Talk:Network interface device

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Telecommunications (Rated C-class)
WikiProject icon This article is within the scope of WikiProject Telecommunications, a collaborative effort to improve the coverage of Telecommunications on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
C-Class article C  This article has been rated as C-Class on the project's quality scale.
 ???  This article has not yet received a rating on the project's importance scale.

RfC: Status of a NIU as a loopback device[edit]

As you can see below, this has become an edit war. Based on the old Bellcore standards, the Bellsouth TR-73569 standard located here: the ATIS Telecom Glossary 2000 telcom glossery and the stat sheet from Westell, a NIU needs to be a loopable device. I have posted this information repeatedly, only to have the refrences removed\ignored, repeatedly, and I am requesting that a third party review this information. ~As stated below, section 2.3.1 of TR-73569 states "On command the connector must loop the LEC signal from the side one transmission path back to the Local Exchange Carrier on the side two path." And on the Westell page for their NIU\Smartjacks, , they say "Without Autoterm Plus, your NIU cannot be put into loopback in any of these events. The NIU cannot do what it is made for,requiring an on site call." So, the NIU can not do what it was made for, being put into a loopback. Westell is one if not the biggest makers of NIUs out there too. Yet these refrences are ignored or removed repeatedly. Any help\feedback on ending this edit war would be helpfull, as I just do not have the time due to work (a job that involves looping up smartjacks daily, I should add) Ozzy 98 (talk) 03:12, 12 February 2008 (UTC)


Ok, since some people do not want to try this, I'm just going to rewrite the top part and list the primary functions of a NIU per Westell\Hyperedge and Bellsouth, plus the FCC's defination. It'll need a copy edit when done, but it will have citations.Ozzy 98 (talk) 15:01, 10 February 2008 (UTC)

Well, soon as I get home anyways, still at a motel. And please read: before making any more changes based on what you THINK an NIU is. I pasted links from the makers of NIUs and the users of NIUs both stating one of the primary functions is proving a loop without the use of a loopback plug. Once we get THIS part done then, we can actualy start adding the real tech info, like loopback codes used and the Layer-3 issues with some Hyperedge cards. Remember, our goal is to improve this page, so let's all work for this. Ozzy 98 (talk) 15:11, 10 February 2008 (UTC)
Please put the citations where you believe they should go. Or you can put them in the external links section. The way they are, I cannot fix them because I do not know even what you intended. You could also just copy and paste the proper citation syntax I put in the article. -- Schapel (talk) 15:40, 10 February 2008 (UTC)
There, I updated the page, I'm sure it needs some copy editing now. Please though, READ the links posted before you remove them, again. This is getting a bit silly, and will require outside assistance if it continues. Every statment is refrenced in the Bellsouth standard, it covers everything if you read it. (That's the refrence you removed last time) Ozzy 98 (talk) 19:16, 10 February 2008 (UTC)
This is getting silly.. You keep insisting that an NIU is a loopback device, yet I do not see a reference that makes such a statement. It looks like you're deliberately trying to mess up the article and daring me to clean it up. I'll see what I can do. If your edits look like it makes the article worse, as it looks like they do again, I'm going to revert them. Why not discuss such sweeping changes before making them? -- Schapel (talk) 01:19, 12 February 2008 (UTC)
Another problem is that the terms "front of the jack" and "back of the jack" are used without explanation. It sounds like the front is the interface to the CO and the back is the interface to the CSU, but this should be clarified. -- Schapel (talk) 01:44, 12 February 2008 (UTC)
"You keep insisting that an NIU is a loopback device, yet I do not see a reference that makes such a statement." <-- It's section 2.3.1 of TR-73569, as stated clearly in the refrences. "On command the connector must loop the LEC signal from the side one transmission path back to the Local Exchange Carrier on the side two path." That is a loopback. Most of the things you're putting "citation needed" are refrenced in here. Everything in here is almost paraphrased from the orgional bellcore standard or the updated bellsouth one, that I've posted here (And you've already removed once). You are repeatedly ignoring refrences posted here, even after I explain them to you, just to prove your idea that a NIU is not loopable. I'm requesting moderation for this page. Ozzy 98 (talk) 02:45, 12 February 2008 (UTC)


I tried to clean this page up a bit, but think I made it more of a mess since I do not really know wikiformatting (Nor do I care that much about learning it. I added the refrence to Bellsouth's "Functional Criteria for the DS1 Interface Connector". This is what Bellsouth requires on it's NIUs, and it pretty much is used by all LECs to. If it's not loopable, then it's hard to call it an NIU. Some of the newer citation needed were not needed; the refrence section covered them in it's defniation of an NIU, and since it was the Atis def used by the US Goverment, it's pretty much law. And to help shatter ANY doubts about an NIU being a loopable device, this is from the statsheet of a Westell NIU (One fo the biggest NIU makers):"Without Autoterm Plus, your NIU cannot be put into loopback in any of these events. The NIU cannot do what it is made for, requiring an on site call." As you can see, according to the biggest NIU maker, the point of the NIU is to allow remote testing of the line, so that an on-site call is not required. Now then, anyone want to fix my formatting mess? Ozzy 98 (talk) 04:10, 10 February 2008 (UTC)

