IEEE
`Std 802.3u-1995
`
`SUPPLEMENT TO 802.3:
`
`MDI
`
`110 :2
`
`.
`E 402 9
`
`§ 402 :2
`
`TRANSMIT
`DEV CE
`UNDIER
`TEST
`
`PG
`
`Em
`
`*Resistor matching to 1 part in 100.
`
`23.5.1.3 Receiver specifications
`
`The PMA shall provide the Receive function specified in 23.4.1.3 in accordance with the electrical specifica-
`tions of this clause. The patch cables and interconnecting hardware used in test configurations shall meet
`Category 5 specifications as in ISO/IEC 11801: 1995.
`
`The term worst-case UTP model, as used in this clause, refers to lumped—element cable model shown in fig-
`ure 23-23 that has been developed to simulate the attenuation and group delay characteristics of 100 m of
`worst-case Category 3 PVC UTP cable.
`
`This constant resistance filter structure has been optimized to best match the following amplitude and group
`delay characteristics, where the argumentfis in hertz, and the argument x is the cable length in meters. For
`the worst-case UTP model, argument x was set to 100 m, and the component values determined for a best
`least mean squared fit of both real and imaginary parts of H(fi x) over the frequency range 2 to 15 MHz.
`
`NOTE—This group delay model is relative and does not includes the fixed delay associated with 100 111 of Category 3
`cable. An additional 570 ns of fixed delay should be added in order to obtain the absolute group delay.
`
`PropagationImag(f, x) = j(—l0)
`
`Pr0pagationReal(f, x) = —(7.1 i% + 0.701L()6)(%)
`
`PropagationImag(f, x) + Propagatz'0nReal(f, x)
`20
`
`H(f,x) = 10
`
`23.5.1.3.1 Receiver differential input signals
`
`Differential signals received on the receive inputs that were transmitted within the constraints of 23.5.1.2,
`and have then passed through a worst-case UTP model, shall be correctly translated into one of the
`PMA_UNITDATA.ir1dicate messages and sent to the PCS. In addition, the receiver, when presented with a
`
`This is anlegrchive IEEE Standard.
`
`It has been superseded by a later version of this standard.
`
`bit 1025
`
`
`
`Aerohive - Exhibit 1025
`0140
`
`

`
`CSMA/CD
`
`IEEE
`Std 802.3u-1995
`
`R13
`50
`
`R14
`
`50
`
`
`Receive side
`
`L's :I: 10%
`C's 1: 5%
`R's 11%
`
`link test pulse generated according to the requirements of 23.4. 1 .2 and followed by at least 3T of silence on
`pair RX_D2, shall accept it as a link test pulse.
`
`Both data and link test pulse receive features shall be tested in at least two configurations: using the worst-
`case UTP model, and with a connection less than one meter in length between transmitter and receiver.
`
`A receiver is allowed to discard the first received packet after the transition into state LINK_PASS, using
`that packet for the purpose of fine-tuning its receiver equalization and clock recovery circuits.
`
`NOTE—Implementors may find it practically impossible to meet the requirements of this subclause without using some
`form of adaptive equalization.
`
`23.5.1.3.2 Receiver differential noise immunity
`
`The PMA, when presented with 8B6T encoded data meeting the requirements of 23.5.1.3.1, shall translate
`this data into PMA_UNITDATA.indicate (DATA) messages with a bit loss of no more than that specified in
`23.4.1.3.
`
`The PMA Carrier Sense function shall not set pma_carrier=ON upon receiving any of the following signals
`on pair RX_D2 at the receiving MDI, as measured using a l00BASE-T4 transmit test filter (23.5. 1 .2.3):
`
`a) All signals having a peak magnitude less than 325 mV.
`b) All continuous sinusoidal signals of amplitude less than 8.7 V peak-to-peak and frequency less than
`1.7 MHZ.
`
`c) All sine waves of single cycle or less duration, starting with phase 0° or 180°, and of amplitude less
`than 8.7 V peak-to-peak, where the frequency is between 1.7 MHz and 15 MHz. For a period of
`7 BT before and after this single cycle, the signal shall be less than 325 mV.
`
`This is an Archive IEEE Standard.
`
`It has been superseded by a later version of this stanqigrd.
`
`bit 1025
`
`
`
`Aerohive - Exhibit 1025
`0141
`
`

`
`IEEE
`Std 802.3u-1995
`
`SUPPLEMENT TO 802.3:
`
`cl)
`e)
`
`Fast link pulse burst (FLP burst), as defined in clause 28.
