RE: Patch cord for 1300 WWDM
Dave,
Yes, I was referring to SX not LX in this discussion.
Providing a "conditioned" launch is not required of the 850 CWDM solution in
order to achieve the 100 m objective on the installed base of 160 MHz-km
62.5 um fiber. Nor is it required to meet the 300 m objective on the
installed base of 500 MHz-km 50 um fiber.
To get extended distances on enhanced 62.5 micron fiber, then a conditioned
launch is needed, per the TIA FO2.2 launch recommendation. In order to
achieve the full 2000 MHz-km bandwidth of the new 50 um fiber, a conditioned
launch is also needed. I was simply saying that by meeting that conditioned
launch, one would not cause a degradation of bandwidth on the installed base
of 62.5um fiber. Whether it is practical to design such a launch with low
cost optics is another matter. There are design choices that allow more than
one way to meet this criteria. In any case, conditioned launch is not
required in order for the solution to meet the 100 and 300 m objectives.
With regards to Lew Aronson's work, it does support my assertion that the
850 nm OFL bandwidth is the worst case bandwidth for the installed base of
62.5 um fibers. Lew investigated several ROFL launches and compared them to
OFL launches. There were no cases in his study that contradicted my
statement within the measurement accuracy of the test. There were cases,
however, where the OFL bandwidth was below what we now call minimum
standards compliant bandwidth (160 MHz-km at 850 nm). However, these
installations may pre-date the definition of such standards. (The national
laboratories, where the cable plant was tested, were very early adopters of
fiber). And in fact at least one of the cables tested was known to not be
compliant prior to the test. Since these early installations, fiber
manufacturers have improved their processes and selection criteria to
provide adequate guard bands for variations seen between factory and field
bandwidth performance.
Regards,
Paul Kolesar
----------
From: Dave Dolfi 3764 [SMTP:dolfi@aldolfi.labs.agilent.com]
Sent: Friday, August 04, 2000 3:21 PM
To: stds-802-3-hssg@ieee.org; Kolesar, Paul F (Paul)
Cc: dave_dolfi@agilent.com
Subject: RE: Patch cord for 1300 WWDM
Dear Paul,
I had a feeling I'd hear from you on this, and you didn't disappoint
me! Thank you for your detailed response. You seem to be saying
two things. First, you describe what is basically a doughnut shaped
launch which will both satisfy the launch requirement for the new
fiber
to achieve greater than its OFL bandwidth, and at the same time,
guarantee that the OFL bandwidth is achieved on the installed base
of 62.5 um MMF. However, you also seem to be saying that it doesn't
matter anyway, since ANY launch will achieve the OFL bandwidth on
the
62.5 um installed base. (I assume that you are only referring to SX,
since if it were true for LX as well we wouldn't need patch cords).
With respect to your first statement, I would reply that the launch
you describe can certainly be classified as a "conditioned" launch.
I don't think it could easily be achieved with multiple sources,
at least not with the kind of simple multiplexers which have been
proposed by companies such as Blaze and Agilent. I also question
whether the tolerances required by such a launch would allow the
loose, multimode type tolerances which the 850 nm CWDM advocates
have
claimed for this PMD, and which form the basis of their claim of
lower cost relative to 1300 nm WWDM. If this launch is really
necessary for 850 nm CWDM to work, then these questions need to
be addressed.
With respect to your second statement, I would refer you to the work
that was done by Agilent (HP at the time) during the DMD studies of
802.3z, particularly the presentation to 802.3z by Lewis Aronson at
the March, 1998 meeting of 802.3z (available at the following URL:
http://grouper.ieee.org/groups/802/3/z/public/presentations/).
There are measurements presented there on both 62.5 um and 50 um
MMF at short wavelength. I'm not in a position to challenge your
assertion regarding the absence of launch restrictions at short
wavelength (you're the expert here!) but I would appreciate it if
you look over this paper and verify that the results presented are
consistent with your statement.
