History of Telecommunications T-Carrier | Hacker News Reader

History of Telecommunications T-Carrier

computer.rip

39 pointsaberoham15 comments

As opposed to E1 links in Europe. Same same until you want to patch a US T1 to a European E1. Some sweet hardware money was made dealing with that. ( different framing, 2mbit/s instead of 1.44, 56kbps splits vs 64kbps splits. Clocks are a nightmare.

Story is that Bell tested the T1 in comms pits around Murray Hill NJ and when it worked for a long city block between two manholes they knew they had a product they could sell.

Watched a telco guy fix a broke E1 by sending one of the pairs a bit further round the krone frame to separate the signals better. I've even seen them add more wire to try and dampen down some local noise or reflectance or harmonics or something.

The old E1/T1 lines were sometimes pressurised and came with fancy brass taps to let the water out with its own teeny weeny bucket hanging off the tap. I kid you not.

> Some sweet hardware money was made dealing with that

I used to work in a company that made good money on those

Anything that isn't fiber is still frequently nitrogen pressurized.

Side note: as a teen I always wondered why those tanks never got stolen. When I got a car I stopped one day and checked- they were stamped with the teleco name. Writing this I wonder: did they stamp the regulator too?

Thanks for the bit of history. I read that by the early 1970s, a number of smaller recording studios in Los Angeles were linked by T1 lines to Capitol Records in order to make use of their famous reverberation chambers. I wondered who manufactured such high-performance ADCs and DACs at that time.

Ha! As I remember, each end had an independent clock and a little bit of buffer to smooth out any variance. And it mostly worked!

The word "plesiochronous" is burned into my brain as a result of the few times that it didn't, though...

My career began in the late 90s telecoms boom, which was a gloriously chaotic period in retrospect. It seemed so very obvious at the time that "ethernet for everything" was what we should be working towards, but the legacy telcos didn't arrive at the same conclusion until well after the crash - so there were lots of opportunities for smaller players to undercut them.

"Never bet against Ethernet." Bob Metcalf, probably

The company I worked for in the 2000s had both kinds of T1: They had a single CAS (channel-associated, or "robbed bit", signaling) T1 from one carrier, an two PRI (ISDN) T1s from another carrier. The two PRI T1s only had one channel for signaling, and 47 channels for voice calls.

Our block of phone numbers (we had a block of 500 or 1,000 numbers) was tied to the PRI T1s; for the CAS T1, we were allowed to use our number block as the "calling party" number for outgoing calls. If the PRI T1s went down folks wouldn't be able to call us, but we could call them and the called parties wouldn't notice any difference.

> 24, 4, 7, it has the upsetting feeling of a gallon being four quarts each of which is two pints.

A gallon is two pottles; a pottle is two quarts; a quart is two pints; a pint is two cups; a cup is two gills. And a gill is two jacks!

24 and 4 don’t strike me as unusual numbers (two dozen and two doubled), but seven does.

> a different media

‘A … medium.’ One medium, two or more media.

> Early versions of DS1 only actually carried 7 bits for each sample, which was sufficient for a telephone call when companding was used to recover some of the dynamic range. The eighth bit was used for framing.

A book I have says the eighth bit was initially used to indicate on-hook/off-hook status, not framing.

This is correct. There's more. An incoming call was indicated by a solid "seizure" or activation of this bit. The receiving end would "wink" back to indicate that it would accept the call. This was ~0.5 sec blip "on" in the reverse channel. The sending channel could then transmit whatever information was arranged, e.g. internal extension number, called number, calling party number. After any data was transmitted via audio as DTMF or MF, if the receiving end decided to answer it would "seize" in return. In theory, the call ended when either end "dropped" the bit, however it was often the case that when calling into lesser modern destinations for example, the call only ended when the calling party ended it, leading to some fun empirical solutions :)

As I remember it T-1/DS-1 was the standard demarc with the ISP in North America while I the UK it was PRI ISDN.

The company I work at originally did TDM test eqpt, there was a mad scramble around 2008/2009 to switch over to Ethernet, which is what the US telcos use now for customer call traffic.

Ethernet or Carrier/Metro Ethernet?

> A DS1 is a combination of 24 DS0s, for 1.544Mbps.

> A DS2 is a combination of 4 DS0s, for 96 channels or 6.312Mbps.

What sorcery is this? How can 4 DS0 lines provide more bandwidth than 24 DS0 lines?

It's a typo, 4 DS1s. Quite clear from context.

Reply on news.ycombinator.com