`The link integrity test pulse signal TP_IDL_l00.
`
`23.5.1.3.3 Receiver differential input impedance
`
`The difierential input impedance as measured at the MDI for each receive input shall be such that any reflec-
`tion due to differential signals incident upon each receive input from a balanced cable having an impedance
`of 100 S2 is at least 17 dB below the incident signal, over the fiequency range of 2.0 MHz to 12.5 MHz. This
`return loss shall be maintained at all times when the PHY is firlly powered.
`
`With each receiver connected as in figure 23-19, and with the source adjusted to simulate eopl and eop4
`(50% duty cycle square wave with 3.5 V amplitude, period of 480 ns, and risetime of 20 ns or faster), the
`amount of droop on each receive pair as defined in figure 23-18 shall not exceed 6.0%.
`
`23.5.1 .3.4 Common-mode rejection
`
`While receiving packets from a compliant l00BASE—T4 transmitter connected to all MDI pins, a receiver
`shall send the proper PMA_UNITDATA.indicate messages to the PCS for any differential input signal Es
`that results in a signal Edif that meets 23.5.1.3.1 even in the presence of common-mode voltages Em
`(applied as shown in figure 23-24). Ecm shall be a 25 V peak—to—peak square wave, 500 kHz or lower in fre-
`quency, with edges no slower than 4 ns (20%—80%), connected to each of the receive pairs RX_D2, BI_D3,
`and BI D4.
`
`“'0'
`
`RECEIVE
`DEVICE
`UNDER
`TEST
`
`
`
`* Resistor matching to 1 part in 1000.
`
`23.5.1.3.5 Receiver fault tolerance
`
`The receiver shall tolerate the application of short circuits between the leads of any receive input for an
`indefinite period of time without damage and shall resume normal operation after such faults are removed.
`Receivers shall withstand without damage a 1000 V common-mode impulse of either polarity (Ecm as indi-
`cated in figure 23-25). The shape of the impulse shall be 0.3/50 us (300 ns virtual front time, 50 us virtual
`time of half value), as defined in IEC 60.
`
`23.5.1.3.6 Receiver frequency tolerance
`
`The receive feature shall properly receive incoming data with a ternary symbol rate within the range
`25.000 MHz i 0.01%.
`
`This is anlétrchive IEEE Standard.
`
`It has been superseded by a later version of this standard.
`
`bit 1025
`
`
`
`Aerohive - Exhibit 1025
`0142
`
`

`
`CSMA/CD
`
`IEEE
`Std 802.3u-1995
`
`Mm
`
`RECEIVE
`DEVICE
`UNDER
`
`TEST
`
`§ 402 :2’
`49.9 Q
`-v\,—o
`
`110 Q
`
`Ecm .
`
`§4o2 9*
`
`I||—O8
`
`* Resistor matching to 1 part in 100.
`
`
`
`23.5.2 Power consumption
`
`Afier 100 ms following PowerOn, the current drawn by the PHY shall not exceed 0.75 A when powered
`through the M11.
`
`The PHY shall be capable of operating from all voltage sources allowed by clause 22, including those cur-
`rent limited to 0.75 A, as supplied by the DTE or repeater through the resistance of all permissible MII
`cables.
`
`The PHY shall not introduce extraneous signals on the M11 control circuits during normal power-up and
`power-down.
`
`While in power-down mode the PHY is not required to meet any of the 100BASE-T4 performance require-
`ments.
`
`23.6 Link segment characteristics
`
`23.6.1 Cabling
`
`Cabling and installation practices generally suitable for use with this standard appear in ISO/IEC
`11801: 1995. Exceptions, notes, and additional requirements are as listed below.
`
`a)
`b)
`
`c)
`
`100BASE-T4 uses a star topology. Horizontal cabling is used to connect PHY entities.
`100BASE-T4 is an ISO/IEC 11801: 1995 class C application, with additional installation require-
`ments and transmission parameters specified in 23.6.2 through 23.6.4. The highest fundamental fre-
`quency transmitted by 8B6T coding is 12.5 MHz. The aggregate data rate for three pairs using 8B6T
`coding is 100 Mb/s.
`100BASE-T4 shall use four pairs of balanced cabling, Category 3 or better, with a nominal charac-
`teristic impedance of 100 Q.