Dave Dolfi
Agilent Technologies
> From owner-stds-802-3-hssg@ieee.org Fri Aug 4 09:14:56 PDT 2000
> Return-Path: <owner-stds-802-3-hssg@ieee.org>
> Received: from unicorn.labs.agilent.com (unicorn.labs.agilent.com
[130.29.252.5])
by aldolfi.labs.agilent.com (8.9.3 (PHNE_18979)/8.9.3
AgilentLabs
Workstation) with ESMTP id JAA19492
for <dolfi@aldolfi.labs.agilent.com>; Fri, 4 Aug 2000
09:14:56 -0700
(PDT)
> Received: from alexed.labs.agilent.com (alexed.labs.agilent.com
[130.29.252.59])
by unicorn.labs.agilent.com (8.10.2/8.10.2/Agilent Labs Mail
Hub v 01.00
2000/06/20) with SMTP id e74GEtj28086
for <dolfi@aldolfi.labs.agilent.com>; Fri, 4 Aug 2000
09:14:55 -0700
(PDT)
> Received: from 130.29.252.5 by alexed.labs.agilent.com (InterScan
E-Mail
VirusWall NT); Fri, 04 Aug 2000 09:14:55 -0700 (Pacific Daylight
Time)
> Received: from hplms26.hpl.hp.com (hplms26.hpl.hp.com
[15.255.168.31])
by unicorn.labs.agilent.com (8.10.2/8.10.2/Agilent Labs Mail
Hub v 01.00
2000/06/20) with ESMTP id e74GErv28067;
Fri, 4 Aug 2000 09:14:53 -0700 (PDT)
> Received: from hplms2.hpl.hp.com (hplms2.hpl.hp.com [15.0.152.33])
by hplms26.hpl.hp.com (8.9.3 (PHNE_18979)/HPL-PA Relay) with
ESMTP id
JAA28307;
Fri, 4 Aug 2000 09:14:52 -0700 (PDT)
> Received: from hplms26.hpl.hp.com (hplms26.hpl.hp.com
[15.255.168.31])
by hplms2.hpl.hp.com (8.10.2/8.10.2 HPL-PA Hub) with ESMTP
id
e74GEpA06527;
Fri, 4 Aug 2000 09:14:51 -0700 (PDT)
> Received: from ruebert.ieee.org (ruebert.ieee.org [199.172.136.3])
by hplms26.hpl.hp.com (8.9.3 (PHNE_18979)/HPL-PA Relay) with
ESMTP id
JAA28302;
Fri, 4 Aug 2000 09:14:49 -0700 (PDT)
> Received: by ruebert.ieee.org (8.9.3/8.9.3) id LAA28587; Fri, 4
Aug 2000
11:42:34 -0400 (EDT)
> Message-ID:
<4490F7068AC0D111A7120008C72878EC04E6CBA6@nj7460exch003u.ho.lucent.com>
> From: "Kolesar, Paul F (Paul)" <pkolesar@lucent.com>
> To: stds-802-3-hssg@ieee.org
> Subject: RE: Patch cord for 1300 WWDM
> Date: Fri, 4 Aug 2000 11:42:30 -0400
> MIME-Version: 1.0
> X-Mailer: Internet Mail Service (5.5.2650.21)
> Content-Type: text/plain;
charset="iso-8859-1"
> Sender: owner-stds-802-3-hssg@ieee.org
> Precedence: bulk
> X-Resent-To: Multiple Recipients
<stds-802-3-hssg@majordomo.ieee.org>
> X-Listname: stds-802-3-hssg
> X-Info: [Un]Subscribe requests to majordomo@majordomo.ieee.org
> X-Moderator-Address: stds-802-3-hssg-approval@majordomo.ieee.org
> Content-Length: 9199
> Status: RO
>
>
> Dave Dolfi,
>
> I would like to address your concerns over launch conditions.
>
> The launch requirement for the new 50 um fiber is presently
specified as >=
> 85% encircled flux within a 16 um radius of the center of the
fiber. This is
> not much different than the launch requirement determined to be
optimal for
> enhanced 62.5 um fiber which is <= 25% encircled flux within 4.5
um radius
> and >= 75% within 15 um radius of the center of the fiber. The
main
> difference is that the 62.5 um spec limits the amount of power
allowed in
> the very center of the fiber (within 4.5 um radius) while the 50
um spec has
> no such limitation. But, the similarity of the outer radii
specification
> (85% within 16 um vs 75% within 15 um) permits solutions that meet
both
> requirements simultaneously. In short, the overall power
concentration in
> the center of the 50 um fiber is not required to be much different
than that
> for enhanced 62.5 fiber.
>
> Further, the TIA FO2.2 data indicates launches that meet the above
> requirements for enhanced 62.5 um fibers do not cause degradation
of the
> bandwidth below the OFL specification on installed-base 62.5 um
fibers. In
> fact, such launches usually cause bandwidth enhancement. By
combining these
> launches with 62.5 um fibers tested and determined to produce
higher
> restricted launch bandwidth, we can guarantee enhanced
performance. Perhaps
> even more to the point, the data collected by both the TIA and
IEEE on this
> subject has not shown any launch condition that caused the
bandwidth to
> collapse below the 160 MHz-km OFL spec for 62.5 um fiber. This
includes
> launches produced by single-transverse mode "CD" lasers as well as
> multi-transverse mode VCSELs with various spot sizes and numerical
> apertures. It also includes Radial Overfilled Launches that were
developed
> as a possible test launch condition by the IEEE MBI study group,
which are
> said to extract the "worst case" modal bandwidth for any laser
launch. In
> summary, there is no evidence of any launch condition that causes
less than
> 160 MHz-km bandwidth from the installed base of 62.5 um fiber. The
100 m
> capability of the 850 CWDM PMD is based on 160 MHz-km bandwidths.
