Token ring local area network testing apparatus using time delay reflectory

Description Submit all comments and votes

FIELD OF THE INVENTION

The present invention relates to network testers and, in particular, to a testing apparatus connectable to a token ring local area network.

BACKGROUND OF THE INVENTION

In general two types of local area networks (LANS) are presently available: collision based and token-ring.

Collision based LANs are wired in "parallel" so that each station on the LAN is logically connected to the same cable as every other station. As a result when any one station transmits, its signal reaches all the other stations at roughly the same time, allowing for propagation delays. Each station on the LAN watches the destination address of the transmitted frame and determines if it is the intended receiver, if it is, the station copies the frame. If two or more stations transmissions overlap in time they will interfere with each other additively and the data will be lost. This is called a collision.

Stations attempt to avoid collisions by looking for an idle state on the line before transmitting. However, due to speed of light delays, a station will not always be able to immediately detect if another station is transmitting because that station's signal might not have reached it across the cable yet. The extreme case of this is when each station begins transmitting simultaneously. When any station detects a collision it will stop transmitting, pick a random time interval, wait that interval, and then attempt to retransmit.

In a network where a high volume of frames are being sent collisions can significantly limit the usable data bandwidth on a network.

Token ring LANs avoid this limitation by allowing only one station on the network to transmit at any one time. Token ring LANs are wired in a logical ring with a transmit output from each station connected to a receive input on the next station. As long as no station has a frame to transmit a special frame called a token is circulated around the ring passed from station to station. When a station needs to transmit a frame it must wait until it receives the token. The station then removes the token from the ring, transmits its frame, then releases a new token. Thus, there can be only one token on the ring at a time and therefore only one station can be transmitting at a time.

While logically connected in a ring, the physical wiring of token ring LAN forms a star with the transmit and receive pairs for each station, combined in a single cable, running to a central point where they are connected to a media access unit (MAU). The MAU merely provides the connection of one station's transmitter to the next station's receiver to complete the logical ring. A station that is not currently part of the ring (also referred to as being DEINSERTED from the ring) will be bypassed by relays in the MAU. When that station wishes to become part of the ring (when it INSERTS) it will provide a steady DC voltage signal to cause the relays in the MAU to switch and insert it into the ring.

The protocol that describes the operation of the particular token ring LANs of interest is defined by the IEEE standard 802.5. Each station on the ring implements the Media Access Control (MAC) level of the 802.5 standard in a combination of hardware and software. Acquiring the token before transmitting as described above is one of the basic operations defined by the MAC protocol. Once a station acquires the token, it may transmit data addressed to one or more of the other stations on the ring.

Each station examines data on its receiver to determine if the data is addressed to it. If not, the data is merely buffered and regenerated through its transmitter to the next station downstream on the ring. In this way, each station acts as a repeater. If a station determines that data is addressed to itself, it will copy the data and change a "frame copied" bit in the data frame to indicate that the data has been copied. The data frame with the copied bit set is then sent to the next station downstream on the ring. Thus, when the station which originally sent the data eventually receives its own data frame back, it is able to determine that the data was successfully copied. At that point the originating station strips its transmitted frame from the ring and releases a new token. Now another station which has data to transmit may acquire the token and send data over the ring.

Part of MAC protocol requires that one station assume a leadership role of "active monitor." The active monitor can be any of the active stations on the ring and can change if the current station acting as active monitor deinserts itself from the ring or is otherwise unable to perform the duties of the active monitor. The active monitor performs several key functions in maintaining the MAC protocol. For instance, it watches for certain protocol violations and initiates recovery procedures if they are necessary. It also initiates a "neighbor notification" process by periodically sending (default of every seven seconds) an "active monitor present" frame which causes each station in turn to identify itself to its next downstream station. Neighbor notification allows network management software to obtain a map of the ring topology and inform each station of its nearest upstream active neighbor (NAUN) which is useful information in error recovery and trouble shooting. Most importantly, the active monitor provides the master or reference clock with which every station on the network must synchronize.

A typical station on a token ring network is illustrated in prior art FIG. 1. A station 50 normally comprises a processor 52, monitor 54, and a network interface card (NIC) 56. It is critical that each station include a NIC 56 since the MAC code as well as the hardware which allows a station to insert into the network are normally stored on the NIC. Each NIC is connected to the MAU by a receive pair and transmit pair, in this case 60 and 64, respectively.

