmable, has analog-to-digital conversion capacity, has 192 bytes of general purpose memory, and has 14K words of fourteen-bit wide program memory. Z69 provides power supply decoupling. Z70, Z71, and X3 form its 4 MHz crystal circuit. Z25 and Z72 form its power-up reset circuit. Potentiometer Z81 provides an adjustable voltage reference to analog input AN4 and is used for adjusting threshold 57, 59 (Fig. 5) for recognizing base station ping signal 56. Z4 and D3 provide circuit protec¬ tion for the external line that may optionally lead to an on-board display 42. Analog input AN2 is used for monitor¬ ing battery level from the power supply. Numerous control lines connect to or from the processor U2 via the processor bus.

In LED 92, D5 is an indicator LED, the current of which is controlled by Z97.

In FSK receiver 96, Ul is a Motorola MC13150FTB coil- less receiver chip. It mixes an incoming 10.7 MHz IF signal down to a second IF frequency of 455 kHz, which is further filtered, amplified, limited, and FM detected. XI, Z66, and Z67 form a crystal circuit operating at 11.155 MHz. Z52 AC couples the IF signal into Ul . F3 and F4 filter the 455 kHz second IF. Z54, Z57, and Z56 are used by the IF amplifier. Z55, Z58, Z64, and Z65 are for power supply decoupling. Z59 and Z60 are used by the limiter. Z19, Z20, Z21, and Z61 are used by the FM detector. Z23, Z22, and Z62 control detector gain and bandwidth. Z24 sets RSSI (Received Signal Strength Indicator) gain while Z91 and Z92 low pass filter and thus smooth the RSSI that is forwarded to the processor bus RAO analog input. The demodulated signal from Ul pin 23 is forwarded to data shaper 95.

Data shaper 95 consists of a transistor Q2 that is either off or driven into saturation. The FM detector output is RC low pass filtered by Z27 and Z63 and then AC coupled through Z77 to transistor Q2 , the base bias of which

is established by Z28 and Z29. The output of the transistor is RC low pass filtered by Z98 and Z85 and then forwarded to the processor bus RC7 asynchronous data input.

Power supply 99 consists of a five volt voltage regu¬ lator U6 with input and output capacitors Z73 and Z74, respectively. Nine volt battery or higher automotive voltage is provided on connector P2 pins 1 and 2 via a nine volt Tee connector. This voltage is divided by Z53 and Z93 to approximately fourteen percent of its original level so that the five volt processor can monitor battery voltages as high as thirty-five volts. Separate digital and radio power supplies are provided, AC isolated by Z96. Additional power supply noise filtering and load decoupling are provided by Z75 and Z76.

On-Board Display

Fig. 12 shows the preferred embodiment of on-board display 42. In general, on-board display 42 is a sophisti¬ cated stopwatch. Microprocessor 111 receives trigger information from a wire driven by a transponder 32. This trigger information is to determine lap times, which are stored in memory 114 and displayed in real time on a six- digit numeric LCD display 112. In the preferred embodiment, the most recent ninety-nine laps will be recorded and they will be numbered one through ninety-nine. Display 112 indicates two digits for lap number, two digits for seconds, and two digits for hundredths of seconds, each separated by a period. Higher order minutes and lower order milliseconds are not displayed. At any time, a button 110 may be pressed to recall lap numbers and lap times on LCD display 112. If button 110 is held down for at least three seconds, memory 114 is reset and display 112 indicates six dashes until the next trigger is received.

On-board display 42 may be connected to a printer or personal computer 115 using standard 9600 bps asynchronous communication. If a break signal or data sequence is received by processor 111, it responds by pausing one-half second and then sending back a lap time report containing all memorized lap times. This report is in ASCII and

formatted 117 to produce a suitable report when sent directly to printer 115. In order to generate the break signal or data sequence when using printer 115, a specially wired printer cable 117 is used to loop back the on-board display's own output into its input. On-board display 42 periodically outputs a break or data sequence which will be looped back to itself if the cable is attached. If on-board display 42 is connected to a personal computer 115 instead of to a printer, software on the personal computer may be executed for generating the break and then storing the lap time report for later reference. At twenty-five characters per lap time, ninety-nine laps maximum, and 9600 baud, a complete lap time report consists of less than two and one-half kilobytes and is sent in under two and one-half seconds.

