[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference 7.286::fddi

Title:FDDI - The Next Generation
Moderator:NETCAD::STEFANI
Created:Thu Apr 27 1989
Last Modified:Thu Jun 05 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:2259
Total number of notes:8590

525.0. "SM to MM converters." by STKHLM::WEBJORN (Gullik Webj�rn Product & Technology Group) Thu Apr 02 1992 08:27

    Does anyone have comments on single_mode<->multimode f/o converters?
    
    We have a sales situation where we would like to install single-mode
    concentrator mgmt modules, but *temporarily* use a multimode fiber
    for one leg of the ring. In the next stage ( ~ 6 months) two
    new singlemode fibers will be installed to a third location.
    The mm fiber is already there.
    
    The PTT want's to solve this problem with sm-mm converters.
    
    What is our position in this case? What are our requirements
    on the converters? How  do we test/certify the fiber link?
    
    I guess these converters are phone_trunk converters, designed
    for 140 Mbit's trunk service. Would the expected error-rate
    be (much) worse than the requirement for FDDI? ( Hello Paul!)
    
    Besides, this ring will be used for a fault-tolerant vaxcluster!!
    
    Gullik
T.RTitleUserPersonal
Name		DateLines
525.1Where no man has gone before!POLAR::NEUMANNThu Apr 02 1992 09:3610
    If you connected the multimode fiber directly to the singlemode
    concentrator (you'd have to use the appropriate connectors, then the
    link would almost certainly work, in terms of optical loss budget. 
    BUT, you are entering unknown territory.  There may be a problem with
    modal noise, in which reflections from the multimode fiber back into
    the laser are timevarying and may cause bit errors.  Contact the people
    who have worked on the subject.  Try Bruce Schofield, DTN 226-7547.
    
    Rick
525.2KONING::KONINGPaul Koning, NI1DThu Apr 02 1992 10:0321
Re .0: sm to mm converters MAY work.  Bit error rate is one issue, but another
critical one to FDDI is what happens when the link first turns on.  Single
mode transmitters normally have a fairly significant turn-on delay, and FDDI
has specific requirements in that area.  General purpose converters may not
meet those requirements.

An issue that relates to error rate is jitter: these converters will introduce
some additional jitter.  For short links this sort of setup may work adequately.
An interesting twist is the fact that you're talking about the opposite case
from what we've previously considered: single mode drivers and multimode fiber
rather than the other way around.  For the "usual" case we think that converters
can work but certainly not to the same distances as a "real"  single mode setup.
I'd speculate that the same might apply here -- in other words, I hope you're
not trying to go the full 2 km on the multimode leg.  Clearly you'd need to do
some careful testing here.

Is there some more straightforward alternative?  How about using line cards
intended for this setup (i.e., with multimode transceivers) for now, and then
trading them in for single mode cards when the single mode link arrives?

	paul
525.3SM over MM could be a solution...'STKHLM::WEBJORNGullik Webj�rn Product &amp; Technology GroupThu Apr 02 1992 12:3422
    
    Very interesting, I haven't considered running single mode eqip. into
    multimode fiber, but assuming the loss budget is within specs, why not!
    
    Anyway, Paul, I think the DEMO center has 2 concentrators, we could
    temporarily swap these for the customers. Problem is of cource that
    the people at the DEMO center need to be taught how to handle single 
    mode fiber, i.e. dont touch connectors, always keep *CLEAN* use proper
    tools, torque etc.
    
    While on the subject, someone with real experience, how tricky is S/M
    to handle? I always figured that the stuff deposited form your thumb
    onto a connector would be several �m, so you'd need to be *VERY*
    careful with cleanliness etc. How big a trouble is this? What about
    mating / remating S/M connectors? Doesn't the ferrule compress each
    time? 
    
    Last is there any risk of damaging the APD if you miscalculate fiber
    loss, or is the laser power still harmless for the optics?
    
    Gullik
525.4mm to sm more common!PERE::BRUCEFri Apr 03 1992 16:105
    i know that the auto strada project in italy has been using mm to sm
    converters for quite some time. a contact is BLG01::STRATOS sarissamlis
    in bologna. i don't know if mm-sm has the same problems as sm-mm
    though!
525.5Use the Sales Mechanisms ... avoid the support issuesSYORPD::DEEPBob Deep - SYO, DTN 256-5708Mon Apr 06 1992 11:1917
I had a customer who was trying to get Cisco mm to run over sm with ODS line 
drivers.  Had lots of trouble.

I would suggest that you supply the customer with concentrators configured 
for one port mm, one port sm, until the new links are run.  It will save 
support $$ in the long run.   One solution if to "rent" them the mixed
board, and then sell the sm-sm boards when the links are available.  The
"rent" would be less than or equal to the cost of the sm-mm converters, as
determined by the account manager (or an LOP, if that works better for that
account).

I'd stay away from the line drivers (converters) unless you have time to 
work the support issues.

My $.02

Bob