>
>
> Regards,
> Paul Kolesar
>
> ----------
> From: Dave Dolfi 3764 [SMTP:dolfi@aldolfi.labs.agilent.com]
> Sent: Thursday, August 03, 2000 6:17 PM
> To: stds-802-3-hssg@ieee.org; billw@blazenp.com
> Cc: dave_dolfi@agilent.com
> Subject: RE: Patch cord for 1300 WWDM
>
>
> Dear Bill,
>
> I'm happy that you agree with my summary of the patch cord
> situation.
> Unfortulately, I'm not sure that I agree with everything you
seem
> to be saying about 850 nm CWDM. Specifically, while I agree
that
> you could specify a VCSEL and a mux design which would
achieve 100
> meters on the installed 62.5 um MMF OR > 550 meters on the
new
> enhanced
> bandwidth 50 um MMF, I'm not sure that a single design could
achieve
>
> both at the same time.
>
> The 62.5 um conventional fiber requires a large spot at its
input to
> mitigate potential DMD problems if too much of the
excitation is in
> the central portion of the fiber, while the new fiber
requires a
> small spot centered at its input in order to satisfy the
encircled
> flux
> requirement necessary to achieve the higher bandwidth. Can
you
> really
> achieve both of these at the same time?
>
> I think you need to prove that this is true before you can
claim to
> simultaneously achieve the bandwidths you are claiming over
both
> fiber
> types.
>
>
> Dave Dolfi
> Agilent Technologies
>
>
> > From owner-stds-802-3-hssg@ieee.org Thu Aug 3 11:18:07
PDT 2000
> > Return-Path: <owner-stds-802-3-hssg@ieee.org>
> > Received: from unicorn.labs.agilent.com
(unicorn.labs.agilent.com
> [130.29.252.5])
> by aldolfi.labs.agilent.com (8.9.3
(PHNE_18979)/8.9.3
> AgilentLabs
> Workstation) with ESMTP id LAA18859
> for <dolfi@aldolfi.labs.agilent.com>; Thu, 3 Aug
2000
> 11:18:07 -0700
> (PDT)
> > Received: from alex1.labs.agilent.com
(alex1.labs.agilent.com
> [130.29.252.55])
> by unicorn.labs.agilent.com (8.10.2/8.10.2/Agilent
Labs Mail
> Hub v 01.00
> 2000/06/20) with SMTP id e73II6c16414
> for <dolfi@aldolfi.labs.agilent.com>; Thu, 3 Aug
2000
> 11:18:06 -0700
> (PDT)
> > Received: from 130.29.252.5 by alex1.labs.agilent.com
(InterScan
> E-Mail
> VirusWall NT); Thu, 03 Aug 2000 11:15:34 -0700 (Pacific
Daylight
> Time)
> > Received: from hplms26.hpl.hp.com (hplms26.hpl.hp.com
> [15.255.168.31])
> by unicorn.labs.agilent.com (8.10.2/8.10.2/Agilent
Labs Mail
> Hub v 01.00
> 2000/06/20) with ESMTP id e73II5x16406;
> Thu, 3 Aug 2000 11:18:05 -0700 (PDT)
> > Received: from hplms2.hpl.hp.com (hplms2.hpl.hp.com
[15.0.152.33])
> by hplms26.hpl.hp.com (8.9.3 (PHNE_18979)/HPL-PA
Relay) with
> ESMTP id
> LAA15401;
> Thu, 3 Aug 2000 11:18:03 -0700 (PDT)
> > Received: from hplms26.hpl.hp.com (hplms26.hpl.hp.com
> [15.255.168.31])
> by hplms2.hpl.hp.com (8.10.2/8.10.2 HPL-PA Hub) with
ESMTP
> id
> e73II2e24872;
> Thu, 3 Aug 2000 11:18:02 -0700 (PDT)
> > Received: from ruebert.ieee.org (ruebert.ieee.org
[199.172.136.3])
> by hplms26.hpl.hp.com (8.9.3 (PHNE_18979)/HPL-PA
Relay) with
> ESMTP id
> LAA15314;
> Thu, 3 Aug 2000 11:18:00 -0700 (PDT)
> > Received: by ruebert.ieee.org (8.9.3/8.9.3) id NAA03725;
Thu, 3
> Aug 2000
> 13:43:06 -0400 (EDT)
> > Reply-To: <billw@blazenp.com>
> > From: "Bill Wiedemann" <billw@blazenp.com>
> > To: <stds-802-3-hssg@ieee.org>
> > Subject: RE: Patch cord for 1300 WWDM
> > Date: Thu, 3 Aug 2000 10:38:21 -0700
> > Message-ID: <A0DC08D4103BD411A9DC009027B0B6350E8130@MAIL>
> > MIME-Version: 1.0
> > Content-Type: text/plain;
> charset="iso-8859-1"
> > Content-Transfer-Encoding: 7bit
> > X-Priority: 3 (Normal)
> > X-MSMail-Priority: Normal
> > X-Mailer: Microsoft Outlook CWS, Build 9.0.2416
(9.0.2910.0)
> > X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300
> > In-Reply-To: <A0DC08D4103BD411A9DC009027B0B63514345C@MAIL>
> > Importance: Normal
> > Sender: owner-stds-802-3-hssg@ieee.org
> > Precedence: bulk
> > X-Resent-To: Multiple Recipients
> <stds-802-3-hssg@majordomo.ieee.org>
> > X-Listname: stds-802-3-hssg
> > X-Info: [Un]Subscribe requests to
majordomo@majordomo.ieee.org
> > X-Moderator-Address:
stds-802-3-hssg-approval@majordomo.ieee.org
> > Content-Length: 2823
> > Status: RO
> >
> >
> > Thank You.