Memory 114 for storing lap times is separate from microprocessor 111 to increase capacity. Power supply 116 is designed to produce a regulated five volt supply while being connecting to a variety of sources, including a nine volt battery, a twelve or eighteen volt automotive system or an AC power adapter with a DC output between seven and thirty volts.

Printer cable 117 is designed to carry a standard asynchronous transmit signal from on-board display 42 to a printer 115. It also feeds this same signal back into the on-board display's receive line. On-board display 42 periodically transmits either a break or data sequence while monitoring the receive line. Whenever cable 117 is attached, this break or data sequence will be looped back into the on-board display's receive line where it will be recognized as an indication that printing should ensue.

DB25 printer cable wiring 117 is indicated below. Note that DB25 position three is connected to both on-board display connector positions one and two, forming the loop back. This connection is performed on the display side of cable 117 so that only two wires must run through the cable to printer 115. No handshaking is performed with the printer. The transmit speed is slow enough and the total

byte count small enough that handshaking is not necessary. Printer 115 must be on-line and ready to print before cable 117 is connected to on-board display 42.

Display Display Connector DB25 Connector Printer Signal Position Position Signal

Ground 3 7 Signal Ground

Receive Data 2 3 Receive Data

Transmit Data 1 3 Receive Data

Following is an example of a lap time report printed by on-board display 42 directly to a printer 115. Note that the elapsed minutes will roll back to zero after sixty seconds. The first elapsed time is about fourteen minutes after on-board display 42 was reset and its power left on. The lap time for this line is defaulted to the same time and is typically not a valid lap time. Each subsequent lap time equals the difference in the elapsed time for that line and the elapsed time for the previous line.

Lap Elapsed Lap Time

1 14:15.751 14:15.751

2 15:47.318 1:31.567

3 17:17.950 1:30.632

4 18:48.731 1:30.781

5 20:18.686 1:29.955

(lines omitted)

99 07:02.501 3:12.549

Each on-board display 42 block illustrated by Fig. 12 is functionally based on a single off-the-shelf component. Most of the other components in each block of on-board display 42 are passive components acting in support of the central off- the-shelf component, as recommended by its manufacturer.

As shown in Fig. 16, in power supply 116, U2 is a five volt regulator. Connector P2 positions four and three provide twelve volts and ground, respectively, and are typically connected to the automotive battery. Z44 protects against noise entering through this connection while Z42

protects against reverse polarity. Z36 and Z37 assist in stabilizing the five volt output.

In memory 114, U4 is a Microchip serial EEPROM of the 93LC46/56/66 family into which lap times are stored and retained even if the power is removed. Z20 provides power supply decoupling.

In processor 111, U2 is a Microchip PIC16C73A micropro¬ cessor which is one-time-programmable, has one hundred ninety-two bytes of general purpose memory, and has fourteen K words of fourteen-bit wide program memory. Z21 provides power supply decoupling. Z22 , Z23, and XI form its four MHz crystal circuit. Z2 and Z24 form its power-up reset cir¬ cuit. Z41 and Z8 provide circuit protection for the exter¬ nal trigger line that is connected to a transponder 32. Numerous control lines connect to or from processor U2 via the processor bus.

In communication 113, U6 is a Maxim MAX232 communica¬ tions chip which uses Z25 through Z29 to produce EIA standard communication voltages. Connector P2 positions one and two provide asynchronous transmit and receive signals, respectively. These signals are fed to and from processor 111 at TTL voltage levels.

In button 110, S2 is a rubber button with a conductive puck, centered over an interlaced finger design on the printed circuit board, such that depression of the button effects a contact closure between the processor's RB6 input line and ground. Z9 serves as a pull up to five volts which is worked against whenever the button is pressed, and in combination with Z33 provides a. limited amount of contacting debouncing.

On-board display block 112 shown in detail in Fig. 17 is functionally based on two off-the-shelf components. Most of the other components in block 112 are passive components acting in support of the central off-the-shelf components, as recommended by their manufacturers.

