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

RE: [EFM] AW: [EFM-P2P] RE: Single wavelength, single fiber PMD f or P2P




Walt, Tom,

Bad optical connection (high reflection or attenuation) is equivalent to
cold solder joint.  It is unpredictable in its performance over time,
humidity and temperature, and hence un-reliable.  

Different optical network distribution specifications place requirements
on the network performance, both for attenuation and reflection, high
enough to guarantee detection of bad connection at the time of
installation or repair.  

Similar issue on fiber patch panels.  Marginal connection is highly
sensitive to mechanical movements when other connections are made on the
same panel due to fiber movements.

Currently service providers have learnt how to deal with fibers in mass
installations, and they use the stringent specs for a reason.

Meir

-----Original Message-----
From: Thomas.Murphy@xxxxxxxxxxxx [mailto:Thomas.Murphy@xxxxxxxxxxxx]
Sent: Tuesday, March 05, 2002 12:59 AM
To: walt.kutzavitch@xxxxxxxxxxx; Thomas.Murphy@xxxxxxxxxxxx
Cc: owner-stds-802-3-efm-p2p@majordomo.ieee.org;
stds-802-3-efm@ieee.org; stds-802-3-efm-p2p@majordomo.ieee.org
Subject: AW: [EFM] AW: [EFM-P2P] RE: Single wavelength, single fibre PMD
f or P2P



Hello Walt,
 
Thanks for your interesting input on this issue.  We have spent
considerable time discussing this and other
related issues in the preparation of the preparation.  We are proposing
a system 
which will work in a worst case ORL environment of 20 dB, which based on
feedback
from a number of network providers is a very safe number for the lower
limit.  
 
The scenario you present is of relevance to all PMD types, not just 
single wavelength, single fibre solutions.  If you examine required ORL
of connectors
from 802.3 you will see a value of >25dB, irrespective of PMD type.  A
dust
particle in the connector will violate this condition and hence render
the 
network as 'out of spec'
 
The situation with splices is then same.
 
Regards
 
tom
 
 

-----Ursprüngliche Nachricht-----
Von: walt.kutzavitch@xxxxxxxxxxx [mailto:walt.kutzavitch@xxxxxxxxxxx]
Gesendet am: Montag, 4. März 2002 21:11
An: Thomas.Murphy@xxxxxxxxxxxx
Cc: owner-stds-802-3-efm-p2p@majordomo.ieee.org;
stds-802-3-efm@ieee.org; stds-802-3-efm-p2p@majordomo.ieee.org
Betreff: Re: AW: [EFM] AW: [EFM-P2P] RE: Single wavelength, single fibre
PMD f or P2P

Just a side track that may be of interest. 

I've had 30+ years of experience in designing telecommunications 
systems for residence and small and medium business. 
Two things come to mind. 

First: 
I think it's important to remember where the ONU could be  located. 
The ONU could be on the outside of the building. In fact that is the 
likely spot. What happens when a service visit is required? 

The service technician can't order a nice day for the visit. 
The service call could be at a time of  dirty dusty conditions. 
The service call may be in the rain or snow. 
There may be mud and building material all around. Even plaster. 
A service call could result in disconnecting the fiber. 
What will happen to the exposed end of the fiber during such a service
visit. 
What will happen to the ONU end? 
The cleanliness of the fiber connection may be compromised. 
Up until now, fiber existed in wiring closets. The SOHO environment is
quite different. 

Second: 
What about moves and changes. 
I've seen situations where the equipment had to be moved many feet to
accommodate 
remodeling, etc. How may times can the fiber be spliced? How much
lateral force can be applied 
to the connector without changing the characteristics of the interface. 

We had better consider such conditions. 

The single wavelength P2P that requires a high return loss from the 
fiber link may be impractical. 




	Thomas.Murphy@xxxxxxxxxxxx 
Sent by: owner-stds-802-3-efm-p2p@majordomo.ieee.org 


02/27/2002 01:57 AM 


        
        To:        plee@xxxxxxxxxxxxxxxxxxxx
Thomas.Murphy@xxxxxxxxxxxx
Jack.Jewell@xxxxxxxxxxxxx
mdudek@xxxxxxxxxxxx
Hans.Mickelsson@xxxxxxxxxxxxxxx 
        cc:        stds-802-3-efm@ieee.org
stds-802-3-efm-p2p@xxxxxxxxxxxxxxxxxx
bob.barret@xxxxxxxxxxxxxxxxxx
david@xxxxxxxxxxxxxx
doravv@xxxxxxxxxx
FEffenberger@xxxxxxxxxxxxxxxxx
jradcliffe@xxxxxxxxxxxxxxxxxxxx
jstiscia@xxxxxxxxxx
mark.sankey@xxxxxxxxx
meir@xxxxxxxx
n.kleiner@xxxxxxxxxxxx
PengL@xxxxxxxxxxx
raanan@xxxxxxxxxxxxxxx
rbrand@xxxxxxxxxxxxxxxxxx
sasaki144@xxxxxxx
schelto.vandoorn@xxxxxxxxx
Tonyshouse@xxxxxxx
wdiab@xxxxxxxxx
Vipul_Bhatt@xxxxxxxx
(bcc: Walt Kutzavitch/TMP/SC/PHILIPS) 
        Subject:        AW: [EFM] AW: [EFM-P2P] RE: Single wavelength,
single fibre PMD f         or P2P 