As shown in prior art FIG. 2, MAU 68 includes a plurality of relays and ports to connect stations on the token ring. For example, the transmit pair from port 72 is connected to the receive pair of port 76, and the transmit pair of port 76 is connected to the receive pair of port 80 and the transmit pair from port 80 is connected to the receive pair of port 82. Each port also contains relays including relays 84, 86, which connect to incoming transmit and receive pairs from a station's NIC. Relays 84, 86 remain connected to each other, thereby looping the transmit and receive pair from a station back on each other, until the station 50 applies a "phantom voltage" of +5 v to open the relays 84, 86 and insert the station 50 into the network. Until the phantom voltage is applied, the transmit and receive pairs of a port are electrically connected to each other.

Occasionally, problems arise in the operation of a token ring. For instance, sometimes the ring will enter a "beacon state" in which the entire operation of the ring will be shut down until the problem is solved. One example of an error which would cause a beaconing ring is if a station on the ring has a bad NIC and is not correctly repeating frames. This could cause the token to be lost, and no station could then send data. Another problem that occasionally arises relates to the inability of a station to insert into the ring. In order to trouble-shoot problems such as these, testers have been developed to try to focus in on the cause or source of the error(s).

A variety of tests are currently performed during a trouble-shooting procedure. Testers can currently measure phase jitter over a ring and test the continuity of cables used to connect stations in the ring. Prior art testers can also determine the speed at which the network is operating and not insert if the station is not at the correct speed. Prior art tempters are also able to detect if a station, which has been unable to insert into the network, is trying to use an address that is already taken by a station on the ring. However, in order to perform these tests with prior art testers one or all of the users on the network are sometimes inconvenienced and/or the tests are not always conclusive.

For instance, in the case where it is desired to perform tests on the network when all of the ports in the MAU are full, one of the stations must be disconnected to allow for access to the network by the tester. This situation is illustrated in prior art FIG. 3. Station 50 had to be disconnected from MAU 68 in order for prior art tester 90 to have access to the network. Depending on the tests being performed, the user of station 50 might be off the network for a substantial period of time. Also, if station 50 had been the cause of the error and it was removed from the network to make room for a prior art tester, it would not be clear to the user of the prior art tester 90 whether the problem was merely intermittent or if the removal of station 50 rectified the error.

Prior art testers which measure phase jitter on the token ring require that the ring be rendered inoperable for the duration of the test. Depending on the size and usage of the ring, a delay of such a duration could be very undesirable if not intolerable.

Prior art testers are also deficient in a number of other respects. In testing for a fault at the end of a conductor cable adjacent to the tester, an intermediate cable of several feet must be inserted between the tester and the cable conductor end when using time delay reflectory (TDR) to determine whether a fault exists in the cable connector end. In using TDR to identify faults along the cable conductor, prior art testers typically require a number of circuit elements to be changed in matching the tester's input impedance to the cable conductor impedance in order to avoid unwanted multiple return or reflected signals. Prior art testers also lack the ability to check and modify the speed at which they transmit frames into the token ring so that this rate conforms to the operating speed of the token ring. Prior art testers also lack certain information gathering techniques that are useful in isolating faults on the token ring. In addition, in order to perform the necessary token ring tests, prior art testers are required to be configured in a variety of ways. Prior art testers do not have the capability to perform all of these tests from one location in the network.

SUMMARY OF THE INVENTION

In accordance with the present invention, a testing apparatus is provided for use with a token ring local area network (LAN). The testing apparatus includes hardware to enable it to be connected between a token ring media access unit (MAU) and a station on the ring. Because of this insertion by the testing apparatus into the token ring, the station remains on the ring since both the station and the testing apparatus utilize the same MAU port in communicating with other stations on the ring. The hardware includes a control network comprised of a field programmable gate array (FPGA) and switches. In connection with the insertion of the testing apparatus into the token ring, the control network is used in controlling the procedures and tests associated with a proper insertion. In that regard, an electrical connection is first made between the testing apparatus and the MAU, while the station is electrically disconnected from the MAU. A lobe test is conducted between the testing apparatus and the MAU. Depending upon the results of this test, a further test might be run to check the integrity of the cable conductor between the testing apparatus and the MAU. The testing apparatus also determines whether a beacon state exists on the token ring, as well as determining whether the data transmission rates for the station and the ring match. Subsequent to these tests and checks, a lobe test is then conducted by the station after electrical connectivity is established between the station and the testing apparatus. If the station should fail this test, a check is made of the integrity of the cable conductor between the testing apparatus and the station. If the lobe test passes, the control network electrically connects the station to the MAU. Simultaneously, the control network also reconfigures itself so that the testing apparatus is electrically connected in series with the station and the MAU.

