Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [802.3_10SPE] Single Twisted Pair Ethernet over AWG 22-24 type cable ?

Yong – thanks for bringing this up. (and good to hear from you – it’s been a little while)


There are many overlaps in the general functional characteristics of the xDSL systems. We had a nice overview presentation on SHDSL-based systems from Mario Traeber in San Antonio, which fits within the space of 10SPE.  The and 10PASS-2 systems have somewhat different operating characteristics than we have been considering, and, if you are willing, I would welcome a presentation on them.  We should consider how the technology base there relates to our needs, and therefore not re-invent wheels.


As far as Patrick’s original comment is concerned, we do not specify the gauge of the cable in the IEEE specification, but rather, specify the transmission characteristics of the link segments. (I’ve often said, that if we could make the link segments out of string and meet the transmission requirements, that would be acceptable!)


If there are additional cabling configurations that are relevant to the problem, it would help us if you could bring a presentation on the transmission characteristics of the cabling used in the application spaces you wish to address.  Particularly, think of it not as a “per meter” specification, but the specification of the insertion loss, return loss, crosstalk, and EMC-related characteristics for the link segments you are mentioning.   I expect that the gauge of the wire used will likely be a function of the reach and powering levels in the application (as it is in many places).


Patrick, I would also urge you to become/remain engaged with cabling standards organizations like TIA TR42 and ISO/IEC JTC1 SC25 WG3 which have projects on single-pair cabling.  These groups MAY define things like wire gauge, connectors, etc.



From: Yong Kim [mailto:000006d33765285e-dmarc-request@xxxxxxxx]
Sent: Tuesday, December 20, 2016 8:37 AM
To: STDS-802-3-10SPE@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_10SPE] Single Twisted Pair Ethernet over AWG 22-24 type cable ?


Don’t we have xDSL and that operates over the POTS cables that may have (most likely than not) ‘ugly’ cable stubs (mostly unterminated and sometimes very long)?


What’s the rationale to go slower and invest further R&D?   Just curious.




From: patrick Mourot [mailto:patrick.mourot@xxxxxxxxxxxxxxxxx]
Sent: Tuesday, December 20, 2016 8:01 AM
To: STDS-802-3-10SPE@xxxxxxxxxxxxxxxxx
Subject: [802.3_10SPE] Single Twisted Pair Ethernet over AWG 22-24 type cable ?


Dear 802.3cg,

In industrial building Profibus type cable is deployed but some other types of cable are present in those structures.

Today, in most of them we also find single twisted pair cables (screened or unscreened) cabling for telecommunication purposes particularly for TDM (Time Division Multiplexing) telephonic systems.The typical application is to transport Voice and Signalling messages between devices.

This cabling system is provided at a low cost and lot of cabling installers are well aware of its specifications.

Having Ethernet over the existing wiring system in the buildings will provide a solution to migrate traditional TDM equipments towards the IP world. While keeping the current cabling the solution becomes cost effective : it requires only changing devices (i.e. VoIP phone sets instead of TDM ones). The solution needs only minor investments (put in place some Ethernet switches) but no civil engineering is required to deploy specific Ethernet cables (e.g. Category 5 or Category 6). De facto, this optimizes the overall cost of the IP infrastructure.

The single twisted pair cable is similar to AWG 22, corresponding roughly to Æ 0,6 mm copper or AWG 24 corresponding roughly to Æ 0,5 mm copper.

The cable characteristics are :

-80-135 Ohms at 700 kHz with a 25dB max attenuation between 340 and 680 kHz.

-types : SYT0.5 & SYT0.6 or 278 or ITT 26626 BBEB DEBE or TsR-R depending on the country with a range between 800m and 1200m.

The decreasing loss with a constant of 15dB per decade of frequency and a 44dB loss at 700 kHz (90% level of confidence).

An artificial line equivalent to 900m of LY0.5 cable (Æ 0,5 mm copper) could be as follow:

We are wondering if the802.3cg group could have some interests for consideration in standardization of Ethernet (10Mb/s ?) on this type of cable.

Feel free to comment and thank you for your feedback.


P. Mourot