At 07:42 AM 12/11/98 -0600, Mike Mayer wrote: > >[...] > >="Coding Scheme for Gigabit Ethernet >= PAM-5 (Pulse Amplitude Modulation 5 Levels) >= 2 Bits per symbol ( => 4 bit combinations or levels) >= Symbol Rate 125 Mbaud/sec >= Bit Rate per pair 250 Mbits/sec >= Bit Rate using 4 pairs 1000 Mbits/sec >= 5th coding level used for control and management >= Simultaneous Bi-Directional Transmission " >= > >I haven't read the spec, but I am confused in what I have read about >the data rate. Is it 250 Mbits/sec per pair in each direction >simultaneously? Some things I have read imply 1000Base-T is 500 Mbits/ >sec transmit + 500 Mbits/sec receive instead of 1000 Mbits/sec and >1000 Mbits/sec receive simultaneously.
Good call Mike. That one went right by me. Egg all over my face ... But the more I read this, the more I think they should get their units correct for bit rate, baud rate, data rate, transfer rate. The following is a shakedown of some notes I have. 1000BaseT is full duplex. It sends and receives data over four pairs from both ends of each pair at the same time. Each pair carries a full duplex of 250 Mb/s data. The data is encoded PAM 5. A 1000BaseT physical layer has 4 identical transceivers (receiver and transmitter). Each transceiver sends 250 Mbits/s. Since it's 2 bits per symbol, it's really 125 MBaud/sec. 4 transceivers operating at 250 Mbits/sec equals 1000 Mbits/sec BEFORE and AFTER the line coding/decoding. I'll explain why I emphasized this further below. 1000BaseT with PAM 5 was created to be compatible with 100BaseT. This allowed a dual data rate 100/1000BaseT transceiver to be developed. The baud rate of 1000BaseT equals 125 MBaud/sec. 100BaseT has the same baud rate. Thus the reason for why some people are convinced that the same CAT5 cable used for 100BaseT can also be used for 1000BaseT. Now, to explain why the emphasis above. Strictly speaking, by using PAM 5 line coding, there is no 1000 Mbits/sec total transferred on the cable itself. The 1000 Mbits/sec is before and after the coding. The PAM 5 coding turns it into a baud rate and is 125 Mbaud/sec for each line. There are no "bits" being sent down the line. So my concern about a contiguous on/off bit stream generating a square wave burst at 500 MHz is unfounded. --------- This message is coming from the emc-pstc discussion list. To cancel your subscription, send mail to [email protected] with the single line: "unsubscribe emc-pstc" (without the quotes). For help, send mail to [email protected], [email protected], [email protected], or [email protected] (the list administrators).
