HBCU-5710R 1000BASE-T Small Form Pluggable Low Voltage (3.3V) Electrical Transceiver over Category 5 Unshielded Twisted Pair Cable Characterization Report Application Note 5044 Summary The Physical Medium Attachment (PMA) electrical characteristics, link segment characteristics and Medium Dependent Interface (MDI) characteristics of HBCU- 5710R have been characterized from case temperature 0 C to +70 C and supply voltage 3.1V to 3.5V. The results show that the HBCU-5710R is compatible with the IEEE 802.3:2000 specifications. Introduction The HBCU-5710R 1000BASE-T SFP electrical transceiver is designed to provide an IEEE 802.3 compatible link for 1.25 Gb/s applications. It offers full duplex throughput of 1 Gb/s by transporting data over unshielded twisted pair category 5 cable with 5-level PAM (Pulse Amplitude Modulation) signals. The Avago 1000BASE-T module takes signals from both the twisted pair category 5 cable and the SerDes interface. It is intended for use over the case temperature range 0 C to +70 C. This report details the characterization of HBCU-5710R transceivers at case temperatures of 0 C, +25 C and +70 C and at supply voltages of 3.1V, 3.3V and 3.5V. Up to ten transceivers were utilized to collect data at various temperature and voltage combinations. Definition of terms The following acronyms and terms are used in this document. They are explained and defined below: BER Bit Error Rate CAT-5 Category 5 EMI ESD MAC MDI MSA NEXT RD SFP TD Electro-magnetic Interference Electrostatic Discharge Media Access Control Medium Dependent Interface Multi-Source Agreement Near End Crosstalk Receiver Data (output) Small Form Factor Pluggable Transmitter Data (input) Temperature The temperature specified is the HBCU-5710R SFP module case temperature measured on the EMI cage of the SFP evaluation board, unless otherwise specified. Voltage The voltage specified is measured at the 20-pin connector on the SFP evaluation board, which normally is about 0.1 volt lower than that of the power supply itself due to the voltage drop across the inductor of the SFP MSA power supply filter. Link Link test is the measure of BER of the module while transmitting and receiving data over a specified length of CAT- 5 cable. IEEE specification for the link performance is that BER should be less than over 100meter CAT-5 cable. For each link test, approximately 30 million packets (or about 20 billion bits) were sent. And several parameters were varied during link testing. They include temperature, voltage, cable length, packet size, utilization rate, input TD amplitude, and supply noise. Channel BER channel is defined as in IEEE 802.3:2000 clause 40.6.1.1.1. The test channel consists of a 1.2m 120ohm CAT-5 cable followed by a 103m 100ohm CAT-5 cable, then a 1.48m 120ohm CAT-5 cable, and lastly a 4m 100ohm CAT-5 cable. The specification is BER over 100m. Supply Current and Power Dissipation This is the current supplied to and the power dissipated by the module at the relevant supply voltage. They are measured under link up condition. Differential Output Templates The templates are described in IEEE 802.3:2000, section 40.6.1.2.3. As a module was configured to be in test mode 1 as described in IEEE 802.3:2000, section 40.6.1.2, the differential output voltage waveform around points A, B, C, D, F and H for each of the four module cable-side channels was tested for compliance to the templates.