In display 112, U8 is a Hitachi LCD controller chip and U9 is a Standish translucent LCD display with six digits, each composed of seven segments and a decimal point. Z34

provides power supply decoupling. Z13 sets the internal oscillator frequency of U8 which is used for timing the multiplexing of LCD lines. Several connections are made from U8 to ground to establish standard LCD contrast values. Data bytes (eight bits wide) are provided from processor 111 over the processor bus to U8 inputs D7 through DO . Four control lines are also provided from processor 111 over the processor bus. Backlighting of the LCD is provided by ten side-facing LED's (D90 through D99) which illuminate the printed circuit board beneath the translucent LCD, this light reflecting off of the printed circuit board and through the LCD to provide adequate illumination of the display at night. Z90 through Z99 control the current level through the LED's while Ql turns them on and off via a control signal through Z89 from the processor output RB7.

Base Station

Fig. 13 indicates the preferred embodiment of base station 21. In general, the base station is a micropro¬ cessor controlled full-duplex FSK transmitter and receiver. The base station continuously transmits a "ping" signal at 926.5 MHz. Meanwhile it receives intermittent "ID packets" from the transponders at 915.8 MHz.

Microprocessor 74 (the preferred embodiment actually contains an interconnected pair in order to have enough resources to perform all of the required jobs with an inexpensive microprocessor) provides encoded ping data to PLL-based FSK transmitter 75 which uses a 926.5 MHz carrier. This signal is fed through a circulator 76 which routes it out antenna 51.

Meanwhile, incoming ID packets are picked up by the high-gain, highly directional, yagi antenna 51 and fed through circulator 76 which routes them to a bandpass filter 79. These ID packets, modulated on a 915.8 MHz carrier, are isolated from off-frequency noise by the filter and then forwarded to a chain of low-noise amplifiers 82. The amplified signal is then bandpass filtered by 84 again and then mixed by mixer 81 with an 870.8 MHz local oscillator 78 signal to produce a 45 MHz intermediate frequency signal.

An FSK receiver 80 demodulates the data. The resulting baseband data is further shaped by data shaper 77 and then forwarded to microprocessor 74 where it is decoded and possibly error corrected. Properly decoded ID packets may be further analyzed by processor 74 or forwarded to host computer 30 via an asynchronous communications interface 73. The ID packet contains data at 62,500 bits per second, as indicated in Fig. 10. Because of the "missing pulse" FSK modulation scheme, a special but simple data shaper 77 is required in addition to the data shaper built into FSK receiver 80.

For diagnostic purposes, a keypad 71 and display 72 are included with base station 21. Keypad 71 may be used to invoke various test modes, including a mode where the ping is turned off and on every five (or one hundred) seconds in order to perform a stationary transponder test. Display 72 in turn may indicate the ID number of a transponder whose packet was successfully received or alternatively it may indicate diagnostic information such as the number of valid encoded bytes received over an interval of time. Such diagnostic information may be used to evaluate both trans¬ ponder 32 and base station 21 performance.

Some of base station 21 blocks of Fig. 13 are shown in detail in Fig. 18, and each base station 21 block is func¬ tionally based on a single off-the-shelf RF component. Most of the other components in each block of base station 21 are passive components acting in support of the central com¬ ponent, as recommended by its manufacturer.

Circulator 76, bandpass filter 79 and amplifier chain 82 are each off-the-shelf items complete with connectors and they are not included in the circuit diagram.

In PLL-based oscillator 78, U5 is a Motorola MC1317D PLL chip. Crystal X2 with load capacitors Z36 and Z35 provide a 27.2125 MHz reference frequency. The PLL loop filter connects to pins six and seven of U5. Z10, Z12, and Z31 form a lag-lead filter, the effect of which is amplified by transistor Ql , the base bias of which is in turn estab¬ lished by Z13 and Z14. Z13 and Z14 also act in combination

with Z10 to set the gain of Ql . Control current then flows through Z9 and Z8, augmented by Z7, into control pin six of U5. The PLL high frequency oscillator consists of Z80 and Z79, and operates at exactly thirty-two times the crystal reference. Components Z30, Z37, Z38, and Z78 provide power supply decoupling. Current through Z16 enables U5. Current through Z6 sets the PLL power level. Only one of the two PLL outputs is used, although both are provided with a DC path to power through Z83 and Z84. Z40 and Z86 perform impedance matching with mixer 81.

