Jump to content

Talk:Manchester code

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

Purpose

[edit]

So, what does this code do?

Provides a simple way of encoding arbitary binary sequences without ever having long periods where there is all one voltage, thus losing clock synchronization. Also ensures that the DC component of the encoded signal is zero. -- The Anome 12:13, 18 Sep 2004 (UTC)

Baseband

[edit]

Is it worth mentioning that the codes are only useful on baseband mediums, or would that be implicit? It doesn't seem to me that it's entirely clear in the articles on telecoms how modulation and encoding relate to each other. Probably such a change doesn't belong here, but it's where I thought of it...--Ktims 22:24, 31 March 2006 (UTC)[reply]

This is a basic attribute of a line code. There is some discussion of this at Line code#Transmission and storage but improvement there is welcome. ~Kvng (talk) 13:43, 26 March 2020 (UTC)[reply]
I suppose in non-baseband cases it could be used as the line code before modulating some other signal. I don't know that anyone does that. Otherwise, there is a question about what is a 1 or 0. On magnetic media, only flux changes generate a voltage. Gah4 (talk) 08:12, 3 July 2020 (UTC)[reply]

Encoding conventions

[edit]

The difference between the "Thomas" and IEEE conventions is so trivial (inversion of the signal) that its description causes more confusion than illumination. If any conventions are to be compared, perhaps an explanation of "differential" encoding would be more use (if someone can do this concisely). Mike Shepherd 11:36, 12 July 2007 (UTC)[reply]

Manchester code#Conventions for representation of data now seems clear on this. ~Kvng (talk) 13:47, 26 March 2020 (UTC)[reply]

Transformer coupling

[edit]

I remember learning when I was on an aircraft databus course that this encoding is still (it was a couple of years ago when I was on the course) used whenever devices are transformer coupled on to a databus (or are current rather than voltage sensitive); as it ensures that the signal is constantly alternating. I'll try and do some research/remember it.

Would that be worth adding to the article? 212.17.152.102 (talk) 18:24, 25 August 2009 (UTC) Hibbo[reply]

This is now covered in the second sentence of the first paragraph. It is possible readers may be thrown by terminology: no DC component (constantly alternating) and galvanically isolated (transformer coupled). ~Kvng (talk) 13:51, 26 March 2020 (UTC)[reply]

ethernet

[edit]

For one, coaxial ethernet uses a negative voltage on the cable, convenient for the ECL drivers. So if you look at it as no-voltage and yes-voltage, the sign is the other way around. It needs to be non-symmetric to make collision detection work. UTP ethernet can easily be wired with the polarity wrong. Many 10baseT transceivers detect and correct polarity errors, and then have an LED to tell you that. They can then advertize this LED as a feature. I believe they test the polarity at the end of the preamble. Gah4 (talk) 08:20, 3 July 2020 (UTC)[reply]

cells and cell boundaries

[edit]

Reading ECMA-36, which I believe is (close enough to) the official definition for PE encoded 1600BPI magnetic tape, the description is different. In their description, the data transitions are on the cell boundaries, and the transitions needed to get in the right state for the next one in the middle. That is, a half cell shifted from the description here. That doesn't change what the data looks like, but is important to know when reading descriptions. Gah4 (talk) 22:40, 15 September 2021 (UTC)[reply]