km@emory.UUCP (Ken Mandelberg) (02/02/86)
I would appreciate any info (or a pointer) on thin ethernet.
In particular I need to know what limitations there are in
using it compared with regular ethernet. This just suddenly
became interesting because of Sun's new Model 3/50 workstations
with built in thin ethernet transceivers.
Unless I misunderstand, I can just wire a series of offices with
thin ethernet, leaving an exposed T-connector in each. Then as the
the need arises just cable the Sun to open tap on the T. If I
need to connect a conventional piece of ethernet equipment in
an office, I cut out the T, solder in some N-connectors and
screw in a standard transceiver.
--
Ken Mandelberg
Emory University
Dept of Math and CS
Atlanta, Ga 30322
{akgua,sb1,gatech,decvax}!emory!km USENET
km@emory CSNET
km.emory@csnet-relay ARPANETsutter@osu-eddie.UUCP (Bob Sutterfield) (02/03/86)
> Unless I misunderstand, I can just wire a series of offices with > thin ethernet, leaving an exposed T-connector in each. Then as the > the need arises just cable the Sun to open tap on the T. If I > ... > Ken Mandelberg > Emory University > {akgua,sb1,gatech,decvax}!emory!km USENET > km@emory CSNET > km.emory@csnet-relay ARPANET *** REPLACE THIS ailing cable WITH YOUR wire cutters *** If by "a series of offices" you mean to daisy-chain a series of stations via cables in the walls or above suspended ceilings, please consider what will happen if you have a cable fault - how much gypsum board and suspended ceiling dust do you consider reasonable on your equipment and down the back of your collar, before you dig out enough cable to find the fault? And how many workstations "downstream" do you want out of commission during that time? DEC, at least, suggests installing a DELNI-like thing (that they have just announced) that is a repeater between a thick backbone and eight skinny legs. Each of these legs, they suggest, should go to uniquely one wall jack, and any daisy-chaining of workstations should be done on the bench or along the baseboard. This way, you can isolate the faults in the open air and a fault on one leg will only affect the workstations in that office or lab - presumably all easily visible from each other. Also, "downstream" users will probably be officemates or coworkers in the same lab, and therefore (we hope) more sympathetic... It sounds pretty reasonable to me, but then I haven't worked with this particular type of cable yet, just several others. Any opinions from anyone who has touched the stuff? -- ----- Human: Bob Sutterfield Mail: sutter@osu-eddie.UUCP sutterfield-r%osu-20@osu-eddie.UUCP or: sutter@osu-eddie.ohio-state.CSNET sutterfield-r@osu-20.ohio-state.CSNET
brian@sdcsvax.UUCP (Brian Kantor) (02/04/86)
Actually, a good scheme is to run thin ethernet cable into each office, and somewhere in the office cut the cable, install two BNC male connectors, and a BNC barrel-fitting (double female union) to reconnect the cable. Then when you want to install some equipment in that room, just pop the barrel-fitting out and install one or more BNC T-connectors. You can add a loop of cable at the same time if you have more than one device to install. And if you want to put in a standard "Thick-Ethernet" transceiver, you just use a pair of BNC-to-N between series adaptors. And you can attach your thin ethernet cable to the end of a thick ethernet using the same adaptors. The reasons for not leaving an exposed T-fitting on the cable is that they 1) radiate interference from the open end, and 2) might short out if something metallic (staples, etc) happens to fall into the open end. You can buy non-shorting caps for the T-fittings, but barrel fittings are cheaper in the first place. Believe it or not, you can probably get all those kinds of connectors at Radio Shack. Here at UCSD, the campus storeroom stocks them because the physics and video people all use them heavily. The 3-COM EtherLink book gives the formula for combining the cables of a thin and thick ethernet: (3.28 * T) + E <= 1000 where T is the length in meters of the thin ethernet, E the thick. They then go on to note that this is true only if its soley their equipment on the cable; if you mix other people's transceivers you need to reduce the total length to a maximum of 500 meters (from 1000). Apparently their transceivers are a bit more forgiving than other people's and allow you to violate some spec by a factor of 2. Or something. Brian Kantor UCSD Office of Academic Computing Academic Network Operations Group UCSD B-028, La Jolla, CA 92093 (619) 452-6865 decvax\ brian@sdcsvax.ucsd.edu ihnp4 >--- sdcsvax --- brian ucbvax/ Kantor@Nosc
