[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

1429.0. "FDDI/10BaseT mismatch consequences ?" by NWD002::TUTAK_PE (It's like working in Bosnia) Mon Aug 22 1994 13:57

    
    I have a megalithic, Washington state-based software developer/distributor
    (and birgeoning Gigaswitch customer) whose IS department wants to know 
    the answer to the following question:
    
    What are the ramifications of inadvertently plugging a 10BaseT
    interface connection (PC and NIC) into a twisted pair cabling run that
    connects to an FDDI over UTP device interface ? Will either interface
    and/or associated firmware become disabled, or will it just sit there and 
    do nothing, but with no harm coming to either interface ? What kinds of
    problems are caused ?
    
    Conversely, what are the ramifications of inadvertently plugging a UTP
    FDDI interface connection (PC and NIC) into a run that connects to a
    10BT repeater ?
    
    This question comes in the guise of planning administrative procedures
    for an extensive FDDI to the desktop environment...
    
    Any help appreciated...
    
    Thanks... 
    
    Peter 
T.RTitleUserPersonal
Name
DateLines
1429.1Either network is protectedLEVERS::B_CRONINThu Aug 25 1994 10:0127
    
    The UTP FDDI transmits on pins 1&2, of the RJ45, and receives on pins 
    7&8. The crossover is in the cable plant (i.e. one of the patch cords). 
    
    10Base-T uses pins 1,2, 3 and 6, with the crossover usually in the 
    repeater. So the DTE is transmitting on 1&2, receiving on 3&6, and the 
    repeater is receiving on 1&2, and transmitting on 3&6. If the crossover
    is in the cable plant (as it was with the early 10BASE-T repeaters)
    both ends will transmit on 1&2, and receive on 3&6.
    
    So, there isn't any danger of a completed DUPLEX connection being formed, 
    as the pins don't form the required connections. This means that if a 
    10BASE-T station is connected to an FDDI port, no connection will
    be formed, and the ring is protected (because the FDDI CMT protocols
    won't allow the connection). 
    
    If a UTP FDDI station plugs into a 10BASE-T repeater, it is
    possible that the FDDI transmitter could be seen by the repeater's 
    receiver (i.e. the pins would connect to each other). But, the 
    10BASE-T repeater will need to see linkbeat pulses in order for the 
    link to wake up, and, since the FDDI does not produce linkbeat pulses,
    the repeater will also not recognize the station, so the 10BASE-T 
    network is protected. 
    
    
    
    
1429.2Last reply has an error56821::B_CRONINWed Oct 12 1994 13:5732
    Well, I was almost right..........
    
    It turns out that my answer in -.1 is wrong for the case where the
    FDDI UTP station is connected to a 10BASE-T repeater with all straight 
    through cables. The other cases are correct.
    
    Specifically, if pins 1&2 of the FDDI station connect to pins 1 and 2 
    of a repeater with an internal crossover, the FDDI can open-up the
    10BASE-T repeater, and affect the 10BASE-T network. The level of 
    junk varies by repeater type. 
    
    The oversight was that I forgot how linkbeat really worked. Linkbeat 
    is really only qualifying against a squelch threshold, not a recived
    waveform mask. If there is enough power in the cable, the linkbeat
    detector will open up (its only required to reject +/- 50 mV. So, 
    a +/- 1 Volt UTP FDDI signal will cause the 10BASE-T linkbeat circuit 
    to open up. The receiver will then produce random garbage that gets 
    repeated onto the network.
    
    For the 2 repeaters we looked at, it appears that the receiver 
    opens up and produces very short events, which the repeater 
    extends to be 96 bit runts. So, on the backbone there are a lot 
    of runt packets. They were running around 1000/second in the tests 
    we ran. This will probably vary by repeater. Some repeaters will
    partition, others will just keep repeating the junk. It cannot be 
    predicted in advance. 
    
    Sorry for any confusion,