Peak Differential Output Voltage and Level Accuracy As a module was configured to be in test mode 1 as described in IEEE 802.3:2000, section 40.6.1.2, this is the test for compliance of the peak differential output voltages at points A, B, C and D for each of the four module cable-side channels to IEEE 802.3:2000, section 40.6.1.2.1. The disturbing voltage described in IEEE 802.3:2000, section 40.6.1.1.3, was not used. Previous measurements have demonstrated that the disturbing voltage has no effect on the peak differential output voltage and level accuracy measurements. Maximum Output Droop As a module was configured to be in test mode 1, this is the test for compliance of the magnitude of the peak output voltages at points G and J as compared to that at points F and H, respectively, per IEEE 802.3:2000, section 40.6.1.2.2. All four cable-side channels were tested on each module. The disturbing voltage described in IEEE 802.3:2000, section 40.6.1.1.3, was not used. Previous measurements have demonstrated that the disturbing voltage has no effect on the droop measurements. Transmitter Distortion When a module is in test mode 4, this is the peak distortion on the differential signal output at the MDI as defined in IEEE 802.3:2000, section 40.6.1.2.4. Transmitter Timing Jitter There are six parts to the transmitter timing jitter per IEEE 802.3:2000, section 40.6.1.2.5. The first two are the output jitters for each channel relative to the master TX_TCLK when a module is in test mode 2 or test mode 3. For the next four, the module was linked up to a golden unit with a test channel. While the module is in master and slave modes, the unfiltered and filtered jitter of TX_TCLK was measured with reference to an external master 25 MHz clock. Common Mode Output Voltage The magnitude of the total common-mode output voltage on any transmit circuit as measured per IEEE 802.3:2000, section 40.8.3.3. Modules were configured to be in test mode 4. Common Mode Noise Rejection This is a measurement of a module s ability to withstand common-mode noise from the cabling system generally as a result of electromagnetic field subjection. Modules were tested per IEEE 802.3:2000, section 40.6.1.3.3. The result is shown via link BER. Alien Crosstalk Noise Rejection This is a measurement of a module s ability to withstand alien crosstalk that simulates a 100BASE-TX noise signal coupled into each individual channel on the CAT-5 cable. The level of the noise signal is 25 mv. Modules were tested per IEEE 802.3:2000, section 40.6.1.3.4. NEXT Noise Rejection This is a measurement of a module s ability to withstand crosstalk at the near end of a link segment, injected using a PCB that allows for an RC connection between each channel. Modules were tested per IEEE 802.3:2000, section 40.7.3.1.1 with result shown via link BER. MDI Impedance Balance This is a measurement of the impedance difference between the two MDI contacts to ground of each channel of a module. Modules were tested per IEEE 802.3:2000, section 40.8.3.2. ing was repeated with the module in idle and test mode 4. Transmitter Clock Frequency This is the transmitter clock frequency as tested per IEEE 802.3:2000, section 40.6.1.2.6. Modules were configured to be in test mode 2. Receiver Frequency Tolerance This is a measurement of the module s ability to properly receive incoming data over a frequency variation as specified per IEEE 802.3:2000, section 40.6.1.3.2. The frequency variation applied here was +/- 0.04% as compared to the specification of +/- 0.01%. Link BER was measured. MAC-side Frequency Tolerance This is a measurement of the module s tolerance to the frequency variation in the data going from the MAC to the module TD+/- lines as expressed in link BER. Link Delay Skew Tolerance This is a measurement of a module s tolerance to the difference in propagation delay, or skew, between all duplex channel pair combinations of a link segment, as determined by the link BER. Link delay skew tolerance was tested with module pairs. Each pair of modules was connected with a modified 103meter Cat5 cable, in which channel 2 (twist pair 4 and 5) had an extra delay of 63ns. This is beyond the IEEE specification in 802.3:2000 section 40.7.4.2, which is to tolerate 50ns delay skew with at most 10ns variation over environment.