> > Everything you say is correct. 850CWDM has been designed
to
> directly meet
> > the 100 meter objective over installed (DMD challenged)
62.5u MMF
> and the
> > 300m objective over MMF. In addition we can meet 300
meters over
> existing
> > 50 micron fiber and greater than 550 meters with the new
high
> bandwidth MMF.
> >
> > Finally experimental evidence has shown greater than 300
meter
> performance
> > over installed 62.5 micron fiber that is not DMD
challenged.
> >
> > Bill Wiedemann
> > Blaze
> > 925-560-1610 x169
> >
> > At 02:11 PM 8/2/00 -0700, David W Dolfi wrote:
> >
> >
> > >Everyone,
> > >
> > >
> > >There seems to have been some confusion at the La Jolla
> > >meeting over the necessity for an offset patch cord for
> > >1300 nm WWDM. Because of this, and additional comments
> > >made on the reflector since the meeting, I am writing
> > >this email to clarify the situation.
> > >
> > >Fact 1. An offset patch cord is NOT required for 1300 nm
> > >WWDM in order to meet the current MMF objectives of
802.3ae.
> > >That is to say, it is NOT required in order to achieve a
100
> > >meter link length on the installed base (this includes
both
> > >62.5 and 50 um standard MMF, which both have a 500 MHz-km
OFL
> > >bandwidth length product at 1300 nm), NOR is it requred
to
> > >achieve a 300 meter link length on the new enhanced BW
> > >MMF, which also has a 500 MHz-km OFL bandwidth at 1300
nm.
> > >
> > >Needless to say (but I will for the sake of completeness)
> > >1300 nm WWDM also supports single mode fiber up to 10 km,
> > >again without a patch cord.
> > >
> > >
> > >Fact 2. The ONLY time you need to use a patch cord with
> > >1300 nm WWDM is if:
> > >
> > >1. You want to extend the link length of the MMF
installed base
> > >to 300 meters
> > >
> > >AND IN ADDITION TO THIS
> > >
> > >2. The fiber in question is "DMD challenged".
> > >
> > >
> > >Please note that if you are in this particular situation,
none
> > >of the 850 nm based PMDs will satisfy your need, patch
cord or
> > >not (but see Note below). Your only alternative in this
> situation,
> > >with an 850 nm PMD, is to install new fiber, either the
enhanced
> BW
> > >multimode fiber or single mode fiber. Therefore, the
notion that
> > >the patch cord is some sort of "penalty" you pay for
using 1300
> nm
> > >WWDM is really the wrong way to think about it. Rather
than a
> > >shortcoming, it is actually a benefit, since it gives you
the
> (rel-
> > >atively speaking) low cost option of using a patch cord
in a sit-
> > >uation where your only other alternative is to pull new
fiber.
> > >
> > >Note: The 850 nm 4 channel CWDM PMD will allow you a 300
meter
> link
> > >length, without a patch cord, on the installed base of 50
um
> fiber
> > >ONLY. However, this is a small benefit, since the great
majority
> > >of the MMF installed base is 62.5 um fiber, on which 850
nm CWDM
> > >will only support a 100 meter link length (due to the
fact that
> > >62.5 um fiber has an OFL bandwidth length product of only
160
> > >MHz-km at 850 nm).
> > >
> > >
> > >David Dolfi
> > >Agilent Technologies
> >
> >
>