The testing apparatus also includes a measuring apparatus for use in measuring phase jitter on the token ring. Because of the insertion capability of the testing apparatus, the measuring apparatus is able to conduct phase jitter tests without disruption of normal token ring operation. That is, while the phase jitter test is being performed, normal ring traffic can continue. The measuring apparatus transmits frames having special data associated with conducting the phase jitter test. Accompanying the phase jitter test data is address information. Such address information includes the source address of the phase jitter test data and the destination address thereof. The testing apparatus address can be used as both the source and destination address. The measuring apparatus includes hardware for becoming the active monitor on the token ring. Once it has become active monitor, the measuring apparatus, such as the NIC of the testing apparatus, is able to transmit the phase jitter test pattern. The measuring apparatus also includes a state machine for determining when this pattern has returned to the testing apparatus from the token ring. In that regard, the state machine compares address information in the pattern with a predetermined pattern. The measuring apparatus includes a dual slope interpolation circuit for increasing the resolution of any difference between the token ring reference clock signal and the clock signal that accompanies the returned phase jitter test pattern. Phase jitter is a function of the difference between the rising or falling edge of the token ring reference clock signal and the corresponding rising or falling edge of the recovered clock signal. The interpolation circuit is useful in enhancing the determination of any difference between these two signals. The interpolation circuit communicates with an interpolation counter that monitors the time duration that is enhanced or interpolated by the interpolation circuit. The output of the interpolation counter is indicative of one sample of the phase jitter test. The measuring apparatus also includes a clamp circuit for use in ensuring that the interpolation process always begins at the same reference voltage. The clamp circuit controls an input to a comparator, which comparator is used in controlling the counting of pulses or the monitoring of time by the interpolation counter. The clamp circuit also includes circuitry for making sure that the comparator changes state when an interpolation cycle is completed whereby the interpolation counter does not erroneously continue to count pulses when the interpolation cycle has finished.

With regard to testing the integrity of cable conductors, the testing apparatus includes time delay reflectory (TDR) circuitry for matching its impedance with the impedance of the cable conductor. Impedances of cable conductors used in the token ring are commonly either 150 ohms or 100 ohms. The impedance matching circuitry includes a single impedance element such as a resistor, that can be changed or switchably included in the circuitry in order to match the cable conductor impedance. The single impedance matching element communicates with a directional coupler for enabling the circuitry to require only a single impedance element that is used in matching the cable conductor impedance.

In connection with conducting TDR tests on a cable conductor, the directional coupler transmits a TDR signal and, when a cable conductor fault exists, a reflected or returned signal is coupled in only one direction where it is detected by a detecting resistor. This capability also enables the TDR circuitry to test for faults in the cable conductor end that is directly connected to the testing apparatus without the need for extra or lengthy intermediate cable between the testing apparatus and the cable conductor being tested. That is, when a fault exists in the cable conductor end, this TDR related circuitry and particularly the directional coupler is able to receive the reflected pulse that occurs substantially simultaneously upon transmission of the inputted TDR test pulse. TDR circuitry also includes a masking device for obtaining information on cable conductor integrity when more than one fault is present. The masking device is able to take into account faults that may exist in the cable conductor upstream of another cable conductor fault.

The testing apparatus is also able to check whether or not it is transmitting data at a rate that corresponds to the operating speed of the token ring, as reflected by the frequency of the master or reference clock signal that accompanies frame traffic around the ring. The NIC of the testing apparatus includes two selectable clock signals having different frequencies. Each of the two frequencies is based upon one of the two data transmission rates that are used with typical token rings. These token ring operating speeds are either 16 Mbps or 4Mbps and the clock frequencies are two times these rates, i.e. 32 MHz and 8 MHz, respectively. When the testing apparatus is inserted into the token ring, it selects one of its two clock signals in connection with receiving frames and/or tokens from the token ring. The FPGA is used in determining whether a frame or token from the token ring is accurately recovered using the selected clock signal. In particular, the FPGA includes a hardware state machine that generates control signals when frames or tokens are recovered. When such control signals are not generated, this is an indication that the selected clock signal is incorrect and the other clock signal should be selected for use by the testing apparatus. The testing apparatus processor monitors the time during which the testing apparatus is determining whether or not the correct clock frequency was selected. If a predetermined time passes before a determination is made that the current selected clock frequency is correct, the processor initiates a process by which the other of the two clock signals is selected for accompanying data transmitted by the testing apparatus into the token ring. In one embodiment, the processor modifies one or more bits in a control register, whose contents are monitored by the NIC. Depending upon the state of the bit or bits, one of the two available clock signals is selected.

