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

RE: [EFM] Cable plant question




Kent,

Thanks for the clarifications.  Bottom line, 20dB ORL is adequate, and
this is what we are working on.

Meir

-----Original Message-----
From: kmccammon@tri.sbc.com [mailto:kmccammon@tri.sbc.com]
Sent: Monday, February 25, 2002 1:13 PM
To: Vipul_Bhatt@ieee.org; stds-802-3-efm@ieee.org
Subject: RE: [EFM] Cable plant question



Hi Vipul,
Thank you for posing the question to operators.

SBC uses SC/UPC connectors in all the 13 states for new systems and new
fiber crossconnect arrangements.  We do have legacy cables with ST that
are
in use. We even have a few Biconic used for older asynchronous systems
in
use on the older cables. However, in the effort to reduce costs, SC/UPC
is
the connector in the central offices for the future at this time. 

SBC would be given a large operational burden to replace UPC connectors
with
APC polish connectors. Requiring APC moves cost to the operator in some
cases and especially in this case. Systems specified today by our
company
are required to have SC/UPC and are required to operate over such a
fiber
plant created by SC/UPC. Our fiber panels and CO jumpers are much easier
to
maintain if only one type is universal, plug we get some real good
volume
pricing with this strategy. 

New systems that are more sensitive to reflections like OC-192 and
analog
video may need either proper field methods to sustain higher return loss
with UPC or require APC type connectors. SBC will try to maintain a
single
type of connector/polish going forward until proven otherwise
technically.
Studies are underway to assess how our field people can clean and
measure
return loss to ensure a certain ORL or point reflectance is maintained
over
time. Systems up to this point in time were very robust to return loss
degradation, so UPC has been acceptable as an SBC Method.  

IEEE EFM should specify systems at rates of one Gigabit and below to
operate
nominally under optical return loss near -20 dB to allow flexible use of
existing plant,  not just a pristine plant measured and verified to
sustain
-32 dB. This would allow use of older cables and connectors without
transmission issues from reflections.  OC-48 systems operate on our
legacy
plant and that is twice the rate of Gigabit Ethernet.  

Bottomline: Don't force operators to shift to APC for carrying digital
content at Gigabit or lower rates. SBC does not want the burden of
maintaining -32 dB or better ORL as a precondition to equipment working
especially when we know other types of transport does work more robustly
at
these speeds. 

-Kent

> -----Original Message-----
> From: Vipul Bhatt [mailto:Vipul_Bhatt@ieee.org] 
> Sent: Sunday, February 24, 2002 11:38 PM
> To: EFM Reflector
> Subject: [EFM] Cable plant question
> 
> 
> 
> Hello,
> 
> This is a question aimed at service providers and those who 
> believe they understand the implementation practices of 
> service providers.
> 
> Based on a couple of recent email threads on the P2P 
> reflector, I think it would be due diligence if I expose this 
> topic on the main EFM reflector. Let's keep it at a top level 
> here. Also, the question relates to P2P links only, not P2MP links.
> 
> Suppose the Optical PMD group came to you and asked: We are 
> examining the merits and demerits of a particular type of 
> PMD. It requires that the cable plant return loss be high, 
> say, more than 21 dB. It means the fiber at both ends of the 
> link will be have to be terminated with APC connectors. In 
> addition, you will have to guarantee that ALL fiber joints, 
> end to end, in your EFM cable plant will have to be either 
> splices or APC-to-APC connector joints. This goes for any 
> patch panels (fiber distribution boxes) you may have, either 
> at the CO end, or at the subscriber end. If they have been in 
> place before you deploy EFM links, you will have to search 
> and replace all PC/UPC connectors with APC.
> 
> Without entering the PMD debate (which is best done on the 
> P2P reflector), please tell me: do you find such a cable 
> plant proposition undesirable or acceptable? Very expensive 
> or marginally more expensive?
> 
> I recognize that the final decision on a PMD would depend on 
> many other factors, like the cost comparison of PMD options 
> and its system level implications. With all those caveats in 
> place, I am asking your opinion exclusively on the cable 
> plant question.
> 
> Thank you in advance.
> 
> Regards,
> Vipul
> 
> vipul_bhatt@ieee.org
> 408-857-1973
> 
> ======================
>