You made edits that make the article less readable and less verifiable, and took out my requests for citations without providing citations again. I'm going to revert the edits and try to improve the article. Please don't make edits unless they truly improve the quality of the article. For one thing, describing the NIU as a loopback device is totally confusing. Yes, it supposedly includes loopback capabilities, but that is not its main function. -- Schapel (talk) 12:41, 10 February 2008 (UTC)
Yes, it is one of the main functions, as the refrences FROM THE NIU MAKERS said, that you removed. I'm starting to believe you are not trying to improve this page but vandalize it. You keep adding in citation needed for thats that are proven in the external links, under defination of a NIU. This has gone beyond trying to improve the page, and seems almost personal on your part. Please, read the refrences given. One was from the maker of NIUs, and the other one was stating the requirments of NIUs, and showed EXACTLY how they were built. This page is an industry standard, and *EVERY* NIU out there follows this standard now. I added citations, but I DO NOT KNOW HOW TO CITATION PROPERLY. Hence, right above, I listed them for others to add, as the pages you linked me to recomended. I'm not sure why you are so die-hard aginst calling a NIU a loopback device, did you just completely ignore every link I posted? Or not read the statment above? Please just use the citations I listed above, and THE ONES YOU REMOVED, to improve the page, and stop trying to make NIUs into something they're not, so we will not require moderation on this page. Ozzy 98 (talk) 14:43, 10 February 2008 (UTC)

NI vs NIU[edit]

Is NIU commonly shortened to NI by service providers?

(Response) Yes, it is. Flagmichael 02:25, 30 March 2007 (UTC)


I still have doubts about the correctness of the statement about loopback testing. When I ordered a T1 connection from AT&T in 2007, they gave me a loopback plug to plug into my CSU. When that test failed, they had me plug it into the smartjack. Why would they ask me to do that if they could initiate a loopback test from the CO? We need a reliable source that truly states that, or that statement should be removed. Of course you can initiate a loopback test if you use a loopback plug; it is not necessarily a quality inherent in the smarkjack itself. -- Schapel (talk) 21:40, 9 February 2008 (UTC)

The loopback test loops the front of the smartjack, the loopback plug tests through the smartjack, to make sure the jack is wired correctly, and is not defective itself. The loopback plug isn't a very good test too, cause it only tests half the pairs (There's 4 wires used, tip and ring on both send and recieve pairs). You use loopback plugs to test extended demarks (They use "biscut jacks" and not NIUs) a lot, but again, this isn't a good test cause you can have the pairs rolled, and really need something like a T-BERD unit. So most likely what happened, it tested good to the smartjack, and they wanted to test through the smartjack to make sure they wired it correctly in the back, or, what you're looking at was actualy a demarc extension they thought they might have wired wrong.
Course, this entire page needs reworked. I'm not 100% sure what is meant by the part about transmission protocols part. (talk) 23:00, 9 February 2008 (UTC)
Also, if you plugged the loopback into your CSU, and it FAILED, then your CSU was bad. I think you might be mixed up on something, cause a CSU with a loopback plug in it should come up\up(Looped) (talk) 00:01, 10 February 2008 (UTC)

Its much more simple than confusing words. The NIU or more frequently called "SmartJack" is also known as a DNID. Its called smart because it has software and microchips. If I plug an "Adtran" loopback adapter into a Full T-1 with a regular RJ-45 B patch cord I'm testing the smartjack I'm also making sure my helpdesk engineer can "See" my loop. If he can't see my loop but the smartjack loops up I will goto the CO "central office" and I will use my sunset tester at the correct DSX panel to tell the smartjack at the customers site to loop up. If it loops up that smartjack is working properly. Now I will test the other end going toward my DSLAM/MUX and also make sure the pairs are correctly going In's to out's and that they are in the correct position. I work for a CLEC so my process for this ends at me having proof their is a problem with the circuit(copper) the LEC gave to me. Oh yeah and the Copper pairs that energize the "Smartjack" yes there is 4 wires... but try not to confuse because the smartjack is energized by a single pair (1) at the customer end. Inside the CO I have 2.5 pairs 5 Wires each circuit is on a DSX panel. They are Tx and Rx ins and outs=4 wires and a single green monitor lead. It gets more confusing so I won't bother but I have worked for LECs before and every NIU"Smartjack" I installed if only 1 t-1 circuit was ordered at the customer site there was a tag with the circuit ID attached to a single pair .. 2 wires...That connects to a gas protection block or sometimes not depending on area. Now if your plugging a T-1 adtran loopback adapter into a direct fed PRI running on the same type of card inside the same smartjack thats a different story. T-1's and PRI's flow the same copper...Use the same smartjack... use the same type of card inside the smartjack... but are totally different in many ways. Company name drops that I've worked for Verizon (chicago), McleodUSA (chicago), Covad (Chicago). And yes LECs and CLECs use 4 or 5 wires for circuits inside the CO. But they are all a single copper pair(2 wires) in the field. XRAYRON (talk) 02:49, 17 April 2009 (UTC)

Rewrite by DragonHawk[edit]

Well. I don't even quite remember how I got here, but I saw an article that needed some help, so I was bold and started in on it. By the time I was done I had basically rewritten it. Hopefully for the better.  :) It didn't even occur to me to check this talk page until now. I see there's some history here. I thought I'd explain some of my process.

I took what I knew and did some web searching. I found a few documents from major telcos which give their take on what an NID can be. Which is pretty varied, anything from a glorified phone jack to a fairly sophisticated device, as the article now implies.

Even without seeing the discussion here, I tried to avoid making definitive statements like "an NIU must provide loopback" because there's no central authority for the entire world on what an NIU must do. Even before the AT&T breakup, the US might have had different standards from UK or Japan or whatever. And these days, anyone with a punch-down tool and a copy of Asterisk can call themselves a phone company, so it's real hard to declare "must".

It would be worthwhile, I think, to perhaps get into somewhat more technical detail for specific applications, but we should remember to put things in terms of that context. In other words, avoid the "NIU must do X" argument. Instead, talk about typical and common cases, explain what a standard specifies, state what particular companies do, etc.

I added ONT to this article because it's basically the same concept, but didn't have enough unique information to warrant an article of its own.

Comments, commendations, and condemnations are welcome. —DragonHawk (talk|hist) 04:10, 4 June 2009 (UTC)