The testing apparatus also includes hardware and software for providing station history information. Such information relates to the presence or absence of each station during a measurement interval. The history information is obtained while relying upon conventional MAC protocol. In particular, such protocol includes a periodic conducted neighbor notification process during which addresses of all token ring stations that are present or active on the ring are obtainable. The hardware of the present invention includes a last station memory list that contains addresses of all stations that were present on the token ring during a previous neighbor notification process. A current station memory list is also provided and contains addresses of all token ring stations that are present on the ring during the current neighbor notification process. By comparing each station in the current list with each station in the last list, the status or history of token ring stations can be continuously obtained and updated. In addition to address information in the two memory lists, each station address is accompanied by a state character indicating the history of the particular token ring station in the context of its presence or absence in the ring. The address information and the accompanying state character are controllably viewable using a display unit.

Further useful information gathered by the testing apparatus relates to the expected domain or location of a token ring fault. In accordance with MAC protocol, a beacon state is generated in the token ring when a fault occurs. If a first station should experience a loss of signal, it will eventually generate a beacon frame that includes its address and the address of its nearest active upstream neighbor (NAUN). Such address information is useful to the network manager or user in isolating the fault since, in isolating a fault, it is logical to first check the station that originated the initial beacon frame. In addition to beacon frames generated by the station first experiencing the signal loss, further beacon frames might be generated by other stations on the ring. That is, because of steps taken by the first station in attempting to recover from its signal loss, downstream stations would also experience a signal loss and generate their own beacon frames. These frames include "false or phantom" domains since the apparent token ring fault associated with the first station and its NAUN does not involve such other stations. The present invention distinguishes a beacon frame from a first or originating station from those generated by downstream stations. In one embodiment, phantom domain information is ignored or not stored, while beacon information from the station that originated the first beacon frame is stored. In determining whether the beacon frame received by the testing apparatus is a first beacon frame or subsequent beacon frames including those containing phantom domain information, a software state machine analyzes the received beacon frames. In conjunction with this analysis, a status information memory storage is utilized. This memory storage indicates whether a beacon frame has been previously received for the particular ring fault and, if so, any beacon frame that is received subsequently can be ignored since information has already been obtained from a previous beacon frame that included address information related to the beacon domain. The testing apparatus also includes a memory communicating with the software state machine for storing beacon domain information including the source or station address of the station originating the beacon frame and its NAUN. Under testing apparatus processor control, a video display unit receives the beacon domain information and, when desired, displays the same without displaying other beacon domain information that includes phantom domains, or other information, that is not deemed useful to the network manager in isolating the token ring fault.