        Classification:         






Pete,

You are correct in your appraisal of the temperature comparison 
of the two systems. My reasoning for this was that some DFBs demonstrate
an intensity variation over temperature, however as was pointed out by
me and others,
this can be accounted for by chip design. VCSEL temperature response is
assumed to be
the same but not yet proven. I am willing to accept that both a 1-lambda
and 2-lambda
have comparable temperature responses, hence making this a non-issue.

Tom



-----Ursprüngliche Nachricht-----
Von: Lee, Pete [mailto:plee@xxxxxxxxxxxxxxxxxxxx]
Gesendet am: Mittwoch, 27. Februar 2002 00:25
An: 'Thomas.Murphy@xxxxxxxxxxxx'; Jack.Jewell@xxxxxxxxxxxxx;
mdudek@xxxxxxxxxxxx; Hans.Mickelsson@xxxxxxxxxxxxxxx
Cc: stds-802-3-efm@ieee.org; stds-802-3-efm-p2p@majordomo.ieee.org;
bob.barret@xxxxxxxxxxxxxxxxxx; david@xxxxxxxxxxxxxx; doravv@xxxxxxxxxx;
FEffenberger@xxxxxxxxxxxxxxxxx; jradcliffe@xxxxxxxxxxxxxxxxxxxx;
jstiscia@xxxxxxxxxx; mark.sankey@xxxxxxxxx; meir@xxxxxxxx;
n.kleiner@xxxxxxxxxxxx; PengL@xxxxxxxxxxx; raanan@xxxxxxxxxxxxxxx;
rbrand@xxxxxxxxxxxxxxxxxx; sasaki144@xxxxxxx;
schelto.vandoorn@xxxxxxxxx; Tonyshouse@xxxxxxx; wdiab@xxxxxxxxx;
Vipul_Bhatt@xxxxxxxx
Betreff: RE: [EFM] AW: [EFM-P2P] RE: Single wavelength, single fibre PMD
f or P2P

I don't see where the one-wavelength system has the temperature
advantage.
My reasoning is this:  For the 2-wavelength syetem, choose the upstream
wavelength as 1310nm, then one would still pick an FP laser as the ONU
laser.  So, both systems use an FP laser at the ONU.  If 1550nm were
chosen
for the donwstream wavelength, then this laser is located in the C.O.,
which
would see much less of a temperature extreme, and the advantage is less
pronounced.

Pete Lee


-----Original Message-----
From: Thomas.Murphy@xxxxxxxxxxxx [mailto:Thomas.Murphy@xxxxxxxxxxxx]
Sent: Monday, February 18, 2002 9:36 AM
To: Jack.Jewell@xxxxxxxxxxxxx; Thomas.Murphy@xxxxxxxxxxxx;
mdudek@xxxxxxxxxxxx; Hans.Mickelsson@xxxxxxxxxxxxxxx
Cc: stds-802-3-efm@ieee.org; stds-802-3-efm-p2p@majordomo.ieee.org;
bob.barret@xxxxxxxxxxxxxxxxxx; david@xxxxxxxxxxxxxx; doravv@xxxxxxxxxx;
FEffenberger@xxxxxxxxxxxxxxxxx; jradcliffe@xxxxxxxxxxxxxxxxxxxx;
jstiscia@xxxxxxxxxx; mark.sankey@xxxxxxxxx; meir@xxxxxxxx;
n.kleiner@xxxxxxxxxxxx; PengL@xxxxxxxxxxx; raanan@xxxxxxxxxxxxxxx;
rbrand@xxxxxxxxxxxxxxxxxx; sasaki144@xxxxxxx;
schelto.vandoorn@xxxxxxxxx; Tonyshouse@xxxxxxx; wdiab@xxxxxxxxx;
Vipul_Bhatt@xxxxxxxx
Subject: AW: [EFM] AW: [EFM-P2P] RE: Single wavelength, single fibre PMD
f or P2P



What I meant is that with changes in temperature, the VCSEL/DFB grating
wavelength position
shows small shifts (arising from changes in the refractive index),
however,
what 
happens is that the gain curve moves away from this fixed grating
position
and
hence the intensity drops. With an FP,
the gain curve and centre wavelength show the same temperature
dependence,
hence the
intensity variations are smaller.  This is why fewer (not none) DFBs are
specified over 
an extended temperature range.

