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

Re: [10GMMF] Loss of worst case test configuration

Title: Message
in the past (1985-1997) we applied the 70/70 launch for the specification and test of all passive multimode products like ESCON connectors. The reason was that we think the mandrel wrap definition is not good enough in order get reproducable results: i) it depends on the tensile force when you are wrapping the fiber, ii) the amount of mode selection depends on the characteristics of the fiber coating of the individual fiber.
I think that because of the reason ii) it is very difficult to calculate the modal power distribution for such a setup, therefore we never tried to do this. 
-----Original Message-----
From: owner-stds-802-3-10gmmf@IEEE.ORG [mailto:owner-stds-802-3-10gmmf@IEEE.ORG] On Behalf Of Paul Kolesar
Sent: Donnerstag, 2. September 2004 02:48
Subject: Re: [10GMMF] Loss of worst case test configuration

thanks for filling in the missing piece on 70/70 launch.  I note that the 70/70 launch always produces higher loss than David's steady state launch for a given transverse offset.   My steady state loss data agrees most closely with David's calculation, and is empirically collected using the mode power distribution at the end of a long fiber of a vintage of at least 10 years ago.  

The differences between the 70/70 and the other steady state launch conditions lead me to ask you about what appears to be an inconsistency within fiber measurement standards regarding launch prescriptions.  The standards where such prescriptions can be found include TIA 455-50 (FOTP-50) and IEC-60793-1-40 appendix A.   FOTP-50 defines two launch methods. One is the 70/70 beam optics approach.  The other is the overfilled launch followed by mandrel wrap.   These are both described as creating steady state conditions that are roughly equivalent.  My question to you is whether there exist any means of analytically determining the mode power distribution caused by a particular mandrel wrap.    For example, the 62.5 um fiber prescription is 5 turns around a 20 mm diameter mandrel.  The 50 um prescription is 5 turns around a 25 mm mandrel.  In other words, is there a way to calculate the loss of high order mode energy due to a bend in the fiber and deduce the remaining guided mode power distribution so that one could plot the loss of a connection as a function of transverse offset?  There are discussions in IEC 86 that would benefit from this type of information.  Perhaps some of your colleagues are involved in them.

Paul Kolesar
1300 East Lookout Drive
Richardson, TX 75082
Phone:  972.792.3155
Fax:      972.792.3111

"Joerg-R. Kropp" <Joerg.Kropp@INFINEON.COM>
Sent by:

08/31/2004 02:02 PM
Please respond to "IEEE 802.3aq 10GBASE-LRM"

        cc:        (bcc: Paul F Kolesar/CommScope)
        Subject:        [10GMMF] Loss of worst case test configuration

Hi everyone,
In the discussion for the worst case test configuration there was some discussion about the loss of connector transverse offsets if the 70/70% launch condition (which should represent the so called steady state power distribution) is applied.
I calculated the connector loss vers transverse offset for various launch conditions including the 70/70% launch.
It turns out that the loss with the 70/70% launch is somewhat larger than the 23Ám offset launch. Therefore the total loss of our worst case test configuration would be larger than 1,5dB if such 70/70% launch would be applied.

It is common agreement today to use the defined offset patchcord and I do not want to change that but I want to share the results for the sake of completeness because it shows the sensitivity of our selection on the launch conditions.


Infineon Technologies Fiber Optics GmbH
Dr. Joerg-R. Kropp
Senior Staff Expert
Wernerwerkdamm 16
D-13629 Berlin / Germany
Phone -49-30-85400-4918
Fax     -49-30-85400-4900