`d) When using Category 3 cable for the link segment, clause 23 recommends, but does not require, the
`use of Category 4 or better connecting hardware, patch cords and jumpers. The use of Category 4 or
`better connecting hardware increases the link segment composite NEXT loss, composite ELFEXT
`loss and reduces the link segment insertion loss. This lowers the link segment crosstalk noise, which
`in turn decreases the probability of errors.
`The use of shielded cable is outside the scope of this standard.
`
`e)
`
`This is an Archive IEEE Standard.
`
`It has been superseded by a later version of this stanqigrd.
`
`bit 1025
`
`
`
`Aerohive - Exhibit 1025
`0143
`
`

`
`IEEE
`Std 802.3u-1995
`
`SUPPLEMENT TO 802.3:
`
`23.6.2 Link transmission parameters
`
`Unless otherwise specified, link segment testing shall be conducted using source and load impedances of
`100 Q.
`
`23.6.2.1 Insertion loss
`
`The insertion loss of a simplex link segment shall be no more than 12 dB at all frequencies between 2 and
`12.5 lVlHz. This consists of the attenuation of the twisted pairs, connector losses, and reflection losses due to
`impedance mismatches between the various components of the simplex link segment. The insertion loss
`specification shall be met when the simplex link segment is terminated in source and load impedances that
`satisfy 23.5.1.2.4 and 23.5.l.3.3.
`
`NOTE—The loss of PVC-insulated cable exhibits significant temperature dependence. At temperatures greater than
`40 °C, it may be necessary to use a less temperature-dependent cable, such as many Fluorinated Ethylene Propylene
`(FEP), Polytetrafluoroethylene (PTFE), or Perfluoroalkoxy (PFA) plenurn-rated cables.
`
`23.6.2.2 Differential characteristic impedance
`
`The magnitude of the differential characteristic impedance of a 3 m length of twisted pair used in a simplex
`link shall be between 85 Q and 115 Q for all frequencies between 2 MHz and 12.5 MHz.
`
`23.6.2.3 Coupling parameters
`
`In order to limit the noise coupled into a simplex link segment from adjacent simplex link segments, Near-
`End Crosstalk (NEXT) loss and Equal Level Far-End Crosstalk (ELFEXT) loss are specified for each sim-
`plex link segment. In addition, since three simplex links (TX_Dl, Bl_D3, and Bl_D4) are used to send data
`between PHYS and one simplex link (RX_D2) is used to carry collision information as specified in 23.1.4,
`Multiple-Disturber NEXT loss and Multiple-Dist11rber ELFEXT loss are also specified.
`
`23.6.2.3.1 Differential Near-End Crosstalk (NEXT) loss
`
`The differential Near-End Crosstalk (NEXT) loss between two simplex link segments is specified in order to
`ensure that collision information can be reliably received by the PHY receiver. The NEXT loss between each of
`the three data carrying simplex link segments and the collision sensing simplex link segment shall be at least
`24.5 — 15><log10(f/12.5) (wherefis the fi'equency in MHz) over the frequency range 2.0 MHz to 12.5 MHz.
`
`23.6.2.3.2 Multiple-disturber NEXT (MDNEXT) loss
`
`Since three simplex links are used to send data between PHYS and one simplex link is used to carry collision
`information, the NEXT noise that is coupled into the collision, sensing simplex link segment is from multi-
`ple (three) signal sources, or disturbers. The MDNEXT loss between the three data carrying simplex link
`segments and the collision sensing simplex link segment shall be at least 21.4 — l5xlog10(f/ 12.5) dB (where
`f is the frequency in MHz) over the frequency range 2.0 to 12.5 MHz. Refer to 12.7.3.2 and Appendix A3,
`Example Crosstalk Computation for Multiple Disturbers, for a tutorial and method for estimating the MDN-
`EXT loss for an n-pair cable.
`
`23.6.2.3.3 Equal Level Far-End Crosstalk (ELFEXT) loss
`
`Equal Level Far-End Crosstalk (ELFEXT) loss is specified in order to limit the crosstalk noise at the far end of
`a simplex link segment to meet the BER objective specified in 23.1.2 and the noise specifications of 23.6.3.
`Far-End Crosstalk (FEXT) noise is the crosstalk noise that appears at the far end of a simplex link segment
`which is coupled from an adjacent simplex link segment with the noise source (transmitters) at the near end.
`ELFEXT loss is the ratio of the data signal to FEXT noise at the output of a simplex link segment (receiver
`input). To limit the FEXT noise fi'om adjacent simplex link segments, the ELFEXT loss between two data car-
`
`This is anlegrchive IEEE Standard.
