Conference npss::gigaswitch

    
    This is also the case with the GIX ( global internet exchange ) at the
    Royal Institute of Technology in Stockholm Sweden.
    
    ( This is where all the telco's Cisco's are connected with a Giga/FDDI)
    
    Gullik
    

    
    
    According to AT&T in Sweden, the 1237J lightwave transmitter should
    mate with any ST connector having specified ferrule diameter.
    
    They are not sure whether there should be an polymer alignement insert,
    or if the alignement should be ferrule to metallic body. They are
    looking into the matter.
    
    It seems likely that there is something wrong in the manufacturing
    process, since standard ST connctors fit badly in the transmitter.
    
    Gullik
    

    
    Could someone knowledgeable please respond...!
    
    Current findings is that several other links using DEFXS-BA indeed
    do work, even though the connector is NOT firmly aligned within the
    connector receptable. After checking at the Royal Inst. of Tech we
    even got our own link to loop, by placing the flat surface of the
    two SC shells against a glass plate(to get z-axis ok) and adjusting the 
    distance and position until the yellow led started blinking.
    
    Obviously the transmitter to st connector interface is not so critical
    since a the transmitter connector on a good working link can we wiggled
    enough so you can feel the lack of alignment easily. The subjective
    feel is that the gap is tent's of mm rather than thousand's ( which
    you'd not be able to feel)
    
    I always thought that SMF equpment required the outmost mechanical
    presision. How come this works anyway?
    
    Gullik
    

    The following mail entries should put this issue to rest. The ATT laser
    uses a split sleeve alignment, which does allow some wiggling of the
    ST connector. However, when the connector is allowed to rest in its
    locked position, the laser/connector alignment is correct, and links
    work.
    
    Dick Kirk
    Network Product Support
    
    
From:	NPSS::KIRK         "Dick Kirk DTN 226-7048"  7-MAR-1997 16:35:16.09
To:	STKHLM::WEBJORN
CC:	DELNI::ARIVERA,KIRK
Subj:	DEFXS-BA connector recommendations and a question

Gullik,

     I have some information regarding the ST connector recommended by ATT
     (Lucent) for the laser on the DEFXS-BA.  ATT has responded with
     recommendations to use ST2 connectors made by ATT (Lucent), Molex, 
     Siecor, or AMP.

     I also examined some of the lasers we have here. As I insert an ST
     connector into the receptacle, there is little resistance to the
     movement of the connector until the front edge of the body is 
     about 1 millimeter before the locking tabs. Some lateral freedom of 
     movement also is possible in this position.  As the connector is 
     inserted further, resistance is noticeable. The ferrule reaches
     the seated position when the front edge of the body is even with
     the locking tabs.  Pushing further compresses the spring so that the
     body can be rotated to engage the locking tabs.

     When I perform the same insertion with the same connector into
     the HP receiver, resistance is noticed for the whole insertion 
     distance.

     I believe that this resistance difference is due to the different
     receptacle design between the laser and the receiver. The ATT Laser
     has a split sleeve design in the receptacle. The HP receiver
     has a solid receptacle sleeve.  Resistance is not found in the
     ATT receptacle until the ferrule enters the split sleeve area.

     I think this explains why we feel some looseness of the connector
     while inserting.

     But I need to ask if a problem exists after the connector is fully
     inserted and locked in place.  If the links fail to get established,
     then we need to examine how much optical power is coupled into the
     single mode fiber.  Do you have an optical power meter?

Regards,

     Dick Kirk
     
    
From:	STKHLM::WEBJORN      "Gullik Webj�rn ( Gullik Webjorn ) Network advisory" 10-MAR-1997 04:35:31.94
To:	NPSS::KIRK
CC:	WEBJORN
Subj:	RE: DEFXS-BA connector recommendations and a question


	First, we have got our link up, using the patchcables supplied.

	All links at the KTH Giga are 'loose', i.e. can be wiggled back
	and forth, up and down. This does not seem to impair function.

	There is indeed a dimensional mismatch, but as you say, the
	AT & T transmitter device apparently does not  care.

	Our problems were in other areas, the cable network guys had
	sent a attenuation report with wrong fiber ID's at one end.

	This combined with a connector mismatch problem in the KTH
	cable plant caused us several days til the link came up.

	As it now looks, we are connected to the Internet backbone
	with a 100 mb single mode fiber, and the link seems OK.

	Regards,

		Gullik