In bandpass filter 84, Fl is an off-the-shelf ceramic bandpass filter centered at 915.8 MHz. ZM2 AC couples the amplified signal from amplifier chain 82. Fl removes unwanted off-band signals and then Z45 AC couples the signal to mixer 81.

In mixer 81, U3 is a Motorola MRFIC2001 mixer. It receives a 870.8 MHz local oscillator signal from PLL-based oscillator 78 and an amplified transponder signal at 915.8 MHz from bandpass filter 84, mixing them down to an inter¬ mediate frequency of 45 MHz which is impedance matched through Z49 and forwarded to FSK receiver 80. Z89 and Z90 provide DC paths to power while Z46, Z47, Z48, Z50, and Z51 provide power supply decoupling.

In FSK receiver 80, Ul is a Motorola MC13158FTB FM receiver. It uses a Butler Crystal Oscillator circuit producing a 55.7 MHz reference frequency. This crystal circuit includes components Z65, ZB1, Z64, ZB23, ZB6 , X4 , ZB3, and ZB5. An off-the-shelf 45 MHz IF filter F2 is included within block 80. The input of filter F2 comes from mixer 81 and is impedance matched by Z39. The output of filter F2 goes to receiver chip Ul and is AC coupled through Z52. Receiver chip Ul mixes the input signal with the 55.7 MHz reference frequency to produce a second intermediate frequency signal at 10.7 MHz. This signal is amplified and filtered by numerous components (F3, Z59, ZA5, ZA6, ZA7, Z60, Z54, Z56, Z57, Z61, ZA8 , ZA9, ZB0 , Z62, and F4) accord¬ ing to Ul manufacturer recommendations. Note that the ceramic filter pair F3 (likewise F4) may be used in lieu of

discrete components ZA5, ZA6, and ZA7 (likewise ZA8, ZA9, and ZBO) . ZM3 , ZM4 , Z58, Z55, and Z58 provide power supply decoupling. Z63, Z77, and Z78 are used by Ul ' s limiter according to manufacturer recommendations. Z87, Z79, ZAl , and Z19 provide a 10.7 MHz tank circuit for use by Ul 's quadrature detector. Z20, Z21, and ZA2 set the detector gain and bandwidth. ZA3 couples the detector output to Ul's on-board data slicer, the performance of which is adjusted with ZA4. Z22 provides a pull up for Ul's open collector data output, and this RFDATA is forwarded to data shaper block 77. Z23 , Z24, and Z99 control RSSI (Received Signal Strength Indicator) gain and bandwidth.

In power supply 83 separate voltage regulators are provided to prevent noise injected into the power supply by one block from affecting another block. U6 is an off-the- shelf five volt regulator with input and output capacitors ZN2 and AN3. Identical circuits surround the other off- the-shelf regulators, U4 , Ull, and U11B. Regulators U6 , U4 , Ull, and UllB are identical. External power and ground are provided to base station 21 via connector P2. Z95 and D4 provide noise isolation and reverse polarity protection, respectively.

Additional base station 21 blocks of Fig. 13 are shown in detail in Fig. 19, and each base station 21 block is functionally based on a single off-the-shelf component. Most of the other components in each block of base station 21 are passive components acting in support of the central component, as recommended by its manufacturer.

In processor (s) 74, two identical Microchip PIC16C73A microprocessors U2 , U12 are used. They each have analog-to- digital conversion capability, one hundred ninety-two bytes of general purpose memory, and fourteen K words of fourteen- bit wide program memory. Z 69 and ZP1 provide power supply decoupling. Z70, Z71, and X3 as well as ZP3, ZP4 , and X6 form two 10 MHz crystal circuits. Z25 and Z72 as well as ZP6 and ZP7 form power-up reset circuits. ZR1, ZR2, and ZR3 provide an analog signal to U2 ANl for detecting battery voltages up to 27.7 volts. Z97 and D5 provide an indicator

LED circuit which illuminates whenever a transponder signal 36 is detected. Numerous control lines feed to other blocks of base station 21, as described in association with those blocks. The two processors are interconnected by several lines. RBO is asserted by U2 to produce and interrupt on U12 whenever transponder data is received. U12 then exer¬ cises control over the SCI bus (SS, SCL, SDA, and SDO) and reads the data from U2. The SDA and SDO lines are crossed on U2 to maintain communications compatibility. Both U2 and U12 connect to the TXHOST (communication with host computer or printer 30) , TXDATA (data transmitted by the PLL-based FSK transmitter 75) and TX Control lines (feeding the PLL-based FSK transmitter output level controls DI , D6 , and D7) although only U2 drives these lines. U12 connections are in the high-impedance input state.