`
`It has been superseded by a later version of this standard.
`
`bit 1025
`
`
`
`Aerohive - Exhibit 1025
`0144
`
`

`
`CSMA/CD
`
`IEEE
`Std 802.3u-1995
`
`rying simplex link segments shall be greater than 23.1 —20><1og10(f/12.5) dB (where f is the frequency in
`MHz) over the frequency range 2.0 lVlHz to 12.5 MHz. ELFEXT loss at frequencyfand distance I is defined as
`
`V
`ELFEXT_Loss ()’,l) = 20 X loglo (ids) — SLS_Loss (dB)
`Vpm
`
`where
`
`Vpds
`Vpcn
`SLS_Loss
`
`is the peak voltage of disturbing signal (near—end transmitter)
`is the peak crosstalk noise at the far end of disturbed simplex link segment
`is the insertion loss of the disturbing simplex link segment
`
`23.6.2.3.4 Multiple-disturber ELFEXT (MDELFEXT) loss
`
`Since three simplex links are used to transfer data between PHYS, the FEXT noise that is coupled into an data
`carrying simplex link segment is fi'om multiple (two) signal sources, or disturbers. The MDELFEXT loss
`between a data carrying simplex link segment and the other two data carrying simplex link segments shall be
`greater than 20.9 — 20><log10(f/ 12.5) (where fis the frequency in MHz) over the frequency range 2.0 lV[I-Iz to
`12.5 lVlHz. Refer to 12.7.3.2 and Appendix A3, Example Crosstalk Computation for Multiple Disturbers, for a
`tutorial and method for estimating the MDELFEXT loss for an n-pair cable.
`
`23.6.2.4 Delay
`
`Since T4 sends information over three simplex link segments in parallel, the absolute delay of each and the
`differential delay are specified to comply with network round-trip delay limits and ensure the proper decod-
`ing by receivers, respectively.
`
`23.6.2.4.1 Maximum link delay
`
`The propagation delay of a simplex link segment shall not exceed 570 ns at all flequencies between 2.0 lVlHz
`and 12.5 NH-Iz.
`
`23.6.2.4.2 Maximum link delay per meter
`
`The propagation delay per meter of a simplex link segment shall not exceed 5.7 ns/m at all frequencies
`between 2.0 MHz and 12.5 MHz.
`
`23.6.2.4.3 Difference in link delays
`
`The difference in propagation delay, or skew, under all conditions, between the fastest and the slowest sim-
`plex link segment in a link segment shall not exceed 50 ns at all frequencies between 2.0 MHz and
`12.5 MHz. It is a further functional requirement that, once installed, the skew between all pair combinations
`due to environmental conditions shall not vary more than i 10 ns, within the above requirement.
`
`23.6.3 Noise
`
`The noise level on the link segments shall be such that the objective error rate is met. The noise environment
`consists generally of two primary contributors: self—induced near—end crosstalk, which affects the ability to
`detect collisions, and far-end crosstalk, which affects the signal-to-noise ratio during packet reception.
`
`This is an Archive IEEE Standard.
`
`It has been superseded by a later version of this stanqigrd.
`
`bit 1025
`
`
`
`Aerohive - Exhibit 1025
`0145
`
`

`
`IEEE
`Std 802.3u-1995
`
`23.6.3.1 Near-End Crosstalk
`
`SUPPLEMENT TO 802.3:
`
`The MDNEXT (Multip1e—Disturber Near-End Crosstalk) noise on a link segment depends on the level of the
`disturbing signals on pairs TX_D1, BI_D3, and BI_D4, and the crosstalk loss between those pairs and the
`disturbed pair, RX_D2.
`
`The MDNEXT noise on a link segment shall not exceed 325 mVp.
`
`This standard is compatible with the following assumptions:
`
`a)
`
`b)
`
`c)
`
`Three disturbing pairs with 99th percentile pair—to—pair NEXT loss greater than 24.5 dB at 12.5 lVlHz
`(i.e., Category 3 cable).
`Six additional disturbers (2 per simplex link) representing connectors at the near end of the link seg-
`ment with 99th percentile NEXT loss greater than 40 dB at 12.5 MHZ (i.e., Category 3 connectors
`installed in accordance with 23.6.4.1).
`All disturbers combined according to the MDNEXT Monte Carlo procedure outlined in Appendix A3,
`Example Crosstalk Computation for Multiple Disturbers.
