About CAT-5 for 2 phone line and ethernet

Cat3 is good enough for this service, and it was designed specifically for this shared-sheath service. Nowadays, people prefer to run separate cables. 50m is well within spec (100m), but you seriously risk lightening damage if it is run between buildings.

Some phones installed can draw high current during ringing, which _might_ interfere with ethernet. I would be suspicious of old mechanical-contact bells/buzzers. But even these should only ruin a few frames, and ethernet has to be designed for collision-damaged frames.

Yes. Make _sure_ you do not split any pairs (wire correctly). Twisting is done to balance interference and reduce net noise. Paired signals must travel on paired conductors.

-- Robert

You can certianly do that, although it would not be a good idea to base a

100-outlet network on this kind of design. A cable that's split like that between three jacks is not going to support Gigabit Ethernet, which you may want going forward in the future. Make sure you leave pairs #2 and #2 (orange and green) connected to the data jack,first line on the blue pair and second line on the brown one. The rings are not going to create any problem as the fre

oops, sorry I've been typing too quick and hit the send button. So, below is the full message:

You can certainly do that, although it would not be a good idea to base a

100-outlet network on this kind of design. A cable that's split like that between three jacks is not going to support Gigabit Ethernet, which you may want going forward in the future. Make sure you leave pairs #2 and #2 (orange and green) connected to the data jack, first line on the blue pair and second line on the brown one.

The rings are not going to create any problem as the frequency of the EMI that occurs is radically different from that of Ethernet. The Ethernet devices will easily filter that EMI out, irrespective of the actual amount of energy coupled between the pairs. It would be similar to an experience of an ocean-going ship crew that meets a tsunami wave in an open ocean: it's so big (long) that you don't even know you rode the wave, even though it is big enough to wipe out a city when it hits the land.

Hello, Thank you for your information.

Has anyone done an analysis on the bit error or frame error on ethernet data because of the ringing voltage of the telephone line? ..If so please let me know the information.

This is used for commerical use. So i need more information about the interference.

Waiting for reply, Thanks and regards Praveen

If it's for commercial use, follow the standards and pull another wire. More wire is always cheaper than non-standard installations in commercial applications.

Does it explicitly disallow sharing or does it simply require that the premise wiring be jack to jack with all conductors of a specific cable terminated on a single jack and any sharing done using splitters that are plugged into the jacks and thus outside the scope of the standard?

If by saying "for commercial use" you actually mean that you are worried about liabilities, then you should just abandon the idea of splitting the cable and go by the cabling standard for commercial buildings - TIA/EIA-568-B that specifically disallow cable sharing.

Do not carry POTS and Manchester coded data together although it does work in the lab here. I tried this just for fun and I did get an effect on our 75meter run using 10baseT and just the one POTS 90Vrms 10Hz ACD ringdown and it worked OK. But I know there are risks on induced voltages and NIC's would fry for sure.

Again, I did this just for curiosity but 568 specifically mentions separation of services. NICs are just not FCC part 15/68 DOC CS03 and they are definately not ready for those voltages. POTS lines may come from outside the dwelling where all sorts of high energy discharges exist including Lightning. Yikes. If something happens, damages will occur and there is a risk to Human contact where future maintenance crews were not expecting the hazard. Up here in Canada, that would be considered as a "man-trap", and can result in manslaughter charges against designers or installers of the hazard should a death occur.

That said, I have installed a Nortel PBX digital set line right into pair 4/5, knowing its 18vdc and 80Kb/s

3Vp-p transmission, and *not* a POTS line.

John

Twisted pair ethernet, was originally designed to use existing phone cabling, which means it was built to tolerate the signals found in a phone system. Also consider the frequencies involved. An analog phone line operates at frequencies below 4 KHz, with ringing usually at 20 Hz. In comparison, 10baseT uses signalling at between 10 and 20 MHz, which is at least 2.5 thousand times the frequencies used in the phone system. The coupling transformers in the network cards are designed to efficiently pass the ethernet signal, but not the phone signals. This means that interference is unlikely.

If 20 Hz ringing is causing problems for data carried at frequencies over 10 MHz, there's something seriously wrong with the NIC. Also, the main reason for twisted pair, is to prevent interference, both between pairs and with external devices.

While 90V ringing would certainly cause problems, if connected directly to the NIC, the transformers that connect the line are supposed to handle a few hundred volts, without creating a safety problem (that doesn't mean no equipment failures).

Twisted pair NICs originally were originally designed to share phone cables. While it's not considered good practice to combine the two on one cable, there are occasions, when it's necessary.

Also, phone cables coming into a building, must have protection against surgers. That said however, a close enough lightning hit, is going to cause problems, no matter what. Incidentally, many computers are connected to the phone line through dial up and DSL modems.