Welcome to our community

Host Computer Services    hostcomputers.infopop.cc    Forums  Hop To Forum Categories  Miscellaneous Archive    Data encoding - NRZ & NRZI

Read-Only Read-Only Topic
Go
New
Find
Notify
Tools
-star Rating Rate It!  Login/Join 
Anorak
Posted
Hi,

I was chatting with a colleague the other day who is in the process of Cisco certification.

The topic of data encoding used on some Cisco kit came up and after some investigation - I thought it may be interesting to share this on the forum. The data encoding in question is NRZ (Non-Return-to-Zero) and NRZI (Non-Return-to-Zero-Inverted).

Details of NRZ here

Details of NRZI here


Tea break over - back on yer heads.....
 
Posts: 57Report This Post
How do I get out of here?
Posted Hide Post
Forgive my scepticism........

The "Wikipedia" links remind me of the Monty Python sketch about the Serbo-Croat dictionary.
Can you believe a word it says ???

I can't find any references to RS232 using NRZ signaling (as it claims). I always thought it was a simple "no voltage" = 0 .........."some voltage" = 1

NRZ and NRZI are "unbalanced" and have the potential to send long strings of zero's or one's down your piece of wire that look like a "dead" transmitter. So while I dont have the
time or inclination to research all thier stuff - I doubt if the higher speed transmission methods adhere much to the original implementations of NRZ/NRZI nowadays. USB does use NRZI (of sorts) - they have to use "bit stuffing", (i.e. shoving a "zero" up it's data stream every six consecutive "ones"), to ensure signal balancing - but it's highly modified, and uses differential signals.

Wikipedia - worth all you pay for it.

FT
 
Posts: 142Report This Post
I should get out more.
Posted Hide Post
This site gives some useful information Here


"Free will is an illusion. People always choose the perceived path of greatest pleasure"
 
Posts: 407Report This Post
How do I get out of here?
Posted Hide Post
Wait a bit - that's cheating.......

It says (as I said with RS232), that NRZ is "+V=1" and "0V=0" - well thats not encoding at all !!! - it's like saying "I'm going to set a standard for a large grey animal with a big ears and a trunk - it's called a penguin!"

Excuse my ignorance of NRZ, I guess I need to apologise to Wikipedia. The last time I looked at NRZ encoding was about 23 years ago! (that's my excuse!!).

I'll stick to my guns about NRZI though, if the encoding method has significantlly changed - (unbalenced to balanced, single-ended to differential, bit-stuffed, etc) they ought to call it something different!!.........how about diff/balanced/non-return from zero inverted/bit-stuffed....or "DBNRZIBS".....maybe "Penguin" would be better

FT
 
Posts: 142Report This Post
Member
Posted Hide Post
I rather like Dogbert's reference link that says about RZ encoding, "So a '0' is twice as long as '1', because a '1' is sent as a +v and then a 0v." That reminds me of a certain person's data compression algorithm.

I was told, many years ago, that the I in NRZI has nothing to do with the word, "inverted", and that it is in fact a misrepresentation of the 1 character. i.e. NRZI is really NRZ1 and stands for non-return to zero on ones - which describes exactly what it does. It's a lot easier to say NARZY than NARZ-ONES. NRZ describes how the signal toggles from one polarity to another without coming to rest at a zero or null level, by comparison with RZ which does exactly that. Maybe "toggle on ones - TOO " would have been a better acronym than NRZI, as it provides a better description of what really happens. I suspect it's too late in the day to change it now.

The major benefit of NRZI is, of course, that it's the most efficient data encoding technique in terms of signal transitions per data bit. On the down side, a stream of successive zeros produces no signal transmissions, so pure NRZI is not self-clocking. One way of making it self-clocking is for every group of four successive data bits to be re-coded to five bits in such a way that any two adjacent 5-bit groups will never have more than two successive 0s. (But you knew that anyway.)

We should of course realise that NRZI isn't just used for electrical signal transmission. It is used also in digital magnetic recording, in optical recording and in fibre-optic communications. Certainly a voltage can switch from positive to negative and a magnetic domain can flip N-S, but so far as I know, current optical technologies haven't yet exploited polarised light. They just switch the light on and off. So are we correct in referring to these as using NRZI - whatever that means?

I like to start a discussion at the bar then leave when it gets heated. I'll stay out of this. Anyway, aren't ones thinner than zeros?
 
Posts: 5Report This Post
Woolly Mammoth
Anorak
Posted Hide Post
Only when they're stood upright. Horizontally they are larger, but this would make them more aerodynamic, and so less likely to fly off because of the centrifical force.
This, of course, all goes out of the window if you write your zeros with a slash through them because this can then dig in for better purchase. Wink


Woolly M.
 
Posts: 71 | Location: HOST ComputersReport This Post
  Powered by Eve Community  

Read-Only Read-Only Topic

Host Computer Services    hostcomputers.infopop.cc    Forums  Hop To Forum Categories  Miscellaneous Archive    Data encoding - NRZ & NRZI