rosenb@fuhainf.UUCP (Rosenberger) (12/05/89)
While I was reading the discussion about utp on this newsgroup, a few
quesstions arose and I hope you can help me to answer these:
1) does this work on really unshielded wires?
2) is a pair ( 2 ) of wires really sufficient?
3) do you have to stick to one single companie's type of utp equipment or
is there some standart, so you could mix different hardware?
4) what are the limitations of this approach:
-minimum/maximum distance between two stations,
-sensivity of surrounding environment, etc
5) what are your expierences and what hardware can you recommend, I only
know about STARLAN (I think from AT&T), what other companies offer this?
Thanks in advance
hans
rosenb@fernuni-hagen.de
rosenb@dhafeu61.bitnetrnicovic@polyslo.CalPoly.EDU (Ralph Nicovich) (12/06/89)
In article <338@fuhainf2.fuhainf.UUCP> rosenb@fuhainf.UUCP (Rosenberger) writes: >While I was reading the discussion about utp on this newsgroup, a few >quesstions arose and I hope you can help me to answer these: > >1) does this work on really unshielded wires? Yess this works on really unshielded wires. In fact shielded wires limit the distance and do not work as well. >2) is a pair ( 2 ) of wires really sufficient? In my experiance it takes 2 pair. One pair is used for transmit the other for received data. There are 1 pair products on the market but for me at least they have not performed well. Lots of interfearance problems. >3) do you have to stick to one single companie's type of utp equipment or > is there some standart, so you could mix different hardware? At the present time you must use one companies equipment on each subsystem. An over all network might mix several vendors since they are compatible at the 802.3 or Ethernet level (ie tranciever connection or Ethernet attachment) but what feeds the twisted pair from the wire closet must be matched by the same brand on the other end. We are all looking forward to 10BaseT which should solve this, but for now they are not gaurenteed to inter-operate. That is not to say they wont since many of these systems were co-developed (ie ATT and Ungerman Bass) but it is not something to bet on at this point. >4) what are the limitations of this approach: > -minimum/maximum distance between two stations, > -sensivity of surrounding environment, etc Since this is a star wiring plan the minimum distance between stations does not realy apply as it does in "ETHERNET". Most of the products I have looked at have a maximum of 100 meters. I have personaly streched it in one case to almost 700 feet. This is not a recomendation but I had a user that had two offices that far appart and other solutions were cost prohibited. Either this would work or the user would not be able to network. I have had no problems with sensitivity to surrounding environment, including sharing cables with standard telephones. In fact since the signals in this type of system are "ballanced" and not ground refferanced I beleive it is superior to ETHERNET in some cases. Take for instance a ETHERNET between two adjacent buildings. The Ethernet must be grounded at one place. If it is grounded in building 1 and a ground potential diferance is seen between buildings then the sheild is not ground in the second building. Of cource the tranceivers are isolated if installed properly but their are other considerations. >5) what are your expierences and what hardware can you recommend, I only > know about STARLAN (I think from AT&T), what other companies offer this? > My direct experiance is with Ungermann Bass and AT&T, they work well. Others that I have second hand knowledge of but would recommend (and/or use) are Cabletron and Latisenet. I strongly reccomend you stay away from the single pair or balun aproach. In the case of baluns I installed one and it seemed to work well. We then ran another signal thru the wires and it blew it away. > Ralph Nicovich Network Engineering Cal Poly State University San Luis Obispo, Ca
kratz@bnrgate.UUCP (Geoff Kratz) (12/10/89)
In article <338@fuhainf2.fuhainf.UUCP> rosenb@fuhainf.UUCP (Rosenberger) writes: >1) does this work on really unshielded wires? Yes. (Why are people so suprised to hear this? :-) >2) is a pair ( 2 ) of wires really sufficient? Actually, to get 10 Mbps, you need to use 2-pair UTP (standard stuff). The 2-wire systems I heard of only run about 3 Mbps. >3) do you have to stick to one single companie's type of utp equipment or > is there some standart, so you could mix different hardware? The answer is (I believe) "that depends". Since the 10BASET standard still isn't finalized, some vendors have gone off on their own somewhat. Full interoperability isn't guaranteed on all equipment. >4) what are the limitations of this approach: > -minimum/maximum distance between two stations, > -sensivity of surrounding environment, etc UTP doesn't go nearly as far as either of the coax solutions, so if distance is a problem UTP isn't the answer. The distance is something like 100-200 meters (as opposed to the 500m length of thick). >5) what are your expierences and what hardware can you recommend, I only > know about STARLAN (I think from AT&T), what other companies offer this? The STARLAN-10 product is from Hewlett-Packard and is the hub. We are also using HP Bridges. So far, no complaints. -- Geoff Kratz Bell-Northern Research, Ltd. Ph: (613) 763-5784 Internet Systems P.O. Box 3511, Station C FAX:(613) 763-3283 Ottawa Ontario Canada K1Y 4H7 kratz@bnr.ca ...!uunet!bnrgate!kratz #include <disclaimer.h>
kratz@bnrgate.UUCP (Geoff Kratz) (12/10/89)
In article <283@bnrgate.UUCP> kratz@bnrgate.UUCP (Geoff Kratz) writes: >is a problem UTP isn't the answer. The distance is something like 100-200 Ooops. Its 100m. (I shoulda checked first!). -- Geoff Kratz Bell-Northern Research, Ltd. Ph: (613) 763-5784 Internet Systems P.O. Box 3511, Station C FAX:(613) 763-3283 Ottawa Ontario Canada K1Y 4H7 kratz@bnr.ca ...!uunet!bnrgate!kratz #include <disclaimer.h>