I'm baffled by wifi. I see people quoting ranges of over 100 feet or more with wireless access points inside buildings and I can hardly stretch to 12 feet. I've just done some tests using the following: Netgear WG602 v1 Wireless Access Point D-Link AirPlus XtremeG DWL-G520 Wireless PCI card Netgear WPN802 RangeMax Wireless Access Point Netgear WPN311 RangeMax Wireless PCI card With the units only 12 inches apart the results are: Link Quality Signal Strength WG602 + D-Link 70% -29 dbm WPN802 + D-Link 60% -38 dbm WPN802 + WPN311 77% -21 dbm Shouldn't I be getting close to 100% quality at that distance? With the units 12 feet apart (with some obstructions) the results are: Link Quality Signal Strength WG602 + D-Link 34% -61 dbm WG602 + WPN311 36% -61 dbm WPN802 + WPN311 31% -65 dbm Why should there be such a drop in quality/signal at that distance? How do other people get the range that they apparently do? What sort of range/signal do you get? I don't understand those minus figures for signal strength except that -21 is better than -61. What is the strongest signal that one can get using that notation? -- Barry Samuels http://www.beenthere-donethat.org.uk The Unofficial Guide to Great Britain
On Thu, Dec 20, 2007 at 03:59:40PM +0000, Barry Samuels wrote:
Why should there be such a drop in quality/signal at that distance? How do other people get the range that they apparently do? What sort of range/signal do you get?
Do you live on a housing estate? and are there lots of other access points nearby? Dect phones, microwave ovens, wireless video senders, baby monitors, bluetooth etc. etc. will all have the potential to cause interference. Adam -- jabberid = quinophex@jabber.earth.li
On 20/12/07 16:48:13, Adam Bower wrote:
On Thu, Dec 20, 2007 at 03:59:40PM +0000, Barry Samuels wrote:
Why should there be such a drop in quality/signal at that distance? How do other people get the range that they apparently do? What sort of range/signal do you get?
Do you live on a housing estate? and are there lots of other access points nearby? Dect phones, microwave ovens, wireless video senders, baby monitors, bluetooth etc. etc. will all have the potential to cause interference.
Adam
Detached house in country lane. No other access points nearby. Of the others in your list we do have a cordless telephone in the same room as the computer. Would that have a marked effect? -- Barry Samuels http://www.beenthere-donethat.org.uk The Unofficial Guide to Great Britain
On 20 Dec 2007, at 5:10 pm, Barry Samuels wrote:
Detached house in country lane. No other access points nearby. Of the others in your list we do have a cordless telephone in the same room as the computer. Would that have a marked effect?
Unplug it and find out ;) -- David Reynolds david@reynoldsfamily.org.uk
On 20/12/07 17:28:15, David Reynolds wrote:
On 20 Dec 2007, at 5:10 pm, Barry Samuels wrote:
Detached house in country lane. No other access points nearby. Of the others in your list we do have a cordless telephone in the same
room as the computer. Would that have a marked effect?
Unplug it and find out ;)
-- David Reynolds david@reynoldsfamily.org.uk
I did and it doesn't have any effect. -- Barry Samuels http://www.beenthere-donethat.org.uk The Unofficial Guide to Great Britain
Barry Samuels wrote:
On 20/12/07 16:48:13, Adam Bower wrote:
On Thu, Dec 20, 2007 at 03:59:40PM +0000, Barry Samuels wrote:
Why should there be such a drop in quality/signal at that distance? How do other people get the range that they apparently do? What sort of range/signal do you get?
Do you live on a housing estate? and are there lots of other access points nearby? Dect phones, microwave ovens, wireless video senders, baby monitors, bluetooth etc. etc. will all have the potential to cause interference.
Adam
Detached house in country lane. No other access points nearby. Of the others in your list we do have a cordless telephone in the same room as the computer. Would that have a marked effect?
Have you checked the router rf power is set to 100%?/ Ian
On 20/12/07 18:03:45, Ian Thompson-Bell wrote:
Barry Samuels wrote:
Why should there be such a drop in quality/signal at that distance? How do other people get the range that they apparently do? What sort of range/signal do you get?
Do you live on a housing estate? and are there lots of other access points nearby? Dect phones, microwave ovens, wireless video senders, baby monitors, bluetooth etc. etc. will all have the potential to cause interference.