In keypad 71, an off-the-shelf three by four membrane keypad is used. Diodes DIO, Dll, and D12 are individually pulled to ground in order to pull selected keypad columns to ground. If any key is pressed in the selected column, a corresponding row will be pulled to ground, this voltage being provided to the processor on lines RBI through RB4. Note that internal to the processor on these lines are pull up resistors to five volts. Control of DIO, Dll, -and D12 is handled by octal latch U12B due to a lack of sufficient I/O pins on processor U12. This latch is clocked by the pro¬ cessor control line RCl and fed by processor signals RAO and RBI through RB7. RAO substitutes for the expected RBO because the special interrupt function of RBO is used for other purposes.

In display 72 an off-the-shelf alphanumeric LCD display module is connected using a standard fourteen-pin interface, this interface augmented with two pins to provide back¬ lighting. Communication with the LCD is through selected microprocessor port A, B, and C lines. ZN8 is used for contrast control. Q3 and ZN9 control the backlight.

In data shaper 77, since the RFDATA from FSK receiver 80 is in a "missing pulse" format, traditional asynchronous start bits, data bits, and stop bits must be reconstructed.

Fig. 10 illustrates the timing for the reconstruction of a OxOF data byte (least significant bit first) . LM319AN comparator U8 , referring to the voltage reference produced by the divider ZE2 and ZE3 , is used to convert slightly noisy pulses into a clean signal issued from pin twelve. This signal is pulled up by ZE5 and forwarded to the DM74123 high speed dual retriggerable monostable multivibrator U9. Whenever a OxOF data byte is received, this pin twelve signal will resemble waveform W5 of Fig. 10. The output of the retriggerable monostable multivibrator will then resem¬ ble waveform 6 of Fig. 10. Using timing established by ZE7 and ZJ2 , the first stage of U9 lengthens each pulse to very slightly over one-half bit time (8us at 62,500 bps) . For a string of "1" bits there is a pulse every 8us, and U9 ' s first monostable multivibrator is repeatedly retriggered before it can expire, keeping its output at five volts. However, if a "0" bit occurs and a pulse is missing, it will expire and provide a trigger (from the inverted output) to the second monostable multivibrator the timing of which is established by ZE6 and ZJl such that it's inverted output on pin nine will fall to zero for exactly one bit time (16us) . This pin nine output will otherwise remain at five volts. On the other hand, if a sequence of "0" bits occurs, the second monostable multivibrator will be repeatedly retrig¬ gered before it can expire, keeping its inverted output a zero volts. As a result, of this two-stage process, the output from pin nine of U9 reproduces a standard 62,500 bps asynchronous signal (waveform W7) that is an eight micro¬ second delayed facsimile of the data transmitted by the transponder (waveform W4) .

In communication 73, U10 is a Maxim MAX232 communica¬ tions chip which uses ZJ3 through Z17 to produce EIA stan¬ dard communication voltages. Connector P3 positions one, two and three provide asynchronous transmit, ground, and receive signals, respectively. These signals are fed to and from U12 of processor 74 at TTL voltage levels over the (comm) communication bus.