`
`The MDNEXT noise is defined using three maximum level 100BASE-T4 transmitters sending uncorrellated
`continuous data sequences while attached to the simplex link segments TX_Dl, BI_D3, and BI_D4 (disturb-
`ing links), and the noise measured at the output of a filter connected to the simplex link segment RX_D2.
`(disturbed link). Each continuous data sequence is a pseudo-random bit pattern having a length of at least
`2047 bits that has been coded according to the 8B6T coding rules in 23.2.1.2. The filter is the 100BASE-T4
`Transmit Test Filter specified in 23.5 .1.2.3.
`
`23.6.3.2 Far-End Crosstalk
`
`The MDFEXT (Multiple-Disturber Far-End Crosstalk) noise on a link segment depends on the level of the
`disturbing signals on pairs TX_Dl, BI_D3, and BI_D4, and the various crosstalk losses between those pairs.
`
`The MDFEXT noise on a link segment shall not exceed 87 mVp.
`
`This standard is compatible with the following assumptions:
`
`a)
`
`Two disturbing pairs with 99th percentile ELFEXT (Equal Level Far-End Crosstalk) loss greater
`than 23 dB at 12.5 MHz.
`
`b) Nine additional disturbers (three per simplex link) representing connectors in the link segment with
`99th percentile NEXT loss greater than 40 dB at 12.5 MHz.
`All disturbers combined according to the MDNEXT Monte Carlo procedure outlined in Appendix A3,
`Example Crosstalk Computation for Multiple Disturbers.
`
`c)
`
`The MDFEXT noise is defined using two maximum level 100BASE-T4 transmitters sending uncorrellated
`continuous data sequences while attached to two simplex link segments (disturbing links) and the noise mea-
`sured at the output of a filter connected to the far end of a third simplex link segment (disturbed link). Each
`continuous data sequence is a pseudo-random bit pattern having a length of at least 2047 hits that has been
`coded according to the 8B6T coding rules in 23.2.1.2. The filter is the 100BASE-T4 Transmit Test Filter
`specified in 23.5.1.2.3.
`
`This is anlegchive IEEE Standard.
`
`It has been superseded by a later version of this standard.
`
`bit 1025
`
`
`
`Aerohive - Exhibit 1025
`0146
`
`

`
`CSMA/CD
`
`23.6.4 Installation practice
`
`23.6.4.1 Connector installation practices
`
`IEEE
`Std 802.3u-1995
`
`The amount of untwisting in a pair as a result of termination to connecting hardware should be no greater
`than 25 mm (1.0 in) for Category 3 cables. This is the same value recommended in ISO/IEC 11801: 1995 for
`Category 4 connectors.
`
`23.6.4.2 Disallow use of Category 3 cable with more than four pairs
`
`Jumper cables, or horizontal runs, made from more than four pairs of Category 3 cable are not allowed.
`
`23.6.4.3 Allow use of Category 5 jumpers with up to 25 pairs
`
`Jumper cables made from up to 25 pairs of Category 5 cable, for the purpose of mass—terminating port con-
`nections at a hub, are allowed. Such jumper cables, if used, shall be limited in length to no more than 10 In
`total.
`
`23.7 MDI specification
`
`This clause defines the MDI. The link topology requires a crossover function between PMAs. Implementa-
`tion and location of this crossover are also defined in this clause.
`
`23.7.1 MDI connectors
`
`Eight-pin connectors meeting the requirements of section 3 and figures 1-5 of IEC 603-7: 1990 shall be used
`as the mechanical interface to the balanced cabling. The plug connector shall be used on the balanced
`cabling and the jack on the PHY. These connectors are depicted (for informational use only) in figures 23-26
`and 23-27. The table 23-6 shows the assignment of PMA signals to connector contacts for PHYs with and
`without an internal crossover.
`
`pin 1
`
`23.7.2 Crossover function
`
`It is a functional requirement that a crossover function be implemented in every link segment. The crossover
`function connects the transmitters of one PHY to the receivers of the PHY at the other end of the link seg-
`ment. Crossover functions may be implemented internally to a PHY or elsewhere in the link segment. For a
`PHY that does not implement the crossover function, the MDI labels in the last colunm of table 23-4 refer to
`its own internal circuits (second column). For PHYs that do implement the internal crossover, the MDI
`labels in the last column of table 23-4 refer to the internal circuits of the remote PHY of the link segment.
`
`This is an Archive IEEE Standard.
`
`It has been superseded by a later version of this stanqlgrd.