Adam
Detached house in country lane. No other access points nearby. Of the others in your list we do have a cordless telephone in the same room as the computer. Would that have a marked effect?
Have you checked the router rf power is set to 100%?/
Ian
(WAP) Yes and it is. -- Barry Samuels http://www.beenthere-donethat.org.uk The Unofficial Guide to Great Britain
On Thu, Dec 20, 2007 at 05:10:01PM +0000, Barry Samuels wrote:
Detached house in country lane. No other access points nearby. Of the others in your list we do have a cordless telephone in the same room as the computer. Would that have a marked effect?
Ok, what is your house constructed of? as you did mention something about some obstructions, if you live in a nuclear bunker or a building with 10ft thick granite walls that could explain the problem. :) Anyhow, after re-reading your post... you seem concerned at what the numbers say but are you actually having problems with the wireless lan? While i'm sitting here typing this on my network I have a link quality of 78/100 and I can get it to 94/100 if I touch the laptop to the access point but, I don't care what the numbers say as I have a 54 Mb/s connection and it works with little latency or throughput problems. Are you having any other problems or symptoms of a problem on your network? Adam -- jabberid = quinophex@jabber.earth.li
On 20/12/07 21:46:24, Adam Bower wrote:
On Thu, Dec 20, 2007 at 05:10:01PM +0000, Barry Samuels wrote:
Detached house in country lane. No other access points nearby. Of the others in your list we do have a cordless telephone in the same room as the computer. Would that have a marked effect?
Ok, what is your house constructed of? as you did mention something about some obstructions, if you live in a nuclear bunker or a building with 10ft thick granite walls that could explain the problem. :)
Damn! Now you've spoilt my surprise. :)) But seriously though - the house has wooden internal partition walls which include a 3 inch layer of fibreglass insulation.
Anyhow, after re-reading your post... you seem concerned at what the numbers say but are you actually having problems with the wireless lan? While i'm sitting here typing this on my network I have a link quality of 78/100 and I can get it to 94/100 if I touch the laptop to the access point but, I don't care what the numbers say as I have a 54 Mb/s connection and it works with little latency or throughput problems. Are you having any other problems or symptoms of a problem on your network?
Adam
This all started when I decided that I wanted to move my desktop from a room upstairs to a room downstairs but leaving my wife's desktop in the original room upstairs. There is a telephone point in the upstairs room to which the router is connected and there is a wireless access point to accommodate my laptop. At present the two desktops are connected to the router with ethernet cables. There is no telephone point in the room downstairs. Before I move my desk and computer downstairs I want to be sure that it can connect to the network reliably. I did some tests with the laptop in the room downstairs which gave a 55% signal quality with -67 dbm signal strength. I started a file transfer which ran at about 4-8 Mbs for a while then just stopped. Pinging the machine upstairs gave about 6 replies then stopped. I had to assume from those results that the laptop was on the very edge of it's working range. The laptop always gives a much better signal quality and strength than my desktop so I had to assume that the desktop just would work at that range. On Wayne's reccommendation I had a talk with Solwise and I have to admit that they seem to be a very helpful company. If I purchased two HomePlug units (AV200) they volunteered to give me a full refund if they didn't work. So I now have two of those BUT our house has a two phase electrical supply and these units won't work across the phases (I tried) so they have to go on the same phase. This means that the nearest HomePlug unit is across the hall from the upstairs room. I'm going to run a new telephone cable across the attic and down into the room with the HomePlug unit and move the router into that room which means that we have to either use wireless to connect my Wife's machine or try and use an existing telephone cable as a network cable to connect to her machine. I have no means of testing the wireless connection between the rooms upstairs as I don't have a long enough ethernet cable to try it. I can put the WAP into that room and scan for a signal which is what those figures were all about. I hope that all makes sense. -- Barry Samuels http://www.beenthere-donethat.org.uk The Unofficial Guide to Great Britain
On 21-Dec-07 10:49:08, Barry Samuels wrote:
[...] I'm going to run a new telephone cable across the attic and down into the room with the HomePlug unit and move the router into that room which means that we have to either use wireless to connect my Wife's machine or try and use an existing telephone cable as a network cable to connect to her machine.
