dfc@macs.UUCP (David Caldwell) (03/07/91)
We are installing Ethernet into a new building and we want to install 10-base-T cable. We also want to use the same cables for async serial communications for offices that are not equiped with Ethernet machines but will be equiped as such in the future. We have been told by Cabletron Industries that we need a cable that conforms to the 258-A standard. This cable is an 8 wire flat cable designed for crimp on connections to RJ45 plugs. We have no idea what this 258-A standard is or from whence it comes nor can we find any specifications for cables which conform to this standard from any of the standard cable manufacturers like Belden. BTW, this cable is supposed to have wires 1&2 twisted and 3&6 twisted. What we would like to know is if anyone else has run 10-base-T and asynch over the same cable at the same time and how was it done? Send replies to mdg@macs or griffith@stetson.BITNET Thanks, David Caldwell Stetson University
shepperd@dms.UUCP (Dave Shepperd) (03/13/91)
dfc@macs.UUCP (David Caldwell) writes: > What we would like to know is if anyone else has run 10-base-T and > asynch over the same cable at the same time and how was it done? We do it, however, it isn't clear to me what you are asking. Do you mean do we mix async and 10-base-T in the same feeder to the wall outlet? Yes, we do, along with digital and analog telephone data as well as other analog and digital signals. The only problems have been with people using modems on their phone line at the same time as they were using their async terminal (the modem would sometimes loose carrier). Also one could hear a little noise on their analog phone if their terminal was sending or receiving a lot of data AND the line cord to the phone and/or terminal was rather long. Are you asking how we get the async and 10-base-t from the wall to the workstation/terminal/whatever? Our wall jacks are duplex RJ-11's (3 pair outlets, giving us a total of 6 pair to each location). We use the center pair of each RJ-11 jack for telephone and the outer 2 pairs for 10-base-t or async (one on each outlet). This allows someone to plug a terminal into a phone outlet or a phone into a terminal outlet with no ill effects to the underlying equipment (although the terminal and/or phone obviously won't work). A single 6 pair cable runs from each wall jack location to a phone closet. Each phone closet is cross connected to a multi-hundred pair feeder which connects it to the main computer/telephone room. I had great difficulty running two or more long 10-base-t circuits (each about 150 feet) in the same feeder. One or the other would work just fine but when both were hooked up, neither would work. So I ran a coax from the main room to a mini-repeater in a remote phone closet and 10-base-t from there to the wall jack locations and that has been working just fine (this is the recommended setup, but I tried to cheap out; it won't work to be too cheap). If you plan on using Cabletron's 10-base-t stuff, be advised that their LED indicators are neat, but misleading. If the LED is off, then the circuit is no good. If the LED is on, then the circuit MIGHT be good. You really need a TDR or other signal quality meter to verify the (hardware) circuit before you can believe the LEDs. We make the cords that connect the 10-base-T from the wall to the workstation using RJ-45 and RJ-11 plugs on the appropriate ends using a standard two twisted pair IW cable (exactly the same as the cable from the phone closet to the wall jack, except 2 pair instead of 6). Although this works, I can't recommend it. This cable is solid conductor and the RJ-11 and RJ-45 plugs really want (and need) to be crimped to stranded wire (the kind in "normal" flat telephone line cord). I haven't found any of that flat stranded wire line cord with TWISTED PAIRS anywhere, yet. I'm going to get some if I can find it. Using solid wire is problematic, since it is difficult to handle while making the cable, it doesn't allow for much flexibility and it'll easily break (or at least the connections at the RJ's will) if moved much at all. In our environment, the cable is fastened to the wall and terminates at a 10-base-t transceiver which is also fastened to the wall so the wire doesn't get moved. If I had workstations that required a 10-base-t connection directly, I'd be in trouble. When someone needs to connect either 2 async devices or an async and a 10-base-t device as well as a phone and/or modem to a location, we hand out some 3 pair "T" adapters. These gizmos plug into an RJ-11 jack and offer 2 parallel RJ-11 sockets so one can plug both a phone and terminal into the same outlet. However, when doing this for 10-base-t circuits, we use great care. In this instance, we have to use a 1 pair telephone line cord on the non-10base-t side of the "T". If other than a 1 pair line cord were used, the transmission line characteristics of the 10-base-t circuit would get goofed up and therefore the circuit would stop working or become flakey (although the Cabletron's LED's would say everything is fine). For async devices, no such restriction applies. The cords to the async terminals are the standard 3 pair flat telephone line cords and terminate in an RJ-11 to DB25 adapter which is fastened to the async device. We construct these cords too, although they can be purchased just about anywhere for an outrageous sum. We've been using async on telephone lines now for about 10 years. It works like a champ. The longest run is near 3000 feet. The long runs have trouble at 9600 baud (RS232), but work just dandy at 4800. I suspect if you were using RS422/RS423, you could run 3000 feet at 9600 with no difficulty. We've been using 10-base-t (all Cabletron stuff) for only a few months and I'd say the jury's still out on it. -- Dave Shepperd. shepperd@dms.UUCP or motcsd!dms!shepperd Atari Games Corporation, 675 Sycamore Drive, Milpitas CA 95035. Nobody knows what I'm saying. I don't even know what I'm saying.