Poor connection due to walls?

I have an old house with thick walls.(4) Is there anyway to solve the problem short of hard wiring? I have tried two systems and neither work very well.

Reply to
W. Wells
Loading thread data ...

If you get at least some (poor) signal, then a better antenna on the wireless host (router or access point) may do the trick.

Reply to
John Navas

First move should be some free reflectors. If you have some signal now, these might be all you need.

formatting link
EZ-12, Windsurfer reflector. printed on photo paper for thick stock, with aluminum foil glued to the sail, provides a substantial boost in signal.
formatting link

Reply to
dold

You must be new to the group - the same subject is covered almost weekly. Have a read of some of the posts.

Reply to
Rob

Reply to
W. Wells

Depends on your walls (concrete, metal lath?), but multiple APs or hard wiring is proabably going to give you the best result.

Reply to
William P.N. Smith

"W. Wells" hath wroth:

Soft wiring?

Power line networking. Wireless bridge on either side of the wall(s). See HomePlug.

Phone line networking. Use the phone lines to connect between wireless access points. See HomePNA.

Ethernet over CATV coax cable. URL when I untrash my bookmarks file.

Reply to
Jeff Liebermann

Basically as others will tell you, 2.4 GHz signals don't pass through walls. What happens is the signal bounces around and or finds a 'hole' in the wall. Thicker the wall the less likely of finding enough of a hole to make the trip. Consider one of the devices that allows networking on your power lines already wire in the house or the phone wiring.

fundamentalism, fundamentally wrong.

Reply to
Rico

rico snipped-for-privacy@hotmail.com (Rico) hath wroth:

Generally true for most common construction techniques. Wireless really does bounce around and find a suitable hole, usually a doorway. However, it can only bounce a limited number of times before multipath and path length extension become a problem.

Drivel: One of my fun projects many years ago was trying to figure out how many walls and floors could be penetrated by 2.4Ghz Wi-Fi in an office building. One wall was easy. Two walls were a bit of a challenge. Three walls resulted in a rather weak and unusable signal. But as we added more and more walls and floors, the signal continued to be present at almost the same level. We eventually found about 8 walls had almost exactly the same weak and unusable signal level as 3 walls. It didn't just fade away or disappear as one would expect. What was happening was the signal was going out the window, was reflected off a nearby building, and re-entered via another window.

Different signals have different attenuations. RF will go through most common materials with varying attenuations. As a rule of thumb,

6dB loss is 1/2 of the range. 12dB loss is 1/4th the range. This means if you put a plywood wall between the AP and the client, a 6dB loss through the plywood will results in 1/2 the range that would be obtained without the plywood obstruction. My guess(tm) is anything over about 12dB attenuation is going to be unreliable or flaky.

I did a survey of various web sites and books claiming to have measured or calculated the attenuation of various materials. See: |

formatting link
the wide variation in opinions and measurements. I don't agree with all the numbers.

Dumb story: Descriptions also vary. One client asked me to improve their office indoor coverage. They described the office as having very light weight partitions, with very little metal in them. I expected to find a radio or interference problem. Instead, I found row after row of metal file cabinets and blueprint storage cabinets propped up against both sides of the office partitions. I apparently forgot to ask about office furniture. I gave up immediately and just two more access points on opposite sides of the offending partitions.

Reply to
Jeff Liebermann

Have you tried moving your modem/accesspoint to another possibly more central location? All of your computers can connect wirelessly? If your connection is ADSL, a phone need not be plugged in. If cable modem, you may need a splitter and make sure the 'high output' tap of your outside splitter drives the modem/accesspoint.

My neighbor gained connectivity thruout the house and yard by putting his modem/accesspoint in the attic :)

lee

Reply to
lee_houston

What happens if you put aluminum foil or some such reflector on the wall(s) sort of aimed at the "door"? Would that offer any improvement or just wasted effort?

fundamentalism, fundamentally wrong.

Reply to
Rico

Aimed? How do I aim a wall?

I have no idea how this might work, or what could be done to insure that it does something useful. Waste of time methinks.

Reply to
Jeff Liebermann

Well I was thinking in terms of the foil might not be shear against the wall in all places. Have you never bounced a laser (pointer) off a mirror when playing red dot with a cat?

Makes sense, just a thought since some em waves do reflect off metal surfaces better then say dry wall or plaster.

fundamentalism, fundamentally wrong.

Reply to
Rico

Reply to
W. Wells

Why aren't there any devices that use the COAX? If cable companies run Internet over it, I would think it could be used as a HOME LAN backbone...

Reply to
SD

SD hath wroth:

Ever hear of 10base2, also known as CheaperNet? It's 10 mbit/sec ethernet over coax. You take two 10baseT hubs, that have both a BNC connector and the usual RJ45 connectors and you have a "media converter". I think they go for about $10 ea on eBay. Install a BNC "T" connector and 50 ohm loads at each end. Run RG-58a/u between hubs and you have an ethernet connection. You can snake the coax as a bus through the walls and around the house, install additional taps where necessary, and have a system that was obsolete in about 1995.

There are plenty reason why few people do it this way. I won't go into all of them. The big ones are topology and speed. 10base2 coax is wired as a bus. You really want a "star" topology, where all the various wires around the house go to one central location. If you've ever tried to troubleshoot a problem on a bus type network, you'll understand why. It's NOT easy. In addition, each tap requires that both the coax cables from the wall go to the workstation. That means two BNC connectors sticking out of the wall, with two coax cables going to each workstation. Not very elegant.

The speed problem is that it's limited to 10Mbits/sec. 10base2 will not run at 100Mbits/sec.

I've used such coax arrangements where nothing else is available. For example, if there are only two machines on the endpoints of the system, I can get away with using CATV coax, RG-6/u instead of RG-58a/u. I've also used it where RF shielding is important and for long distances (about 1000ft). It works, but it's not fast or elegant. The nice part about using CATV coax is that most existing installations are a star topology.

Also see:

formatting link
formatting link
ethernet over CATV systems.

There are lots of other possibilities for using coax cable. Too many to itemize. Got any specific problems you want to solve?

Incidentally, Comcast cable, in its previous manifestation as TCI, was considering using the home CATV wiring for distributing 2.4GHz Wi-Fi signals. It's fairly easy to do with a simple diplexer. The high loss is not as bad as trying to go through impossible walls and floors. Figure on 9dB/100ft loss at 2.4GHz, which is less than the loss going through a brick wall. However, I found that most home coax cables and connectors are truely disgusting at 2.4GHz. I once demonstrated a prototype system using multiple 100ft runs of RG-6/u. It worked fairly well, but required replacing all of the junk coax with "satellite grade" coax and using push-on F connectors.

Reply to
Jeff Liebermann

Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.