Based on the foregoing summary, a number of beneficial aspects of the present invention are seen. The testing apparatus is connectable into a token ring lobe cable in series between a regular station and the MAU of the token ring. In this configuration, the testing apparatus is able to non-intrusively monitor the insertion of the regular station with which it is in series. Consequently, the testing apparatus is able to monitor and test this regular station, which may be suspected as causing a token ring fault. If this regular station passes the tests conducted by the testing apparatus, it is successfully inserted into the ring, together with the testing apparatus. Because of this insertion capability, the testing apparatus does not require a separate MAU port in order to insert into the token ring. In conjunction with the insertion feature, the testing apparatus is able to measure phase jitter on the ring without causing the ring to be placed in a non-operational state. That is, the testing apparatus transmits traffic frames around the token ring that include phase jitter test data patterns, while other frames are also transmittable about the ring. With regard to measuring the phase jitter for a particular sample, a dual slope interpolation circuit enhances the resolution of data that is analyzed when measuring the phase jitter. Additionally, a clamp circuit is employed for making sure that the interpolating process is accurately started and terminated. The testing apparatus also checks the integrity of cable conductors using TDR technology. The testing apparatus utilizes directional coupler circuitry to detect cable conductor faults at the cable conductor end to which it is attached. The TDR hardware enables the testing apparatus to determine the magnitude and location of cable conductor faults. In matching the impedance of the measuring circuitry with the cable conductor impedance it is only necessary to include or modify a single impedance element. With respect to properly operating on the ring, the testing apparatus is able to check for and modify, if necessary, the rate at which it transmits frames into the token ring. As a consequence, token ring speed operation is achieved by the testing apparatus without creating a fault or beacon state. The testing apparatus maintains history information related to active stations on the token ring from the start of a measurement. The testing apparatus compares station address information utilizing the contents of current and last station memory lists. Such information includes the identity of new stations that have entered the token ring since the start of the measurement, stations that were present at the start of the measurement but subsequently left the ring and stations which previously left the ring but then returned. Station history information enables token ring troubleshooters to correlate topology changes with network fault conditions which helps them isolate the source of a problem. The testing apparatus also filters or ignores phantom beacon domains. Beacon domains associated with the station that originated the beaconing condition are displayable, while phantom domains are not. Because phantom beacon domains are not displayed, the troubleshooter or network manager is not misled by phantom beacon domain information in attempting to isolate a token ring fault. Furthermore, a user of a station downstream of the station originating the beacon frame, because of reliance on the originating beacon frames and not on phantom domains, will not be inclined to reboot or power off the downstream station since the user is aware that the downstream station is not the cause of the problem.

Additional advantages of the present invention will become readily apparent from the following discussion, particularly when taken together with the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block diagram of a token ring station;

FIG. 2 is a block diagram of a prior art media access unit (MAU);

FIG. 3 is a block diagram of a prior art token ring with a number of stations illustrating the physical disconnection of one of the stations so that a prior art tester is connectable to a MAU port;

FIG. 4 is a block diagram of the testing apparatus of the present invention electrically connected in series between a MAU port and a regular station on the ring;

FIG. 5 is a block diagram of the token ring and tester interconnections for the token ring of FIG. 4;

FIG. 6A is a block diagram illustrating the serial interconnection to a single MAU port by the testing apparatus and the regular station on the ring;

FIG. 6B is a block diagram showing further detail of the testing apparatus illustrated in FIG. 6A;

FIG. 6C is a block diagram showing even more detail of the testing apparatus illustrated in FIG. 6A;

FIG. 7 is a block diagram illustrating electrical connectivity between the MAU port and the testing apparatus in order to conduct desired tests therebetween;

FIG. 8 is a block diagram illustrating the electrical connectivity between the testing apparatus and the regular station in which the testing apparatus is de-inserted from the MAU port, but is configured to avoid further reinsertion procedures:

FIG. 9 is a block diagram illustrating the electrical connectivity between the testing apparatus and the regular station;

FIG. 10 is a block diagram illustrating the electrical connectivity involving the MAU port, the testing apparatus and the regular station;

FIG. 11 is a flow diagram of steps conducted by prior art testers in testing a token ring;

FIG. 12 a flow diagram illustrating steps taken by the testing apparatus in order to insert serially between a MAU port and a regular station;

FIG. 13 is a flow diagram illustrating steps related to tests conducted and error reporting that can occur when a fault is present during insertion of the testing apparatus into the token ring;

FIG. 14 is a block diagram of basic hardware elements of the testing apparatus useful in determining and modifying the clock signal that accompanies frames from the testing apparatus so that such inputted data conforms to token ring speed;

FIG. 15 is a flow diagram that illustrates the steps involved in the testing apparatus in order to check for and modify, if necessary, the rate at which it transmits data into the token ring;

FIG. 16 is a block diagram of the measuring apparatus for use in measuring phase jitter;

FIGS. 17A-17D are timing diagrams related to the interpolation process for enhancing the resolution of data useful in obtaining a phase jitter measurement;

FIG. 18 is a circuit schematic of the clamp circuit of FIG. 16;

FIG. 19 is a flow diagram illustrating steps taken by the testing apparatus, in one embodiment, in order to become the active monitor for use in sending phase jitter test patterns;

FIG. 26 is a flow diagram illustrating steps related to the determination of the presence of returned phase jitter test data using a state machine;

FIG. 21 is a block diagram of TDR related circuitry for use in testing the integrity of cable conductors;

FIG. 22 is a circuit schematic of the transmitter/receiver of FIG. 21 including an illustration of the directional coupler and the single element matching impedance;