I have no means of testing the wireless connection between the rooms upstairs as I don't have a long enough ethernet cable to try it. I can put the WAP into that room and scan for a signal which is what those figures were all about.
I hope that all makes sense.
-- Barry Samuels
Barry, if you're going to run a cable anyway, might you not be able to put in a proper ethernet cable? I don't know what the longest is that you can readily buy "off the shelf", but probably you can find a long enough one. I think I've seen 20m cables on sale and maybe longer (I happily have a few 15m ones radiating round the house). And, if need be, you can buy (cheap) little connectors which enable you to join one length of CAT cable to another. Whikle there's some possible degradation at the junction, this is very unliely to be critical. Proper CAT cables have nuch better shielding, and the twisted-pair design gives them better protection from interference. For a telephone cable of any length, you might find that the througput is degraded. Just a thought ... On the theme of Wifi range: Though I don't use wireless myself (it's all cable), my laptop has a Wifi device in it. My next-door neighbour has a BT Home Hub unit, which is situated downstairs near to our dividing wall. My laptop is upstairs (about 8ft higher) again near to our dividing wall. Max distance between them: no more than 10ft. But if I switch on my laptop's Wifi and search for a Wifi device, I pick up his Home Hub with (usually) a "weak" or "poor" signal. The only adverse factor I can think of is that, given the vertical separation and the short horizontal separation, the "line of sight" (as it were) between the two would be almost parallel to the wall, and therefore passing through a relatively long stretch of brick. On the other hand, my friend up the road lives in a converted old church, with very thick internal walls. His Wifi hub is upstairs and to one side, and he can move his Mac laptop all over the place downstairs and still get a good signal. A much tougher proposition, I would have thought, than between me and my next-door neighbour. Maybe these things are sensitive to factors which are not too obvious ... Good luck! Ted. -------------------------------------------------------------------- E-Mail: (Ted Harding) <Ted.Harding@manchester.ac.uk> Fax-to-email: +44 (0)870 094 0861 Date: 21-Dec-07 Time: 12:29:42 ------------------------------ XFMail ------------------------------
On Fri, Dec 21, 2007 at 12:29:53PM -0000, Ted Harding wrote:
On 21-Dec-07 10:49:08, Barry Samuels wrote:
[...] I'm going to run a new telephone cable across the attic and down into the room with the HomePlug unit and move the router into that room which means that we have to either use wireless to connect my Wife's machine or try and use an existing telephone cable as a network cable to connect to her machine.
I have no means of testing the wireless connection between the rooms upstairs as I don't have a long enough ethernet cable to try it. I can put the WAP into that room and scan for a signal which is what those figures were all about.
I hope that all makes sense.
-- Barry Samuels
Barry, if you're going to run a cable anyway, might you not be able to put in a proper ethernet cable?
I don't know what the longest is that you can readily buy "off the shelf", but probably you can find a long enough one. I think I've seen 20m cables on sale and maybe longer
You can definitely buy 30 metre ones and probably longer. Or you can get female<-->female RJ45 connectors -- Chris Green
On Fri, 2007-12-21 at 12:49 +0000, Chris G wrote:
You can definitely buy 30 metre ones and probably longer.
Or you can get female<-->female RJ45 connectors
CAT5e Cable is currently at 15p a metre, the crimp on ends are nothing and the crimp tool is available for the price of a McMeal if you only want one good enough for occasional use. Then you can enjoy network cables made to exactly the right length for the rest of your life. Also every long ready made cable I have seen in PC World etc is not only completely insanely overpriced (in that you could buy enough of the things mentioned above to make 10 of them, including the tools) but also multistrand cable which is not to spec for cable lengths > 10 meters. But anyway I think this is a moot point as Barry was trying to use an existing cable already installed in the house. Oh and the female<>female RJ45 connectors are evil and should only be used when there is no alternative.
On Fri, Dec 21, 2007 at 01:20:59PM +0000, Wayne Stallwood wrote:
CAT5e Cable is currently at 15p a metre, the crimp on ends are nothing and the crimp tool is available for the price of a McMeal if you only want one good enough for occasional use. Then you can enjoy network cables made to exactly the right length for the rest of your life.
