RE: Standards for connectors and attenuators

[Eaves John <JEaves@tycomltd.com>]

As additional clarification, the 'applications' in G.691 are specifically
related to 'application codes' which describe sets of optical interface
parameters and values intended to enable multisupplier compatibility at the
optical level for various point-to-point SDH links.   In the applications
Piers Dawe listed, the shorter of the 2 target distances applies to use of
1310nm devices on G.652 fiber (conventional single-mode fiber) and the
longer distance applies to 1550nm devices on G.652, G.653 (dispersion
shifted single mode fiber) and G.655 (non-zero dispersion shifted single
mode fiber).   G.691 defines only a V application code at 120km for 10Gb/s
systems.  In all cases, the target distances are for classification, not
specification, and are generally related to the historical 40km regenerator
spacing found in many embedded long-haul terrestrial networks. 

I'd like to support the proposal from Bob Hughes for 802.3ae to reference
the G.671 specifications for components.  A revision of the Recommendation
has been prepared and is expected to be approved at the February 2001
meeting of ITU-T Study Group 15.  The numerical values quoted by Bob from
the 1996 version of G.671 have not been changed in the revised version.
Regards
John Eaves
TyCom Labs

-----Original Message-----
From: DAWE,PIERS (A-England,ex1) [mailto:piers_dawe@agilent.com]
Sent: Tuesday, November 07, 2000 7:44 AM
To: 'Bob Hughes'; 'gary.bastin@us.jdsunph.com'; 'd_hanson@tripath.com';
'peter.ohlen@optillion.com'; 'robert@gtran.com'; 'petarp@us.ibm.com';
'dkabal@picolight.com'
Cc: 'Jonathan Thatcher (WWP)'; CUNNINGHAM,DAVID (A-SanJose,ex1);
RAUSCH,DAN (A-SanJose,ex1); 'Serial PMD Ad Hoc Reflector'
Subject: RE: Standards for connectors and attenuators



Bob, great information.

At to S,T,U,V, here are extracts from G.691:

where "application" corresponds to the target distance: VSR- (very short
reach), I- (Intra-office), S- (Short-haul), L- (Long-haul), V- (Very
long-haul), and U- (Ultra long-haul).