FIGS. 23A-23B illustrate equivalent circuits related to the single element impedance that matches cable conductor impedance;

FIG. 24 is a circuit schematic that is similar to FIG. 22 and illustrates the inputting of a TDR test signal into the cable conductor:

FIG. 25 is a circuit schematic that is similar to FIG. 24, and illustrates the return of a reflected pulse indicative of a fault along the cable conductor;

FIG. 26 is a circuit schematic that is similar to FIG. 22 and illustrates application of the circuit in receiving a return pulse clue to a fault at the cable conductor end;

FIG. 27 is a block diagram illustrating hardware used in monitoring station history in the token ring;

FIGS. 28A-28B constitute a flow diagram illustrating steps conducted by the testing apparatus in determining the status of active stations on the ring;

FIGS. 29A-29C are block diagrams illustrating at different times during a particular measurement the identification of active stations on the ring;

FIGS. 30A-30E illustrate address information and accompanying state characters based on station history for the token ring configurations of FIGS. 27 and 29A-29C;

FIG. 31 is a block diagram illustrating hardware of the present invention used in storing beacon domain information from an originating station while ignoring other beacon domains including phantom domains;

FIG. 33 is a flow diagram illustrating steps taken by the testing apparatus in storing and displaying beacon domain information from an originating station while filtering out other beacon frames.

DETAILED DESCRIPTION

FIG. 4 illustrates one embodiment of the present invention. The testing apparatus 100 of the present invention contains circuitry to enable both testing apparatus 100 and station 50 to be inserted into the token ring local area network 104 through a single port on MAU 68. The connection of testing apparatus 100 between a station and MAU 68, will be referred to as the "active T" configuration.

FIG. 5 is a representation of the electrical connections between stations on the token ring 104 including the testing apparatus 100 connected in an active T configuration. As shown, testing apparatus 100 is able to connect both itself and station 50 in series with the other stations 108, 112, 116 in the token ring 104. The method and apparatus which allows testing apparatus 100 to be connected in an active T configuration will now be described in further detail.

FIG. 6A illustrates the physical connections of the testing apparatus 100, MAU 68, and station 50 in an active T configuration. As shown, the testing apparatus 100 includes a control network 120, a NIC 122, and processor 126. Two pairs of twisted wires 130 and 134 connect control network 120 to the receive and transmit (hereinafter Rx and Tx) connections of port 82 in MAU 68. Two additional pairs of twisted wires 138 and 142 connect control network 120 to the Rx and Tx connections of NIC 56 of station 50.

The control network 120 determines the electrical connections between the Tx and Rx wire pairs of NIC 56 and MAU 68 and testing apparatus 100. FIGS. 7-10 depict the electrical connections between MAU 68, testing apparatus 100 and NIC 56 as facilitated by switches within the control network 120 at various stages of the active tee insertion process. As shown, NIC 122 comprises both an analog interface 146 and a digital interface 150 which are connected by NIC multiplexer (MUX) 154.

FIG. 6B illustrates testing apparatus 100 in further detail. The control network 120 includes a field programmable gate array (FPGA) 160 that controls both switching network 164 and MUX 154. Both switching network 164 and MUX 154 are used to create the necessary configurations as shown in FIGS. 7-10. Both analog interface 146 and digital interface 150 include standard IC chips which must be included on any token ring NIC. In general, analog interface 146 controls the synchronization of the testing apparatus 100 to incoming data and clock signals, while digital interface 150 includes circuitry to maintain MAC protocol.

In order to understand the various stages of the active T insertion process, it is first necessary to understand the process by which a station normally inserts into a token ring network. To this end, prior art FIG. 11 is a flow chart detailing the normal insertion process for station 50 in which the NIC 56 is to be directly connected to the MAU 68. At step 200, station 50 (employing NIC 56) sends out 1,024 frames of a known bit pattern on its transmit line. This is known as a lobe test. A lobe test ensures the connectivity of components between station 50 and the ring. At this point, relays 84, 86 in MAU 68 are set such that the transmit and receive lines of station 50 are looped back upon one another. At step 204, the 1,024 frames are returned directly to the station 50 without traversing the token ring. At step 208, station 50 determines whether all 1,024 frames were returned without error. If an error is detected in any of the 1,024 returned frames the insertion process is aborted in step 212. Errors of this kind are normally caused by