Conference 7.286::fddi

    Pardon me for asking, but what does "SDH" stand for?
    
    As far as our FDDI receivers working flawlessly at 1550nm, the
    receivers are most sensitive at 1300nm, and I believe their sensitivity
    drops off beyond 1350, and below 1250. I don't have any experience
    with wave shifters, although, I've seen some shape shifters on tv. :-)
    Sounds like the supplier of the wave shifters is trying to push his
    product, and by claiming that often 1300nm receivers can receive also
    at 1550nm, probably means he doesn't understand FDDI & fiber optics
    well.....Once the supplier sells his product, he'll probably 
    dissapear, and you'll end up being responsible for getting it
    to work realiably.... JMHO.
    I wouldn't recommend telling customer our receivers might work
    marginally at 1550nm. You'd just be back there later trying to get 
    something working that it wasn't designed for. I'd encourage matching
    like wavelengths, and would stay away from trying to get something
    to work outside of its intended design range.
    
    Bob

    SDH = Synchronous Digital Hierarchy, the "international" variant of
    SONET. SDH data rates vary from 155.52 Mbits/s to over 9 Gbit/s
    (although not dynamically -- SDH ports are almost always fixed at one
    data rate).
    
    What you are talking about is 1/2 of WDM (wavelength division
    multiplexing), in which you have taken two independent data streams and
    piped them thru one fiber by keeping using two independent optical
    carriers (one at 1550 and the 1300 nm). The other half of WDM, the
    de-multiplexing of the optical carriers from the fiber, you have not
    done. You imply that both the 1300 and 1550nm optical carriers will be
    received by a receiver that cannot descriminate between the two
    wavelengths. Therefore, the receiver will try to process the SDH and
    FDDI data streams, which will not work. Is this what you really meant?
    
    In any case, there are several questions you will need to answer before
    you can proceed with such a system:
    
    	1) Will the rcvr operate at 1550nm (your original question)?
    	2) Is the fiber in the cable plant capable of operating at 1550nm?
    	3) Will there be an optical de-mux at the receivers that will
    		separate the 1300 from the 1550nm?
    	4) Will the "waveshifter" at TX shift the FDDI to 1550nm with
    		out seriously degrading the signal?
    
    There may be some other concerns, too, that i've not listed.
    
    I expect that another noter, one with optical experience, will also
    chime in here.
    
    bill

    I have asked the vendor of the receiver on our SMF PMD about the
    performance of the receiver at 1550 nm.  We should have a response
    in a few days.
    
    Bill is correct about the issue of de-multiplexing at the receiver.
    Our receiver will not be able to discriminate 1300 nm data from 1550
    nm data. And if the de-muxing is done poorly (i.e. if some of the 1300
    nm data stays 'mixed' with the 1550 nm data) the receiver will get
    the sum of the two data sets.  This could lead to things like bit
    errors.
    
    Bob offers some very good cautions about unique uses and supportable
    uses of products.
    
    As to the performance of the fiber, in GENERAL single-mode fiber that
    is used at 1300 nm is pretty good at 1550.  The attenuation of the
    fiber at 1550 nm may be an issue.  Older SMF product could exhibit some
    bend induced attenuation in the 1550 nm region.  In all cases, the SMF
    cable plant should be measured for loss at each wavelength of use. 
    There are ODTR (Optical Time Domain Reflectometers) that work at 1300
    and/or 1550 nm.
    
    When I receive data from the Rx vendor, I'll post another response.
    
    
    Dick Kirk
    Network Product Support

    Our vendor has replied to our inquiry.  Their response is:
    
    "The responsivity of our PIN diodes remains pretty constant from 
    1200 to 1650nm."
    
    
    This is not very quantitative, but given the relatively short link -
    3500 meters, the receiver should have sufficient sensitivity to run
    at 1550 nm.   PLEASE NOTE that this is untested and not guaranteed to
    work.
    
    Dick Kirk
    Network Product Support