`
`bit 1025
`
`
`
`Aerohive - Exhibit 1025
`0147
`
`

`
`IEEE
`Std 802.3u-1995
`
`SUPPLEMENT TO 802.3:
`
`Table 23-6—MDl connection and labeling requirements
`
`
`
`Additionally, the MDI connector for a PHY that implements the crossover function shall be marked with the
`graphical symbol “X”. Internal and external crossover functions are shown in figure 23-28. The crossover
`fimction specified here for pairs TX_D1 and RX_D2 is compatible with the crossover fimction specified in
`14.5.2 for pairs TD and RD.
`
`When a link segment connects a DTE to a repeater, it is recommended the crossover be implemented in the
`PHY local to the repeater. If both PHYs of a link segment contain internal crossover functions, an additional
`external crossover is necessary. It is recommended that the crossover be visible to an installer from one of
`the PHYS. When both PHYS contain internal crossovers, it is further recommended in networks in which the
`topology identifies either a central backbone segment or a central repeater that the PHY furthest from the
`central element be assigned the external crossover to maintain consistency.
`
`Implicit implementation of the crossover function within a tvvisted-pair cable, or at a wiring panel, while not
`expressly forbidden, is beyond the scope of this standard.
`
`23.8 System considerations
`
`The repeater unit specified in clause 27 forms the central unit for interconnecting 100BASE—T4 tvvisted-pair
`links in networks of more than two nodes. It also provides the means for connecting 100BASE—T4 twisted-
`pair links to other 100 Mb/s baseband segments. The proper operation of a CSMA/CD network requires that
`network size be limited to control round-trip propagation delay as specified in clause 29.
`
`23.9 Environmental specifications
`
`23.9.1 General safety
`
`All equipment meeting this standard shall conform to IEC 950: 1991.
`
`This is anlggchive IEEE Standard.
`
`It has been superseded by a later version of this standard.
`
`bit 1025
`
`
`
`Aerohive - Exhibit 1025
`0148
`
`

`
`CSMA/CD
`
`IEEE
`Std 802.3u-1995
`
`1 TX_D1+
`_[: 2 TX_D1—
`3 RX_D2+
`6 RX_D2—
`it
`%4 Bl D3+
`:%7 B|_D4+
`
`5 B|_D3—
`
`8 B|_D4—
`PHY
`
`TX_D1+
`1
`TX_D1— 2:]_
`RX_D2+ 3
`I:
`RX_D2— 5
`B|_D3+ 4g
`B|_D4+ 7g
`
`Bl_D3— 5
`
`B|_D4— 3
`PHY
`
`a) Two PHYs with external crossover function
`
`MD|-X Label
`MD|
`1
`TX_D1+ — TX_D1+ 1
`_l;2 TX_D1— Z TX_D1— 2
`3 RX_D2+: RX_D2+ 3
`_<]:5 Rx_D2—j RX_D2— 6
`4 BI_D3+ — Bl_D3+
`4
`
`5
`
`Bl_D3— j Bl_D3— 5
`
`%7 Bl_D4+ : Bl_D4+
`
`5
`
`Bl_D4— — Bl_D4— 8
`
`7
`
`Internal Signal
`
`
`
`b) PHY with internal crossover function
`
`23.9.2 Network safety
`
`This clause sets forth a number of recommendations and guidelines related to safety concerns; the list is nei-
`ther complete nor does it address all possible safety issues. The designer is urged to consult the relevant
`local, national, and international safety regulations to ensure compliance with the appropriate requirements.
`
`LAN cable systems described in this clause are subject to at least four direct electrical safety hazards during
`their installation and use. These hazards are as follows:
`
`Direct contact between LAN components and power, lighting, or communications circuits
`a)
`Static charge buildup on LAN cables and components
`b)
`c) High-energy transients coupled onto the LAN cable system
`d)
`Voltage potential differences between safety grounds to which various LAN components are
`connected
`
`This is an Archive IEEE Standard.
`
`It has been superseded by a later version of this standard.
`
`bit 1025
`
`
`
`Aerohive - Exhibit 1025
`0149
`
`

`
`IEEE
`Std 802.3u-1995
`
`SUPPLEMENT TO 802.3:
`
`Such electrical safety hazards must be avoided or appropriately protected against for proper network instal-
`lation and performance. In addition to provisions for proper handling of these conditions in an operational
`system, special measures must be taken to ensure that the intended safety features are not negated during
`installation of a new network or during modification or maintenance of an existing network.