Tom



-----Ursprüngliche Nachricht-----
Von: Jack Jewell [mailto:Jack.Jewell@xxxxxxxxxxxxx]
Gesendet am: Montag, 18. Februar 2002 16:22
An: 'Thomas.Murphy@xxxxxxxxxxxx'; mdudek@xxxxxxxxxxxx;
Hans.Mickelsson@xxxxxxxxxxxxxxx
Cc: stds-802-3-efm@ieee.org; stds-802-3-efm-p2p@majordomo.ieee.org;
bob.barret@xxxxxxxxxxxxxxxxxx; david@xxxxxxxxxxxxxx; doravv@xxxxxxxxxx;
FEffenberger@xxxxxxxxxxxxxxxxx; jradcliffe@xxxxxxxxxxxxxxxxxxxx;
jstiscia@xxxxxxxxxx; mark.sankey@xxxxxxxxx; meir@xxxxxxxx;
n.kleiner@xxxxxxxxxxxx; PengL@xxxxxxxxxxx; raanan@xxxxxxxxxxxxxxx;
rbrand@xxxxxxxxxxxxxxxxxx; sasaki144@xxxxxxx;
schelto.vandoorn@xxxxxxxxx; Tonyshouse@xxxxxxx; wdiab@xxxxxxxxx;
Vipul_Bhatt@xxxxxxxx
Betreff: RE: [EFM] AW: [EFM-P2P] RE: Single wavelength, single fibre PMD
f or P2P

This is incorrect.  The grating (or cavity) structure of a VCSEL or DFB
is 
exactly what allows VCSELs (and to some extent DFBs) to display little
variation in threshold and/or power over large temperature ranges.

Jack

-----Original Message-----
From:                  Thomas.Murphy@xxxxxxxxxxxx
[mailto:Thomas.Murphy@xxxxxxxxxxxx] 
Sent:                 Saturday, February 16, 2002 10:39 AM
To:                 mdudek@xxxxxxxxxxxx; Hans.Mickelsson@xxxxxxxxxxxxxxx
Cc:                 Thomas.Murphy@infineon.com; stds-802-3-efm@ieee.org;
stds-802-3-efm-p2p@xxxxxxxxxxxxxxxxxx; bob.barret@xxxxxxxxxxxxxxxxxx;
david@xxxxxxxxxxxxxx; doravv@xxxxxxxxxx; FEffenberger@xxxxxxxxxxxxxxxxx;
jradcliffe@xxxxxxxxxxxxxxxxxxxx; jstiscia@xxxxxxxxxx;
mark.sankey@xxxxxxxxx;
meir@xxxxxxxx; n.kleiner@xxxxxxxxxxxx; PengL@xxxxxxxxxxx;
raanan@xxxxxxxxxxxxxxx; rbrand@xxxxxxxxxxxxxxxxxx; sasaki144@xxxxxxx;
schelto.vandoorn@xxxxxxxxx; Tonyshouse@xxxxxxx; wdiab@xxxxxxxxx;
Vipul_Bhatt@xxxxxxxx
Subject:                 [EFM] AW: [EFM-P2P] RE: Single wavelength,
single fibre PMD
for P2P


In the analysis of the temperature response of a VCSEL (and DFB) I 
as refering to intensity variations and not shifts in wavelength.
The grating structure of both implies a larger sensitivity in
output power to temperature shift.

Tom



-----Ursprüngliche Nachricht-----
Von: Mike Dudek [mailto:mdudek@xxxxxxxxxxxx]
Gesendet am: Freitag, 15. Februar 2002 19:35
An: Hans Mickelsson (ERA)
Cc: 'Thomas.Murphy@infineon.com'; stds-802-3-efm@ieee.org;
stds-802-3-efm-p2p@xxxxxxxxxxxxxxxxxx; bob.barret@xxxxxxxxxxxxxxxxxx;
david@xxxxxxxxxxxxxx; doravv@xxxxxxxxxx; FEffenberger@xxxxxxxxxxxxxxxxx;
jradcliffe@xxxxxxxxxxxxxxxxxxxx; jstiscia@xxxxxxxxxx;
mark.sankey@xxxxxxxxx; meir@xxxxxxxx; n.kleiner@xxxxxxxxxxxx;
PengL@xxxxxxxxxxx; raanan@xxxxxxxxxxxxxxx; rbrand@xxxxxxxxxxxxxxxxxx;
sasaki144@xxxxxxx; schelto.vandoorn@xxxxxxxxx; Tonyshouse@xxxxxxx;
wdiab@xxxxxxxxx; Vipul_Bhatt@xxxxxxxx
Betreff: Re: [EFM-P2P] RE: Single wavelength, single fibre PMD for P2P

VCSEL change in wavelength as a function of temperature is
similar/slightly
better than the quoted number for the DFB.

