dap@aber-cs.UUCP (Dave Price) (03/10/89)
Frustration on our site is growing rapidly as to which
thin coax is 'correct' for thin ethernet.
1/.Several manufacturers/vendors simply say RG58 while
other most definitely specify RG58A/U with others
equally emphatically state RG58C/U !!!
2/.Observation of the various cable specs seem to
say that even saying RG58A/U (for instance) does
not always get you the same cable.
3/.Many people now seem to be having cable with 'foam'
as apposed to solid dielectric.
4/.Some cable has a propagation velocity of
66% light speed while other has 77 or 78%.
Buying what you believe to be the 'correct' cable,
only to discover that some other expert tells you
'that wont work at 185 Metres' is very annoying.
A separate and equally frustrating problem is the length
of cable sections and distances between joints etc.
We often read 0.5 metres between 'tee' connectors, but
what about simple joints ??
We run a cable along an office row with boxes on the walls
with two bncs fitted. We then run two fly leads from the box
to the tee on the back of the machine. If a machine is removed
we insert one fly lead from one bnc to the other.
How long must my fly leads be ?? If I have no machine
in this office and just loop the bncs how long can
that lead be ?? 4 inches ?? or must it be 0.5 metres ??
Dave Price
--
UUCP : { ENGLAND or WALES }!ukc!aber-cs!dap
JANET: dap@uk.ac.aber.cs PHONE: +44 970 623111 x 3164
Post: University College of Wales, Penglais, Aberystwyth, UK, SY23 3BZ.morgan@Jessica.stanford.edu (RL "Bob" Morgan) (03/13/89)
In article <750@aber-cs.UUCP> Dave Price writes: >Frustration on our site is growing rapidly as to which >thin coax is 'correct' for thin ethernet. This isn't the answer you want, but our experience has shown that stringing as many machines/offices as possible on a piece of thin-net as close to 185 meters as possible is a False Economy and an invitation to frustration and worse. With the prices of multiport repeaters coming down as they have, the complexity of every network increasing, and the time any network manager has to fiddle around with piddly cable problems dwindling to nil, it's really worth it to insist on home runs to offices, or at most 2 to 3 offices per run, and to keep cable lengths well short of the recommended maximum if possible. I'll bet that we must have every brand of RG58-A/U and -C/U made somewhere on our campus, and I don't know of any that have failed (now, the TV guys who mixed RG59 and RG58 on some runs did have some trouble 8^). It's certainly worth it not to mix different cable batches/brands on a run (and the best way to do this is to have it be one piece of cable!). It's my impression that a properly-made connector-barrel-connector combination should present no impedance mismatch to the cable, and therefore its separation from other barrels or tees shouldn't matter at all. - RL "Bob" Morgan Networking Systems Stanford
tim@j.cc.purdue.edu (Timothy Lange) (03/14/89)
An interesting note to the length and type of thin Ethernet cable discussion. In all my Novell Netware manuals, the supplement for 3Com Ethernet cards states that the maximum segment length for thin Ethernet is 1000 feet. Two repeaters may be used to extend to 3000 feet. A thick Ethernet segment may go to 3200 feet. 185M is much less than 1000 feet, what gives?? -- Tim Lange. Purdue U. Computing Center/MATH Bldg./W. Lafayette, IN 47907/317-494-1787 Arpanet=tim@j.cc.purdue.edu/CIS=75410,525/Bitnet=TIM@PURCCVM/Fidonet=1:201/30
ron@ron.rutgers.edu (Ron Natalie) (03/15/89)
Ethernet segment size for regular Ethernet is 500 Meters, for thin, 200 meters. A segment is a set of cables joined togther with connectors (i.e. NOT repeaters). No two stations (transceivers) may have more than 1500 Meters and two repeaters between them. -Ron
morgan@Jessica.stanford.edu (RL "Bob" Morgan) (03/15/89)