`
`23.9.2.1 Installation
`
`It is a mandatory fimctional requirement that sound installation practice, as defined by applicable local codes
`and regulations, be followed in every instance in which such practice is applicable.
`
`23.9.2.2 Grounding
`
`Any safety grounding path for an externally connected PHY shall be provided through the circuit ground of
`the MH connection.
`
`WARNING—It is assumed that the equipment to which the PHY is attached is properly grounded, and not left floating
`nor serviced by a “doubly insulated, ac power distribution system.” The use of floating or insulated equipment, and the
`consequent implications for safety, are beyond the scope of this standard.
`
`23.9.2.3 Installation and maintenance guidelines
`
`It is a mandatory functional requirement that, during installation and maintenance of the cable plant, care be
`taken to ensure that noninsulated network cable conductors do not make electrical contact with unintended
`
`conductors or ground.
`
`23.9.2.4 Telephony voltages
`
`The use of building wiring brings with it the possibility of wiring errors that may connect telephony voltages
`to 100BASE-T4 equipment. Other than voice signals (which are low voltage), the primary voltages that may
`be encountered are the “battery” and ringing voltages. Although there is no universal standard, the following
`maximums generally apply.
`
`Battery voltage to a telephone line is generally 56 Vdc applied to the line through a balanced 400 9 source
`impedance.
`
`Ringing voltage is a composite signal consisting of an ac component and a dc component. The ac component is
`up to 175 V peak at 20 Hz to 60 Hz With a 100 9 source resistance. The dc component is 56 Vdc with a 300 Q
`to 600 9. source resistance. Large reactive transients can occur at the start and end of each ring interval.
`
`Although 100BASE-T4 equipment is not required to survive such wiring hazards without damage, applica-
`tion of any of the above voltages shall not result in any safety hazard.
`
`NOlI‘E—Wh1'ng errors may impose telephony voltages differentially across 100BASE-T4 transmitters or receivers.
`Because the termination resistance likely to be present across a receiver’s input is of substantially lower impedance than an
`off-hook telephone instrument, receivers will generally appear to the telephone system as off-hook telephones. Therefore,
`full-ring voltages will be applied for only short periods. Transmitters that are coupled using transformers will similarly
`appear like ofi‘-hook telephones (though perhaps a bit more slowly) due to the low resistance of the transformer coil.
`
`23.9.3 Environment
`
`23.9.3.1 Electromagnetic emission
`
`The twisted-pair link shall comply with applicable local and national codes for the limitation of electromag-
`netic interference.
`
`This is anlggrchive IEEE Standard.
`
`It has been superseded by a later version of this standard.
`
`bit 1025
`
`
`
`Aerohive - Exhibit 1025
`0150
`
`

`
`CSMA/CD
`
`23.9.3.2 Temperature and humidity
`
`IEEE
`Std 802.3u-1995
`
`The twisted—pair link is expected to operate over a reasonable range of environmental conditions related to
`temperature, humidity, and physical handling (such as shock and vibration). Specific requirements and val-
`ues for these parameters are considered to be beyond the scope of this standard.
`
`It is recommended that manufacturers indicate in the literature associated with the PHY the operating envi-
`ronmental conditions to facilitate selection, installation, and maintenance.
`
`23.10 PHY labeling
`
`It is recommended that each PHY (and supporting documentation) be labeled in a manner visible to the user
`with at least these parameters:
`
`a) Data rate capability in Mb/s
`b)
`Power level in terms of maximum current drain (for external PHYs)
`c) Any applicable safety warnings
`
`See also 23.7.2.
`
`23.11 Timing summary
`
`23.11.1 Timing references
`
`All MII signals are defined (or corrected to) the DTE end of a zero length MII cable.
`
`NOTE—With a finite length M11 cable, TX_CLK appears in the PHY one cable propagation delay earlier than at the
`MII. This advances the transmit timing. Receive timing is retarded by the same amount.
`
`The phrase adjustedfor pair skew, when applied to a timing reference on a particular pair, means that the
`designated timing reference has been adjusted by adding to it the difference between the time of arrival of
`preamble on the latest of the three receive pairs and the time of arrival of preamble on that particular pair.
`
`PMA_UNITDATA request
`
`Figures 23-29, 30, 31, and 32. The implementation of this abstract message is not specified.
`Conceptually, this is the time at which the PMA has been given full knowledge and use of the
`ternary symbols to be transmitted.
`
`PMA_UNITDATA.indicate
`
`Figure 23-33. The implementation of this abstract message is not specified. Conceptually, this is
`the time at which the PCS has been given full knowledge and use of the ternary symbols received.