>From the tables:
S	20 or 40 km
L	40 or 80
V	60 or 120 
U	160

Piers

> -----Original Message-----
> From: Bob Hughes [mailto:bhughes@telect.com]
> Sent: 05 November 2000 22:25
> To: piers_dawe@agilent.com; gary.bastin@us.jdsunph.com;
> d_hanson@tripath.com; peter.ohlen@optillion.com; robert@gtran.com;
> petarp@us.ibm.com; dkabal@picolight.com
> Subject: Standards for connectors and attenuators
> 
> 
> To all,
> 
> During the serial PMD specs telecon, 31 Oct 00, we briefly 
> discussed the
> return loss characteristics of the connectors used in long haul links.
> (See "Draft Minutes of serial PMD specs telecon 31 Oct 00", 
> by Piers Dawe,
> last paragraph.) As discussed, I am writing a brief summary of that
> information and providing some information on the 
> recommendations/standards
> for connectors and attenuators.
> 
> Connector Return Loss
> 
> The following anecdotal information is from the perspective 
> of the fiber
> plant "as shipped" to long-haul carriers, primarily for use 
> with long haul
> SONET installations. Persons working in other areas may have 
> a different
> perspective than this.
> 
> About 7 or 8 years ago, most of the transmission industry 
> switched over to
> "physical contact" connectors. ITU-T G.671 has the following 
> definition:
> "Physical Contact (PC): This finish is typically used in a 
> single fibre
> connector. The end face is polished to a spherical shape in 
> order to obtain
> a perfect contact between the two fibre cores and to improve the
> transmission performances of the connector. A typical 
> spherical radius is
> 10-25 mm"
> 
> PC terminations manufactured in the time frame of 2 - 8 years 
> ago will have
> a return loss of 45 dB or better. PC terminations 
> manufactured in the last
> two years will have a return loss of 50 dB or better. In 
> contrast, non-PC
> connectors manufactured more than 8 years ago could be 
> expected to have a
> return loss of  20 to 30 dB.
> 
> In order to help confirm this information, I asked a associate from
> the OTDR industry what levels of return loss he had seen in the
> installed plant. He has seen very few installations that are 
> worse than 30
> to 35 dB. He commented that any physical plant upgrade in the 
> last 6 to 7
> years would have 40+ dB of return loss.
> 
> As a side note, Angled Physical Contact (APC) connectors are 
> not typically
> used with SONET equipment - at least not with any 
> installations I know of.
> However, APC connectors do seem to be very popular with cable TV
> installations. The attractive feature of APC connectors is a very high
> return loss, e.g. 70 dB.
> 
> 
> Standards and Recommendations
> 
> It is my understanding that one of the goals of 802.3ae is 
> interoperability
> between equipment manufactured by various companies. It also 
> seems that in
> order to provide interoperability, it may be useful to specify the
> requirements of the fiber plant. This is especially true of 
> the 1550 nm PMD
> since it can be performance limited by attenuation and back 
> reflectance.
> Specifically, what I am proposing is that we reference an existing
> performance standard for attenuator and connector performance.
> 
> The standard frequently used in North America is Telcordia (formerly
> Bellcore) GR 326 "Generic Requirements for Single Mode Optical Fiber
> Connectors." However, since IEEE is an international 
> standards organization,
> it is probably more appropriate to cite an ITU standard.
> 
> There are three ITU standards (recommendations actually) that 
> I could find
> relevant to connector and attenuator performance. They are 
> listed below. I
> downloaded all three of these this week from the ITU 
> electronic bookstore.
> The cost is 20 CHF each (about USD$12).
> 
> ITU-T  Recommendation  G.671, "Transmission characteristics of passive
> optical components", 11/96
> ITU-T  Recommendation  L.31, "Optical fibre attenuators", 10/96
> ITU-T  Recommendation  L.36, "Single mode fibre optic 
> connectors", 10/98
> 
> As the title implies, G.671 contains performance 
> specifications for many
> types of passive components including connectors, attenuators, and
> isolators. The two "L" recommendations are specific to 
> outside plant and
> appear to have some overlap with G.671.
> 
> L.31 itself only contains environmental recommendations, e.g. 
> temperature,
> humidity. It references Table 6.3 in G.671 for optical performance
> specifications.
> 
> L.36 was published after G.671. In the Summary, L.36 states 
> "While taking
> into account Recommendation G.671 as far as the transmission 
> parameters are
> concerned, this Recommendation is based on the most recent 
> work carried out
> within IEC 86B Working Groups 4, 6 and 7, namely the future 
> IEC 1753-2-1 and
> the 61300-series." In specifying performance for optical 
> connectors, L.36
> appears to be more comprehensive. But, it is also more 
> complicated to use.
> For example, section 6.1.2 defines four different return loss 
> values for
> four different "Classes" of , i.e. Class S, Class T, Class U, 
> and Class V .
> The meaning or intended usage of these classes are not defined in the
> document.
> In contrast, G.671 contains a single table, section 6.9, that 
> specifies all
> of the connector performance parameters.
> 
> For all three ITU recommendations, as well as GR 326, there 
> is no connector
> type specified. The recommendations apply independent of the 
> connector type,
> e.g. SC, LC, etc. Thus, we do not have to get into the 
> connector wars in
> order to use these recommendations.
> 
> The G.671 performance recommendations for attenuators, and 
> connectors are
> shown below.
> 
> Attenuators
> 
> Insertion Loss Tolerance +/- 15%
> Optical Reflectance (dB) -40 Max
> Polarization Dependent Loss (delta dB) 0.3 Max
> 
> 
> Connectors
> 
> Insertion Loss (dB) 0.5 Max for single fibre (Note 1)
> Optical Reflectance (dB) -35 Max (Notes 1 and 2)
> Polarization Dependent Loss (delta dB) 0.1 Max
> NOTE 1 - When used over an extended operating temperature range, these
> values may be exceeded and are under study.
> NOTE 2 - For networks other than those covered by 
> Recommendation G.982,
> including other access networks, a value of -27 dB is allowed 
> however, care
> should be taken to insure system functionality in systems 
> implemented with
> several optical components with reflectance values at, or 
> near, this limit.
> In consideration of future network evolutions, a value of -40 
> dB is under
> study.
> 
> 
> My specific proposal
> 
> I would like to propose that within 802.3ae, we recommend 
> compliance with
> "ITU-T Recommendation  G.671, "Transmission characteristics of passive
> optical
> components", for attenuators and connectors used with the 
> 1310 and 1550 nm
> serial PMD. The purpose of including such a recommendation is 
> to give the
> end-users guidance in the performance specification of the 
> fiber physical
> plant to provide satisfactory end-to-end performance of the 
> 802.3ae system.
> 
> Thanks,
> Bob Hughes
> Telect
> bhughes@telect.com
> 
>