Base station 21 block 75 of Fig. 13 is shown in detail

in Fig. 20, and block 75 is functionally based on a single off- the-shelf RF component. Most of the other components in block 75 are passive components acting in support of the central component, as recommended by its manufacturer. In PLL-based FSK transmitter 75 , U7 is a Motorola MC1317D PLL chip. Crystal X5 with load capacitors ZH2 and ZH5 provide a 28.953125 MHz reference frequency, which may be shifted by turning on diode D2 and thus providing capaci¬ tor ZH2 an AC path to ground. Diode D2 is turned on by dropping processor bus signal RC2 (TXDATA) to ground, a signal current limited by ZEl . The PLL loop filter connects to pins six and seven of U5. ZD6, ZD8 , and ZH3 form a lag-lead filter, the effect of which is amplified by trans¬ istor Q2, the base bias of which is in turn established by ZD9 and ZEO . ZD9 and ZEO also act in combination with ZD6 to set the gain of Q2. Control current then flows through ZF5 and ZF4 , augmented by ZF3 , into control pin six of U7. The PLL high frequency oscillator consists of ZGO and ZHl , and operates at exactly thirty-two times the crystal reference. Components ZHO, ZE8, and ZJ7 provide power supply decoupling. Current through ZE8 enables U5. Current through Dl, ZDO, and ZFO sets the normal PLL transmit power level. This level may be set to one of two diagnostic levels by providing current instead through either D6 and ZN5 or D7 and ZN7. An off-the-shelf balun transformer ZH7 is used to combine the two outputs. ZN8 provides an AC path to ground while ZN9 AC couples the 926.5 MHz signal to circulator 76.

Parts List Transponder 94 Antenna Switch U4 Motorola Semiconductor

MRFIC2003 98 Band Pass Filter Fl Motorola Ceramics

Division KFF6141A 97 Mixer U3 Motorola Semiconductor

MRFIC2001 100 IF Filter F2 Toko: SK107M1-A0-00

96 FSK Receiver Ul Motorola Semiconductor MC13150FTB

96 455kHz Filters F3, F4 muRata CFU445B2

97 Data Shaper Q2 Diodes Inc. MMBT3904DI

(Digi-Key part number)

91 Processor U2 Microchip PIC16C73A/SO-4

92 LED D5 Industrial Devices 5300H1LC

93 PLL-Based FSK Transmitter U5 Motorola Semiconductor MC13176D

93 Transistor Ql Diodes Inc. MMBT3905DI

(Digi-Key part number)

93 Adjustable Coil Z80 Toko NE5478BNAS-100111

98 Power Supply U6 Linear Technology LT1121CST-5 (Digi-Key part number)

On-Board Display:

111 Processor Ul Microchip PIC16C73A/SO-4

112 LCD Display U9 Standish 3918-365-920 112 LCD Controller U8 Hitachi HD61604

112 Transistor Ql Diodes Inc. MMBT3904DI (Digi-Key part number)

114 Memory U4 Microchip 93LC66

113 Commuication U6 Maxim MAX232CWE 110 Button S2 Advanced Rubber Concepts GL-001

116 Power Supply U2 Linear Technology LT1121CST-5 (Digi-Key part number)

Base Station: 76 Circulator UTE Microwave CT-1423-0

79 Band Pass Filter Microwave Circuits B0109151

82 Amplifier Chain Mini-Circuits ZFL-100LN (2 used in series)

84 Band Pass Filter Fl Toko 4DFA-926A-11

80 Band Pass Filter F2 Toko 663BBX-050

80 FSK Receiver Ul Motorola Semiconductor

MC13158FTB

77 Comparator U8 National Semiconductor

LM319AN

77 Monostable Multivibrator U9 Fairchild Semiconductor

DM74123

74 Processors U2, U12 Microchip PIC16C73A/JW

71 Keypad Brady 2054654

72 Display Standish TIM81SLC-LE4

73 Communication U10 Maxim MAX232CPE

78 PLL-Based FSK Transmitter U5 Motorola Semiconductor

MC13176D

78 Transistor Ql Diodes Inc. MMBT3905DI

(Digi-Key part number)

78 Adjustable Coil Z80 Toko NE5478BNAS-100111

75 PLL-Based FSK Transmitter U7 Motorola Semiconductor MC13176D

75 Transistor Q2 Diodes Inc. MMBT3905DI

(Digi-Key part number)

75 Adjustable Coil ZGO Toko NE5478BNAS-100111

83 Power Supply U6 , U4 , Ull, UllB Linear Technology LT1121CST-5 (Digi-Key part number)

51 Antenna Cushman PC9013N

The invention in its broader aspects is not limited to the specific details shown and described, and departures may be made from such details without departing from the princi¬ ples of the invention and without sacrificing its chief advantages.