"Hans Mickelsson (ERA)" wrote:

> Tom,
>
> Good work, see below for some comments.
>
> "Fewest variations of PMD to track in field". I agree that it will be
less
PMDs to track however if this is a real problem is not clear. Assume
that
you decide to use a certain lamda for downstream then it it ijust to
follow
this approach in all your switches and the problem will be non-existing.
>
> Can you explain "Use of inexpensive TOSA a bit further"?
>
> "Temperature performance". An un-cooled DFB (0.01 nm/K) is less temp
dependent than a FP (0.03 nm/K). Can anyone give some figures for a
VCSEL?
>
> "Leverage of 1310 nm....", True, but you raise an interesting question
here. The use of parallel optics will be very difficult with any single
fiber solution.
>
> Some othe options to consider are;
>
> The need for angled polished connectors for controlling refelctions in
the
network. Not necessary for dual but probably necessary for single.
>
> Will the be any difference between the two options in an up-grade
scenario
to higher bit-rates?
>
>         Brgds//HANS MICKELSSON
>
> > -----Original Message-----
> > From: Thomas.Murphy@xxxxxxxxxxxx [mailto:Thomas.Murphy@xxxxxxxxxxxx]
> > Sent: den 14 februari 2002 17:25
> > To: stds-802-3-efm@ieee.org; stds-802-3-efm-p2p@majordomo.ieee.org;
> > bob.barret@xxxxxxxxxxxxxxxxxx; david@xxxxxxxxxxxxxx;
> > doravv@xxxxxxxxxx;
> > FEffenberger@xxxxxxxxxxxxxxxxx; hans.mickelsson@xxxxxxxxxxxxxxx;
> > jradcliffe@xxxxxxxxxxxxxxxxxxxx; jstiscia@xxxxxxxxxx;
> > mark.sankey@xxxxxxxxx; meir@xxxxxxxx; Thomas.Murphy@xxxxxxxxxxxx;
> > n.kleiner@xxxxxxxxxxxx; PengL@xxxxxxxxxxx; raanan@xxxxxxxxxxxxxxx;
> > rbrand@xxxxxxxxxxxxxxxxxx; sasaki144@xxxxxxx;
> > schelto.vandoorn@xxxxxxxxx; Tonyshouse@xxxxxxx; wdiab@xxxxxxxxx
> > Cc: Vipul_Bhatt@xxxxxxxx
> > Subject: Single wavelength, single fibre PMD for P2P
> >
> >
> > Hello All,
> >
> > First off I apologise for a blanket bomb approach with
> > sending out this e-mail but
> > I have my reasons, as will become clear below.
> >
> > As most of you will know, I am currently co-ordinating the
development
> > of a single fibre, single wavelength PMD proposal for P2P links.
> > During a recent discussion of this work, it became clear that
> > a number of people
> > were unaware of what was happening or still unclear of some
> > of the technical issues involved. Several questions arose
> > which had already been answered in the course of dedicated
> > telephone conferences and
> > some new issues were also raised.
> >
> > Of course this discussion is very good for the quality of the
> > standard. However,
> > in the interest of progress and completion of a baseline
> > proposal for March, it
> > is essential that people allay any concerns they may have in
> > the interim and go
> > into the meeting feeling ready to make an informed decision.
> >
> > I would therefore ask people who have an opinion (or concern)
> > in this direction
> > to speak up, raise the issues and avoid further surprises in St
Louis.
> >
> > Arising from the aforementioned discussions was the idea that
> > a 2 wavelength
> > PMD may be an alternative approach. In order to facilitate
> > comparison of the
> > two ideas, a matrix was proposed which I have included here
> > in a somewhat modified
> > form. I divided the table into Today and Future, the former
> > represent current laser sources (FP and DFB)
> > and the later assuming the use of VCSELs, be that at 1310,
> > 1490 or 1550 nm. I dislike the 1-10
> > approach of comparing as this is too subjective, rather a
> > binary 0 or 1 representing the better solution
> > for a particular criteria. In some cases there are no
> > differences and both receive 0. The 'points'
> > are added and a comparison may be made.  NOTE, this will not
> > be the basis of the decision, rather an aid
> > to objective comparison. I have included comments behind each
> > issue cells detailing my evaluation.
> >
> > So, speak up, play with the table, add issues if necessary,
> > send it back to me,
> > get on the telephone conferences.
> >
> > Best regards and looking forward to further progress.
> >
> > Tom and the P2P group
> >
> >  <<Comparison matrix for 1 and 2 wavelength PMDs.xls>>
> >