On Sun, 16 May 2004 14:05:59 -0700, spammie spam spam wrote:
> Ok here is the situation I'm in. I have a friend about 1/2 mile away and
> for two months I've been trying to share my internet connection with him so
> he doesn't need to use crap MSN anymore. However the problem is that the
> connection on his side drops out quite frequently. I have some ideas as to
> why this is, but I need some input. Below is how the network is set up.
>
> My house: 9dbi omni with 5 feet rj-8
> connected to a DI-614 rev a
> equipment is mounted up about 8 feet from the top of my roof
> so its pretty high up.
This is OK, except for the RG8.......even though it's only 5 feet, you
have quite a loss in that section.
> His house: 24dbi mesh antenna with 1 foot rj8
> connected to DWL-810 wireless-ethernet bridge
> equipment mounted to a tripod on 8 foot mast.
OK at that end.
>
> When the antenna is pointed towards my house, the best signal quality we can
> get is about 30%. That isn't too bad but it drops too frequently to be
> really usable. Occasionally it works, other times there is too much packet
> loss.
Here's where it begins to get tricky....What is 30%? Thirty percent of
WHAT? Do you have diagnostics, (hardware or software) which let you
measure your levels in dbm ? Can you measure the noise values? Are you
then able to display or calculate your total Signal to Noise Ratio?
> Now the di-614 on my side only transmits at 17dbm or about 50mw. If I used
> better equipment for my base station with probably a 200mw atheros pc card,
> would that reduce the random signal drops? Since the di-614 ap is old, I've
> noticed it uses a pc card rather than mini pci, could I just replace this
> card with a 200mw card instead? Or are there any firmware hacks to increase
> the signal strength up to what I need?
What I see as part of your problem is that currently, you simply don't
know whether the issue is primarily an insufficiency of signal, or whether
a major part of the issue is the reception of "noise" which drowns out the
sufficient signal.
The 24dbi simi-parabolic is sufficient on his end, and if noise reception
is a factor for him, the extreme directionality of the antenna is quite a
"plus" for the rejection of noise, unless the noise emmanates from the
direction of your abode.
> The whole point of my little experiment is to do this all on off the shelf
> hardware to reduce costs, so I tcan't really go out and buy an expensive
> prebuilt base station for $600.
>
> Thank You
What I see as the *main* shortcomming in this total setup, is the use of
the "omni" antenna at your end. Omnis are just fine, as long as you have
sufficient driving power. At an average transmit power level of between
20 and 50 milliwatts, you'll NEVER have enough power to decently drive an
omni. Therefore, I submit that your immediate solution is an amplifier at
your end. I would opt for the 1WATT amplifier which is legally purchasable
in US. Something like one of these:
http://www.wirelessinteractive.com/WIPA241W.html
I use one of these exact models and swear by its performance. However, it
is necessary to ensure that you don't have powerful noise sources close to
your home, because this amplifier will pick up and amplify NOISE, as well
as usable signals, and you won't see much performance benefit if it gets
swamped with noise.
When you factor in all other hardware you may purchase, and the time and
effort involved attempting to make your connections work, then the
purchase price of ~$225,00 for the amplifier is quite a bargain.
Attempt to locate and install diagnostic software which will permit you to
determine if noise is affecting your connections. It does in so many
areas which are residential and commercial in nature.
A half mile is a piece of cake......remember that WiFi records of ~ 100
miles have been set, and equipment like the 24dbi parabolic ensure that
reception at the other end is no problem.
Post again if you have further issues!
vg