Also every long ready made cable I have seen in PC World etc is not only completely insanely overpriced (in that you could buy enough of the things mentioned above to make 10 of them, including the tools) but also multistrand cable which is not to spec for cable lengths > 10 meters.
Don't buy cables from PC World then. These days premade cables can be had for amounts of money that make it a waste of time to make them by hand unless you need a particularly long run. They tend to be much better quality too. J. -- Web [ noodles is almost too good to be true ] site: http:// [ ] Made by www.earth.li/~noodles/ [ ] HuggieTag 0.0.23
On Fri, Dec 21, 2007 at 01:20:59PM +0000, Wayne Stallwood wrote:
On Fri, 2007-12-21 at 12:49 +0000, Chris G wrote:
You can definitely buy 30 metre ones and probably longer.
Or you can get female<-->female RJ45 connectors
CAT5e Cable is currently at 15p a metre, the crimp on ends are nothing and the crimp tool is available for the price of a McMeal if you only want one good enough for occasional use. Then you can enjoy network cables made to exactly the right length for the rest of your life.
Also every long ready made cable I have seen in PC World etc is not only completely insanely overpriced (in that you could buy enough of the things mentioned above to make 10 of them, including the tools) but also multistrand cable which is not to spec for cable lengths > 10 meters.
Well don't buy from PC World then! :-) If you do a search on Google Products for "30m cat5e" then you get loads of hits and prices start at less than £5.
Oh and the female<>female RJ45 connectors are evil and should only be used when there is no alternative.
Yes, but sometimes there is no alternative. -- Chris Green
Barry, if you're going to run a cable anyway, might you not be able to put in a proper ethernet cable?
I need the new telephone connection to get ADSL to the router's new position. However I managed to get one of those 'evil' ethernet couplers and, by joining two cables, could then put the WAP in it's new position and connect to the router. From my Wife's machine it was getting a signal of around 42% and a strength of -59 dbm and she said that internet sites were displaying as fast as she would normally expect so it seems that it should be OK. So 'all' I need to do now is to grovel around the attic to lay the new telephone cable, move my stuff downstairs and connect my machine via the Solwise HomePlug devices and all will be hunky dory. I was able, in a test using my laptop, to get around 30 Mbs over the HomePlug link. Thanks for everyone's help. I have probably learned something from this experience but the problem is that I won't be able to remember it when I need it again. :-(( -- Barry Samuels http://www.beenthere-donethat.org.uk The Unofficial Guide to Great Britain
On Fri, 2007-12-21 at 13:55 +0000, Chris G wrote:
Well don't buy from PC World then! :-)
If you do a search on Google Products for "30m cat5e" then you get loads of hits and prices start at less than £5.
True but it doesn't change the fact that most of those premade cables I have seen are using multistrand cable which is wrong for that length. Also it would mean that I have to keep a stock of 2,3,6,8,10 and 15m etc where as one box of cable and 5 minutes crimping the ends will do any length I want. Sure I carry some common lengths because the cable from the wall socket to the computer should be multistrand (for under 10m) as it is more durable and making up a cable every time would be daft. But for one off longer lengths that are going into a reasonably static environment I make them every time.
Oh and the female<>female RJ45 connectors are evil and should only be used when there is no alternative.
Yes, but sometimes there is no alternative.
Unless you are extending an already installed cable because it would be a pain to replace it then there is always an alternative. Like making a cable the right length :)
On Fri, Dec 21, 2007 at 04:40:27PM +0000, Wayne Stallwood wrote:
Also it would mean that I have to keep a stock of 2,3,6,8,10 and 15m etc where as one box of cable and 5 minutes crimping the ends will do any length I want. Sure I carry some common lengths because the cable from the wall socket to the computer should be multistrand (for under 10m) as it is more durable and making up a cable every time would be daft. But for one off longer lengths that are going into a reasonably static environment I make them every time.
I bet you only do that because you bill twice, once for your time making the cable and then again for the cable itself at some huge pc world style markup ;) Adam -- jabberid = quinophex@jabber.earth.li
On Fri, 2007-12-21 at 12:29 +0000, Ted.Harding@manchester.ac.uk wrote:
Barry, if you're going to run a cable anyway, might you not be able to put in a proper ethernet cable?