jay@ethos.UUCP (Jay Denebeim) (02/05/86)
In article <1575@emory.UUCP>, km@emory.UUCP (Ken Mandelberg) writes: > I would appreciate any info (or a pointer) on thin ethernet. > In particular I need to know what limitations there are in > using it compared with regular ethernet. As far as I know, there are no limitations except the maximum length of the lan is much shorter. -- Jay Denebeim "One world, one egg, one basket." {decvax,ihnp4}!mcnc!rti-sel!ethos!jay Deep Thought, ZNode #42 300/1200/2400 919-471-6436
johnl@ima.UUCP (02/09/86)
Our old offices were wired with thin Ethernet. When we moved to new space, we completely prewired it with Information Outlets and abandoned thin Ethernet. What we did was to buy a bunch of DELNIs, and pull transceiver cable from each office to a nearby DELNI, and run one cable from each DELNI to the backbone Ethernet in the ceiling. This gives us regular fat Ethernet everywhere but at a cost of only about $150/station compared to two or three times that if every transceiver cable went to an actual transceiver. The thin ethernet does indeed work electrically just like fat Ethernet and you can connect fat to thin through barrel connectors. The problem with thin cable is that it is very fragile and that it is not so great for prewiring. As someone else suggested, we had little loops coming out of the wall with barrel connectors, and hooked up a station by taking out the barrel connector and putting in a longer loop to the machine, in our case IBM PCs and the like. Thin Ethernet cable is fragile enough that if you step on it, you are likely to damage it enough that it won't work. That combined with the grief of fiddling with the loops of cable drove us back to thick Ethernet with nice sturdy transceiver cables and connectors. I'd go with thick if I could possibly afford it. Thin is fine for patching up stuff around a lab, but it's just too flimsy for permanent wiring in offices. John Levine, ima!johnl
geoff@suneast.uucp (Geoff Arnold) (02/10/86)
What we did was to have a wall plate with two BNC connectors installed in
each office. For those (few) offices whose benighted denizens don't use
Ethernet, we just put in a 6" jumper to maintain the continuity of the
net. Otherwise we run two coax segments to the machine, connecting them to a
T-tap or (via adaptors) to a 3Com screw-in transceiver. We exceeded the 1000'
limitation, and started getting into problems at the hairy edge (our file
server being at one end of the net), so we added a UB repeater. It all
just works, and was far cheaper than putting in the thick stuff. It is also
really convenient if you have several machines in one office (some of us
are just spoilt, I guess).
Geoff
--
#include <sys/disclaimer.h> /* co. lawyers: will this do? */
Geoff Arnold, Sun Microsystems Inc. (East Coast Division)
SnailMail: One Cranberry Hill, Lexington, MA 02173; 617-863-8870 x136
UUCP: [East Coast] linus!security!sunne!suneast!geoff
[West Coast] {hplabs,ihnp4,nsc,pyramid,decwrl}!sun!suneast!geoffhans@erisun.UUCP (02/13/86)
[]
We discovered a while ago that you can actually simulate vampire taps
using a thin ether , T-connectors and standard in-line 3-com Xceivers.
What you do is as follows
Thin ether -> ____________________ ____________________
T-connector -----> T
BNC-N crossover ----> I
__I__
3-Com Xceiver -----> | |
|___|
I
The lower N-connector on the 3-com stays unterminated!
Max permissible length from ether center conductor to 3-com far end
is probably around 10 ".
A screening cover lid might be a good idea for the free end of the
Xceiver.
No extensive tests were performed, but a couple ( 3 ) of stations seemed to
work ok, never jamming the ether while {,dis}connecting stations.
This ( if it really works ) removes one difficulty with thin ether.
--
Two's complement, but three's an int.
Hans Albertsson EIS, USENET/uucp: {decvax,philabs}!mcvax!enea!erix!erisun!hansjack@boring.UUCP (02/15/86)
This has probably been handled before, but I've missed it, I guess, so if anyone could answer the following questions I'd be very grateful: - Is it true that you don't need transcievers when you use thin ethernet? - If you do need transcievers, are they different from normal ones? How are prices? Who makes them? - If you don't need transcievers, how do you connect them to the host? Do you need a different ethernet IF? Mail to me, and I'll summarize. -- Jack Jansen, jack@mcvax.UUCP The shell is my oyster.