Jump to content

Talk:MLT-3 encoding

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

MLT-3

[edit]

People are constantly adding supposed reference to 100BASE-TX to this page. Section 24 of the 802.3-2002 (that deals with it) specifies that Ethernet PMA uses NRZI after 4B5B expansion. Where does MLT-3 comes from?

Good question. I've removed the reference to 100BASE-TX. Guy Harris 17:56, 15 June 2006 (UTC)[reply]
I've fallen on my face. Section 25.2: "The 100BASE-TX PMD (and MDI) is specified by incorporating the FDDI TP-PMD standard, ANSI X3.263: 1995 (TP-PMD), by reference, with the modifications noted below. This standard provides support for Category 5 unshielded twisted pair (UTP) and shielded twisted pair (STP). For improved legibility in this clause, ANSI X3.263: 1995 (TP-PMD), will henceforth be referred to as TP-PMD."oakad 07:37, 16 June 2006 (UTC)[reply]
The MLT-3 encoding, 100BASE-TX, and 10BASE-T#Variants articles are now consistent with each other: they all say that 100BASE-TX encodes the raw bits with 4B5B, and then encodes those bits with MLT-3. I hope we got it right this time :-). --DavidCary (talk) 16:50, 27 June 2013 (UTC)[reply]

Coding efficiency

[edit]

MLT-3 has a coding efficiency of 1 bit/baud - 100 Mbit/s 100BASE-TX after 4B5B expansion has symbol rate of 125 Mbaud and maximum fundamental frequency of 31.75 MHz. --93.136.111.10 (talk) 20:44, 20 November 2008 (UTC)[reply]

"coding efficiency of 1 bit/baud" - baud means symbol per second; so I would think it would be better to use bit/symbol, or, bit/s per baud here. Hoemaco (talk) 15:35, 31 August 2016 (UTC)[reply]

maximum fundamental frequency

[edit]

Shouldn't that be called highest/max. harmonic? As far as I remember, the fundamental is another name for first harmonic. —Preceding unsigned comment added by 80.219.166.161 (talk) 00:27, 8 April 2010 (UTC)[reply]

Yes, the fundamental is another name for the first harmonic.
Theory: As you may already know,
Any signal can be analyzed as if it were a musical tone.
When the spectrum of some repeating signal is analyzed, invariably the results show it can be dissected into a series of one or more simple tones (pure tones) -- a single fundamental tone at some frequency, and harmonics at integer multiples of that fundamental frequency.
The first harmonic (the fundamental frequency) is the *lowest* of all the harmonics.
However, different tones have different fundamental frequencies.
Practice:
If we find the fundamental frequency of each and every possible tone that can be produced with MLT-3 encoding, we find that the tone produced by transmitting a "1111111..." sequence has a fundamental frequency of one fourth (1/4) the baud rate. All other repeating signals that can be produced using MLT-3 have a lower fundamental frequency.
If we find the fundamental frequency of each and every possible tone that can be produced with non-return-to-zero, we find that the tone produced by transmitting "10101010..." has a fundamental frequency of one half (1/2) the baud rate. All other repeating signals that can be produced using NRZ have a lower fundamental frequency.
application to this article
The phrase "the maximum fundamental frequency" in this article is difficult to understand. However, replacing that phrase with "the highest lowest harmonic", or "the maximum minimum harmonic", while technically it means exactly the same thing, would only make the article more difficult to understand.
How could we improve this article to make it easier to understand? --DavidCary (talk) 17:03, 27 June 2013 (UTC)[reply]
Maybe by referring to bandwidth? As far as I understand, bandwidth is related to the first fundamental freq. For a normal unipolar nrz, if I know correctly usually the BW is considered approx the first fundamental (which is 1/2T_bit) (thus getting ideally 2bps per hertz), though in this case the signal will look far from a square wave, but theoretically it can be decoded. Or we can take the first null point in the spectrum, which for a simple nrz would be 1/T_bit. Though if this is correct, then the MLT-3 seems to violate the R/B<2N law (see Hartley's law) (probably the reason is in some definition issue). Can someone clarify this? Hoemaco (talk) 15:35, 31 August 2016 (UTC)[reply]


MLT-3 origin

[edit]

MLT-3 was invented by Mario Mazzola and Luca Cafiero who were two of the founders of Crescendo Communications. Crescendo was later acquired by Cisco and the patent was assigned to Cisco.

Jhamstra (talk) 20:04, 11 June 2010 (UTC)[reply]

For clarification, the third inventor was Maurilio DeNicolo who was also known as "Tazio". MLT-3 = Mario, Luca Tazio - get it? —Preceding unsigned comment added by Wundermac (talkcontribs) 20:14, 22 June 2010 (UTC)[reply]