In article <9113@j.cc.purdue.edu> tim@j.cc.purdue.edu (Timothy Lange) writes: >An interesting note to the length and type of thin Ethernet cable >discussion. In all my Novell Netware manuals, the supplement for 3Com >Ethernet cards states that the maximum segment length for thin >Ethernet is 1000 feet. Two repeaters may be used to extend to 3000 >feet. A thick Ethernet segment may go to 3200 feet. >185M is much less than 1000 feet, what gives?? 3Com's claim is that their transceiver design is superior, so that their units will work on a thin-net segment up to 1000 feet long, despite the IEEE 802.3 10Base2 spec being 185 meters (~600 feet). But "work" means communicate with each other, not necessarily with other brands of transceiver on the same 1000 foot segment. We try to keep our segments well under 600 feet, to maximize the likelihood of everyone's transceivers interoperating correctly. BTW, I seem to recall that the Western Digital 8003E has a jumper to set "long-cable mode" (or something like that) vs. normal mode, which is supposed to make it work at 1000 feet also. I never figured out why you'd want to *not* have it work at the longer length. Anybody know? - RL "Bob" Morgan Networking Systems Stanford
jerry@olivey.olivetti.com (Jerry Aguirre) (03/16/89)
In article <898@Portia.Stanford.EDU> morgan@Jessica.stanford.edu (RL "Bob" Morgan) writes: >BTW, I seem to recall that the Western Digital 8003E has a jumper to >set "long-cable mode" (or something like that) vs. normal mode, which >is supposed to make it work at 1000 feet also. I never figured out >why you'd want to *not* have it work at the longer length. Anybody >know? I don't have the 3Com manual handy but the Western Digital manual warns that the option for 300 meter segments requires that the network consist of a single segment and all the cards on that segment must be configured with that option. In other words this is for isolated networks between 180 and 300 meters that want to avoid the cost of a repeater. Jerry Aguirre
soley@ontenv.UUCP (Norman S. Soley) (03/18/89)
In article <9113@j.cc.purdue.edu>, tim@j.cc.purdue.edu (Timothy Lange) writes: > An interesting note to the length and type of thin Ethernet cable > discussion. In all my Novell Netware manuals, the supplement for 3Com > Ethernet cards states that the maximum segment length for thin > Ethernet is 1000 feet. Two repeaters may be used to extend to 3000 > feet. A thick Ethernet segment may go to 3200 feet. > > 185M is much less than 1000 feet, what gives?? What gives is that 3com is non-standard. As soon as you extend a segment beyond 185m it ceases to be 803.2 But Ethernet is robust enough that you can do all kinds of non-standard things to it and it'll still work. You can actually get away with it with most Ethernet products but only 3com claims to be specifically designed to do it in a reliable fashion. -- Norman Soley - Data Communications Analyst - Ontario Ministry of the Environment UUCP: uunet!mnetor!ontmoh!ontenv!soley | Contents of this message are OR: soley@ontenv.UUCP | my ideas, not the Ministry's "Stay smart, go cool, be happy, it's the only way to get what you want"
prc@maxim.ERBE.SE (Robert Claeson) (03/20/89)
In article <Mar.14.13.37.48.1989.459@ron.rutgers.edu>, ron@ron.rutgers.edu (Ron Natalie) writes: > Ethernet segment size for regular Ethernet is 500 Meters, for thin, 200 meters. > A segment is a set of cables joined togther with connectors (i.e. NOT repeaters). > No two stations (transceivers) may have more than 1500 Meters and two repeaters > between them. I've heard about a thing called "inter-repeater link". It should be used instead of a single repeater to connect two segments when they are some distance apart from each other (such as in two different buildings). According to the information I have (from ISOLAN), such a link is made by using a normal repeater on each of the two segments that should be connected to each other and connecting them with a normal ethernet cable. This extra segment (the inter-repeater link) doesn't count as a segment and can be twice as long as a real segment. But -- there must be no other devices connected to it than the two repeaters at each end. seg 1 seg 2 --------- ------------------------- : : +--+--+ +--+--+ : R : : R : +--+--+ +--+--+ : IRL : ============================================ -- Robert Claeson, ERBE DATA AB, P.O. Box 77, S-175 22 Jarfalla, Sweden Tel: +46 (0)758-202 50 Fax: +46 (0)758-197 20 EUnet: rclaeson@ERBE.SE uucp: {uunet,enea}!erbe.se!rclaeson ARPAnet: rclaeson%ERBE.SE@uunet.UU.NET BITNET: rclaeson@ERBE.SE