`WAVEFORM
`
`Figure 23-29. Point in time at which output waveform has moved 1/2 way from previous nominal
`output level to present nominal output level.
`
`TX EN
`
`Figure 23-30. First rising edge of TX_CLK following the rising edge of TX_EN.
`
`NOT_TX_EN
`
`Figures 23-31 and 32. First rising edge of TX_CLK following the falling edge of TX_EN.
`
`CRS
`
`Figure 23-33. Rising edge of CRS.
`
`CARRIER_STATUS
`
`This is an Archive IEEE Standard.
`
`It has been superseded by a later version of this standard.
`
`bit 1025
`
`
`
`Aerohive - Exhibit 1025
`0151
`
`

`
`IEEE
`Std 802.3u-1995
`
`SUPPLEMENT TO 802.3:
`
`Figure 23-33. Rising edge of cam'er_status.
`
`NOT_CARRIER_STATUS
`
`Figure 23-34. Falling edge of carn'er_status.
`
`RX_DV
`
`COL
`
`No figure. First rising edge of RX_CLK following rising edge of RX_DV.
`
`No figure. Rising edge of COL signal at MII.
`
`NOT_COL
`
`No figure. Falling edge of COL signal at MII.
`
`PMA_ERROR
`
`No figure. Time at which rxerror_status changes to ERROR.
`
`This is anlgirrchive IEEE Standard.
`
`It has been superseded by a later version of this standard.
`
`bit 1025
`
`
`
`Aerohive - Exhibit 1025
`0152
`
`

`
`CSMA/CD
`
`IEEE
`Std 802.3u-1995
`
`23.11.2 Definitions of controlled parameters
`
`PMA_OUT
`
`Figure 23-29. Time between PMA_UNITDATA request (tx_code_vector) and the WAVEFORM
`timing reference for each of the three transmit channels TX_Dl, BI_D3, or BI_D4.
`
`TEN_PMA
`
`Figures 23-30, 31, and 32. Time between TX_EN timing reference and MA_UNITDATA request
`(tx_code_vector).
`
`TEN_CRS
`
`Figure 23-30. Time between TX_EN timing reference and the loopback of TX_EN to CRS as
`measured at the CRS timing reference point.
`
`NOT_TEN_CRS
`
`Figures 23-31 and 32. Time between NOT_TX_EN timing reference and the loopback of TX_EN
`to CRS as measured at the NOT_CRS timing reference point. 111 the event of a collision (COL is
`raised at any point during a packet) the minimum time for NOT_TEN_CRS may optionally be as
`short as 0.
`
`RX_PMA_CARR]ER
`
`Figure 23-33. Time between the WAVEFORM timing reference, adjusted for pair skew, of first
`pulse of a normal preamble (or first pulse of a preamble preceded by a link test pulse or a partial
`link test pulse) and the CARRIER_STATUS timing reference.
`
`RX_CRS
`
`Figure 23-33. Time between the WAVEFORM timing reference, adjusted for pair skew, of first
`pulse of a normal preamble (or first pulse of a preamble preceded by a link test pulse or a partial
`link test pulse) and the CRS timing reference.
`NOTE—The input waveform used for this test is an ordinary T4 preamble, generated by a compliant T4
`transmitter. As such, the delay between the first and third pulses of the preamble (which are used by the car-
`rier sense logic) is very nearly 80 ns.
`
`RX_NOT_CRS
`
`For a data packet, the time between the WAVEFORM timing reference, adjusted for pair skew, of
`the first pulse of eopl, and the de-assertion of CRS. For a collision fragment, the time between the
`WAVEFORM timing reference, adjusted for pair skew, of the ternary symbol on pair TX_D2,
`which follows the last ternary data symbol received on pair RX_D2, and the de-assertion of CRS.
`
`Both are limited to the same value. For a data packet, detection of the six ternary symbols of eopol
`is accomplished in the PCS layer. For a collision fragment, detection of the concluding seven
`ternary zeroes is accomplished in the PMA layer, and passed to the PCS in the form of the
`carrier_status indication.
`FAIRNESS
`
`The difference between RX_NOT_CRS at the conclusion of one packet and RX_CRS on a
`subsequent packet. The packets used in this test may arrive with an IPG anywhere in the range of
`80 to 160.
`
`RX_PMA_DATA
`
`Figure 23-33. Time between the WAVEFORM timing reference, adjusted for pair skew, of first
`pulse of a normal