I think the point was that the telephone cable was already installed between the two rooms he wants to network.
Proper CAT cables have nuch better shielding, and the twisted-pair design gives them better protection from interference. For a telephone cable of any length, you might find that the througput is degraded.
Yes the twist of the pairs is completely essential for ethernet. a straight wired cable or a cable where tx or rx has been wired across pairs will generally not work very well for distances of over 4 meters Standard installation grade CAT5 has no shielding so the twist (combined with differential signalling) is the only way the cable can deal with noise or crosstalk. That said if the wiring in Barry's house is reasonably new then it should have been done with CAT3 or above which as Keith points out is also twisted pair. Old house telephone wiring wasn't twisted pair however and ethernet won't run over this for any distance that you would care about.
On Fri, Dec 21, 2007 at 10:49:08AM +0000, Barry Samuels wrote:
Are you having any other problems or symptoms of a problem on your network?
Adam
This all started when I decided that I wanted to move my desktop from a room upstairs to a room downstairs but leaving my wife's desktop in the original room upstairs.
There is a telephone point in the upstairs room to which the router is connected and there is a wireless access point to accommodate my laptop. At present the two desktops are connected to the router with ethernet cables.
There is no telephone point in the room downstairs. Before I move my desk and computer downstairs I want to be sure that it can connect to the network reliably.
I have everything hard-wired even in our large rambling house. It just takes a little imagination and exploration to route the wires. The master socket (with ADSL splitter faceplate) is in my study so the router and switch are there. My computer and two printers on the network are there too. So far so good, the cables to the printers and the rest of the house have to run under a doorway so I made two small holes in the chipboard floor and 'fished' with a bent coat hanger to pull wires through. Maxine's computer is in the adjacent study (large and rambling remember, two studies) and I have made a hole in the wall for connections to that and also to upstairs. That hole is about the only structural thing I've done and it's under some shelves anyway. Then from Maxine's study there's Cat5 running up a service duct (drains etc.) to the first floor and then under floorboards and in roof void right across the house to my son's room where there was another computer but now there's just some interconnections and thence two wires out of the window to the garage where I have another computer and a NAS drive. Only one is actually necessary now as I have an old hub in the garage for the two things there. Wirelss - just say no! :-) -- Chris Green
Barry The -xx dBm is a relative power level. If you transmit at the maximum 100mW then it that is 0 dBm. I would think that -30dBm is a standard level for being near to the WAP firing out at full power. If the AP is just giving out 10mW then that already -10dBm. Its decibel do -3 is half and -10 is one tenth of the power. You get good reception at -80dBm which is the normal power where most clients try to roam. As other have said the cordless phone could be doing the problem, did you switch the base and all handsets off ? Keith ________________________________ From: main-bounces@lists.alug.org.uk on behalf of Barry Samuels Sent: Thu 20/12/2007 3:59 PM To: Alug Subject: [ALUG] Wifi again I'm baffled by wifi. I see people quoting ranges of over 100 feet or more with wireless access points inside buildings and I can hardly stretch to 12 feet. I've just done some tests using the following: Netgear WG602 v1 Wireless Access Point D-Link AirPlus XtremeG DWL-G520 Wireless PCI card Netgear WPN802 RangeMax Wireless Access Point Netgear WPN311 RangeMax Wireless PCI card With the units only 12 inches apart the results are: Link Quality Signal Strength WG602 + D-Link 70% -29 dbm WPN802 + D-Link 60% -38 dbm WPN802 + WPN311 77% -21 dbm Shouldn't I be getting close to 100% quality at that distance? With the units 12 feet apart (with some obstructions) the results are: Link Quality Signal Strength WG602 + D-Link 34% -61 dbm WG602 + WPN311 36% -61 dbm WPN802 + WPN311 31% -65 dbm Why should there be such a drop in quality/signal at that distance? How do other people get the range that they apparently do? What sort of range/signal do you get? I don't understand those minus figures for signal strength except that -21 is better than -61. What is the strongest signal that one can get using that notation?
participants (9)
-
Adam Bower -
Barry Samuels -
Chris G -
David Reynolds -
Ian Thompson-Bell -
Jonathan McDowell -
keith.jamieson＠bt.com -
Ted.Harding＠manchester.ac.uk -
Wayne Stallwood