MDI Crossover This is a measurement of the module s ability to automatically identify a crossover cable per IEEE 802.3:2000, sections 40.4.4 and 40.8.2. A crossover patch cable was inserted in addition to a non-crossover link cable between a pair of modules. The result is shown via link BER. MDI Fault Tolerance This is a test to determine a module s ability to withstand, without damage, any short circuit applied to any set of wires on the cable-side as per IEEE 802.3:2000, section 40.8.3.4. The various combinations of shorts were applied for 1 minute each. Normal operation was verified after the shorts were removed. Normal operation was considered to be error-free operation during a transmission of 30 million packets with random data and packet length and 100% utilization over 116m CAT-5 cable. MDI Return Loss This is a measurement of the attenuation of the differential signal reflection looking into the module from the cable-side as described in IEEE 802.3:2000, section 40.8.3.1. Modules were configured to be in test mode 4. In-Rush Current In-rush current is the initial peak supply current drawn by the module when the module was hot-plugged into a port. This current was measured through a 1.2 ohm sense resistor placed in series with the VCC. The SFP MSA allows a peak of 30mA above the steady state current. RD+/- Amplitude This is the differential RD+/- output voltage amplitude. It can be configured to its lowest and highest states, outside of the nominal state, through register writes. In its nominal state, the module is configured to output approximately 680-700 mv. The output amplitude for all three states was measured. RD+/- Rise/Fall Times This is the 20% to 80% rise and fall transition times for RD+/- outputs. They were measured with the RD+/- differential output amplitude configured to its nominal, lowest and highest states. Initialization Time Initialization time is the time from power-up until the module begins auto-negotiation with the host system. Reset Assert Time Reset assert time is the time from asserting TX_DISABLE until the module stops sending data or idle pattern to the host system on RD+/-. Reset Negate Time Reset negate time is the time from negating TX_DISABLE until the module begins auto-negotiation with the host system. Radiated Immunity Applying IEC 61000-4-3, this is a measurement of a module s ability to withstand a radiated field of magnitude 10V/m, modulated by 80% at 1 khz, and of frequency 80-1000 MHz. This test is done in a GTEM cell. During which time, data with random length, random packet size and 100% utilization rate were sent to and received from the module over a 103m cable. The BER of the module was monitored. Electrostatic Discharge (ESD) to RJ45 This is a test of tolerance for electrostatic discharge to the RJ45 connector receptacle as per variation of IEC 61000-4-2. NSG-435 model ESD gun was used to apply 40 consecutive pulse discharges, at the rate of one pulse per second. The measure of tolerance is a link recovery after removal of the electrical pulse. Data transmission is with 125 byte packet and 90% utilization rate. Isolation (Hi-Pot) This is the electrical isolation between the port device circuits, including frame ground and all MDI leads. Measure of compliance is described in IEEE 802.3:2000, section 40.6.1.1. MOD_DEF2 Output Levels The measured voltage referenced to ground at the MOD_ DEF2 output for both 1 and 0 states. RX_LOS Output Levels from PHY The measured voltage referenced to ground at the RX_ LOS output for both 1 and 0 states. This is for modules with RX_LOS enabled only. Modules with disabled RX_ LOS has the RX_LOS output tied to ground internally. Note: The following applies to all the tables here after. OT (Over Temperature) = 0/25/70 C; OV (Over Voltage) = 3.1/3.3/3.5 V; OP (Over Packet Size) = 64/125/255/512/1024 byte; OA (Over Differential TD Amplitude) = 0.4/0.8/2.4 V 3
Table 1. Link BER Link BER With Extreme TD Voltages Link BER With Supply Noise - 10Hz To 2MHz 70C, 3.3V, 120m, OP, 100% Utilization 70C, 3.3V, 120m, OP, 50% Utilization 70C, 3.3V, 120m, OP, 10% Utilization 70C, 3.3V, 116m, OP, 10% Utilization 70C, 3.3V, 103m, OP 10% Utilization 70C, 3.1V, 116m, 100% Utilization, 430mV/2500mV TD Voltages 70C, 3.1V, 120m, 100% Utilization, 200mV Supply Noise 70C, 3.1V, 116m, 100% Utilization, 200mV Supply Noise 70C, 3.1V, 116m, 10% Utilization, 200mV Supply Noise 2 7.8E-11 2 1.12E-10 ** 2 1.67E-10 ** 2 1.5E-11 2 5E-12* 10 1E-11* 3 1.2E-10 ** 1 5E-12* 1 5E-12 * No error detected. Error rate calculated with presumed error of 1. ** condition exceeding standard requirement.
