Can I somehow use POTS wire to network a PC to an existing network?

Can I use POTS (Plain Old Telephone Service) wire to network a PC to an existing network? Here's the situation:

I have a small network running Win/XP on several PC's in my office about 200 feet from my home. It's too far for a wireless network connection I think. There is an underground 25 pair telephone cable running from my office to my home. Can I somehow use that cable to add my home PC (Win/ME) to my office network? Obviously, the telephone cable is NOT Cat 5E quality but it works well enough for telephone service. I think that HomePNA is NOT what I want because that sets up another network that then requires a "network bridge." Is that correct? I'm at the limit of my knowledge here. What I think I need is some sort of (hopefully cheap) converter at each end of the telephone cable. Is that possible? Thanks for any suggestions.

 

i think you might have got some bad info on homepna. It will do exactly whatchu want.

Lots of other vendors out there, but for example, check out the marketing whizz on this starter kit.
http://www.linksys.com/products/product.asp?grid=34&scid=33&prid=131

The sticky part of your question comes when you say 'add it to my existing office network'. Do you have an NT server? If so, you could use it as a software bridge. Otherwise, you will need a bridge or something. This isnt a homepna thing though. Whenever you take 'network type a' and try to connect it to 'network type b', it takes something to 'bridge' them together. Would need more info on the other network, and topology/equipment to help you decide.

Another, perhaps easier alternitive, is get a directional wireless ethernet setup. the antennas arent particularly cheap, but probably about the same cost as a hardware bridge.

 

Thanks Joe, for your comments. My "existing network" is just a plain old peer to peer network using Windows XP/Pro. I do NOT have Windows NT, although does XP include the same functionality?? Anyway, my goal is to simply add my home PC to the existing Win/XP network.

I think I understand that HomePNA can create a PC network, but I also understand that the HomePNA network is DIFFERENT from a Win/XP network and the PC's in the two networks can't see each other (without a "bridge" or something like that). Is that correct?

A directional wireless ethernet setup is possible, but would require antennas, as you said, as well as outdoor mounting, and cabling from the PC's through the walls to the antennas. All in all, the physical arrangements seem simpler if the existing telephone cable can somehow be used.

 

Ok, so you have two options. Get a hardware bridge, or do some serious tcpip configuration on the xp machines

This would be the configuration i recommend for someone without an NT server (since you have lots of things, like name resolution, etc to deal with).
http://www.practicallynetworked.com/networking/alternative_net_ex1.htm
You'll need one of these: http://www.homepna.com/HPNA-LinksysBridgeReview.html-ssi


The other alternitive would be to multihome one of the xp machines in the office and setup for routing. This is painless, but can be tricky and tedious without some knowladge of tcpip.

 

Thanks, Joe, I think I understand the Ethernet-HomePNA bridge. Also, I note that HomePNA can support daisy chaining PC's which is great, and I didn't know that. But what did you mean when you said:

"The other alternitive would be to multihome one of the xp machines in the office and setup for routing." I don't know what "multihome" means.

Thanks again.

 

if you were a tcpip guy, you would simply put an ethernet card AND a pna card in an always on XP machine. Then, with some registry fiddling, you could turn the XP box into a router. A router is a generic term for a networking device that routes traffic across two different subnets. This can be a little tricky, since you would need to have some sort of name resolution (normally done with a WINS or DNS server via NT). You can do it old school with a txt file called a HOSTS file, though, and you'd have to do some route adds to make sure your machines could get to both the internet and the homelan.

hehe. thats why i recommended the solid state bridge. The above is certainly achievable, but for 80$ or so for a bridge, probably not worth the configuration time.

 

thanks again Joe. You're probably right that a solid state bridge is the right solution and it's not worth my time to do it myself. On the other hand, I DO like a challenge, and I actually have lots of time, so I might at least try to figure it out. SOOO, is there a website that explains all this (subnet, HOSTS file, etc.), and starts pretty much from scratch? If so, I will study it and see what I can learn.

 

Thanks Tony for your response. Besides the 200 feet distance, there is the additional problem (which I think I forgot to mention) that BOTH of the buildings involved have aluminum or steel siding, and I assume that would pretty well eliminate a wireless connection unless I wanted to try to use outside antennas, which I really don't want to do.

I don't know anything about wireless boosters so I would be glad to know of any information sources for those devices.

The buried cable is NOT in a conduit, so I will have use it as is, I assume.

 

Be careful connecting computers in two different buildings via copper wire. There are safety implications.

• If the buildings have different impedances for their zero (earth) levels (which can occur over very short distances) you can have a potential current over the earth line where current passes from the building with the high zero to the one with the low zero. The effect is a DC voltage running over your "network" cable. In addition, should a circuit fail on the high zero side and this results in the circuit passing current to "earth ", it is possible that the easiest route is over the network cable to earth at the low zero building - that is significant current running over the network cable.

• There is also a risk from lightening. There is the possiblity that should one building be hit by lightening, the current could run down the network cable and pass to the second building

While I would concede that both situations have a low risk of occuring, were either to occur the outcome could be serious and result in fire. For this reason connecting buildings via copper is considered bad practice.

I would recommed:

1. Wireless using external antenaes if you have to, and have line of sight otherwise.

2. Run fibre between the two buildings - this is the most robust solution, but also the most expensive. However, I'd recommend you get a cabler to quote for the job - it might not be as expensive as you may think and it is the best solution if the budget allows.

3. If you really have to go copper, use a system that will provide a breaker so that unwanted current does not flow from one building to the other.

 

Thanks to Reggie and Tony for comments about the grounding issue. I assume that the grounding issue becomes relevant because of the low current used in network cables, is that correct? I ask that because the 25 pair cable clearly carries conventional telephone traffic and the grounding problem has not occurred that I know of. Or is the grounding problem also potentially troublesome even with telephone traffic but just hasn't occured yet?

 

I wouldn't recommend that. If you want to go the physical connection route, use fibre.

I've seen many instances of people running copper network lines between buildings (inclubing gaffer taped CAT5 cables strung between windows). It is done very often. However, I believe it is bad practice and would be doing Roger a disservice by not stating that. He should take my comments along with other advice and come to a decision based on all the information. It is advice and should be taken as such.

I am not an electrician and were I to go into more detail would make too many technical mistakes which would probably detract from my argument. Therefore if you wish to understand the technical reasons why connect two computers together in two different builds via copper is not a good idea, I would refer you to Practical Network Cabling by Frank Derfler. Mr Derfler describes the issues much more succinctly than I can.