Table 2. Supply Current and Power Dissipation Differential Output Templates (Points A, B, C, D, F, H) Peak Differential Output Voltage And Level Accuracy OT, OV, OA, 115m, (a) At 0C, 3.5V, Min TD input (b) At 70C, 3.1V, Max TD input (c) At 70C, 3.3V, Nominal TD input 10 ma/ W 276/ 0.97(a) 308/ 1.07(c) 364/ 1.13(b) OT, OV 10 Pass Pass - Point A voltage OT, OV 10 V 0.67 0.69 0.72 0.67 0.82 - Absolute value of Point B voltage Relative to Point A - Point C voltage relative to one-half the average of Point A and B (absolute value) voltages - Point D voltage relative to one-half the average of Point A and B (absolute value) voltages Maximum Output Droop - Amplitude at Point G relative to Point F - Amplitude at Point J relative to Point H 370/ 1.15 25C, 3.3V 8 % -1.93-0.32 1.23-2*** 2*** OT, OV 10 % -1.19 0.09 1.22-2 2 OT, OV 10 % -1.39-0.16 1.13-2 2 OT, OV 10 % 97.3 98.6 99.7 73.1 OT, OV 10 % 96.3 97.8 99.4 73.1 Transmitter Distortion 70C, 3.1V 3 mv 12.4 14.8 16.8 18**** *** The IEEE specification for this test is ±1%. The HBCU-5710R achieves ±2% at 25 C, 3.3V. Link degradation has not been observed due to this loosening of the specification. **** The IEEE specification for this test is 10mV. However, the distortion of the test system is about 10mV. The modules were considered to pass the specification if the measured transmitter distortion value is less than 18mV. 5
Table 3. - TX Output Jitter 1: 70C, 3.1V, Mode 2 2: 70C, 3.1V, Mode 3 - TX_TCLK Jitter 3: 70C, 3.1V, Master TX_TCLK referenced to 25MHz external clock Common Mode Output Voltage Common Mode Noise Rejection Alien Crosstalk Noise Rejection Near End Crosstalk (NEXT) Noise Rejection MDI Impedance Balance Transmitter Clock Frequency Relative to 125MHz 4 + 1 ( 4: Same as 3 but filtered) 5: 70C, 3.1V, recovered TX_TCLK referenced to 25MHz external clock 6 + 2-4 ( 6: Same as 5 but filtered) 3 ps 120 137 140 300 3 ps 120 136 150 300 3 ps 340 413 500 1400 3 ps 260 283 300 300 3 ps 440 533 720 1400 3 ps 120 156 190 400 25C, 3.3V 10 mv 32.7 36 42.3 50 70C, 3.1V, 100m, 100% Utilization, 2.8V Common Mode Voltage over frequency range of 1-250MHz OT, OV, 103m, 100% Utilization OT, OV, 103m, 100% Utilization 3 5.4E-12 10 5E-12* 10 5E-12* 70C, 3.1V 3 pass pass OT, 3.3V 10 % -0.0008-0.0001 0.0016-0.01 0.01 6
Table 4. MAC-side Frequency Tolerance Link Delay Skew Tolerance 70C, 3.1V, 116m, 64-byte Frame, 10B-encoded 1010 Pattern, 90% Utilization, +/- 0.01% Data Rate Variaton 70C, 3.1V, 103m, 90% Utilization MDI Crossover 25C, 3.3V, 116m, 100% Utilization MDI Fault Tolerance with Short s 1a) Ch (0,1)+ b) Ch (2,3)+ 2a) Ch (0,1)- b) Ch (2,3)- 3a) Ch (0,1,2,3)+ b) Ch (0,1,2,3)- 3 1E-11* 3 6.3E-12 10 5E-12* 3 Error free after fault removal 4uA with fault MDI Return Loss OT, OV 10 pass pass Error free Transmission after fault removal 300mA with fault In-Rush Current OT, 3.3V 3 ma 280 330 (30mA above link up steady state max current of around 300mA) RD+/- Differential Amplitude - Nominal output config OT, OV 10 mv 665 689 720 370 2000 - Low output config OT, OV 10 mv 444 457 479 370 2000 - High output config OT, OV 10 mv 1061 1219 1263 370 2000 7
Table 5. - Nominal output configuration - Low output configuration - High output configuration OT, OV 10 ps 91/95 107/112 128/131 250 OT, OV 10 ps 92/91 108/109 125/133 250 OT, OV 10 ps 93/99 107/115 142/151 250 Initialization Time 25C, 3.3V 10 ms 175 187 191 300 Reset Assert Time 25C, 3.3V 10 us 0.436 0.465 0.473 10 Reset Negate Time 25C, 3.3V 10 ms 170 175 180 300 Radiated Immunity 25C, 3.3V, 103m, 10V/m field ranging 80MHz to 1GHz Electrostatic Discharge (ESD) to RJ45 3 Error Free Inside switch box 2 kv 8 8 Isolation (Hi-Pot) 25C, 3.3V 200 ua pass 2.25kV DC for 60s MOD_DEF2 Output Levels - Voh OT, 3.3V 3 V 3.22 3.26 3.28 2.4 - Vol OT, 3.3V 3 V 0 0.02 0.03 0.4 RX_LOS Output Levels from PHY ***** - Voh OT, 3.3V 3 V 2.5 2.5 2.51 2.4 - Vol OT, 3.3V 3 V 0.012 0.012 0.013 0.4 ***** This is valid only for modules with RX_LOS enabled. Modules with disabled RX_LOS has the RX_LOS output tied to ground internally. For product information and a complete list of distributors, please go to our web site: www.avagotech.com Avago, Avago Technologies, and the A logo are trademarks of Avago Technologies, Limited in the United States and other countries. Data subject to change. Copyright 2007 Avago Technologies Limited. All rights reserved. 5989-0662EN - May 28, 2007