Long Range wifi troubles

I'm trying to set up a long range wifi connection between 2 sites, with
decent line of sight(a few trees but I can see each antenna). Covering
about 1100 feet. I have 2 EtherDesigns Cantennas, a 10" on the
recieving end(a Linksys PCI card WMP54G) and a 5" on the access point,
a Linksys WRT54G. According to the linksys software on the recieving
end, I have a 51% signal, 3 out of 5 bars, and -77 dB with -79 dB of
noise. It says the computer is transmitting at 18mb but recieving at
only 1 megabit. Linksys reports that data in bytes are sent and
recieved but does not recieve packets.

Windows sees the access point and connects to it, but shows packets
recieved as 0 and won't get an IP address.

Is the problem simply not a good enough connection? If so, why does the
linksys show 3/5 signal and transmit fine? Or do I have another problem
going on where I can't get an IP address?

thanks
Jamie

(E-Mail Removed) wrote:
> I'm trying to set up a long range wifi connection between 2 sites, with
> decent line of sight(a few trees but I can see each antenna). Covering
> about 1100 feet. I have 2 EtherDesigns Cantennas, a 10" on the
> recieving end(a Linksys PCI card WMP54G) and a 5" on the access point,
> a Linksys WRT54G. According to the linksys software on the recieving

Etherdesigns shows about 18dB gain for the "Cantenna III", and 10dB for the
"Cantenna II". 10dB would agree with cans that I've built from 4" diameter
coffee cans. I think 18dB is a lot to get from a 10" Cantenna. If it is
18dB, it would be a very narrow beam, and might be difficult to aim, and
maintain aim. The FAQ refers to 5" as 8dB and 10" as 12dB gain.

Can you compare signal with and without the Cantenna, to see what the gain
is? How are you aiming the antennas? Pure line of sight might not be
correct. You should look for a signal peak, very slowly. Azimuth,
altitude, and rotation of the can (parallel verses horizontal of the wire
inside, or connector if you can't see the wire), and in between) will all
have an effect.

There is also some fresnel zone. This isn't a laser beam, but more of a
flashlight. If you can see the can through a hole in the trees, you might
not be getting good signal thought the hole.

I notice that the long distance test on the Etherdesigns site has the can
lying on a car trunk. The trunk would be reflecting and blocking the
signal, and the can is rotated off vertical in a convenient resting spot.
The cantenna doesn't appear to be connected to the Orinoco card, but I'll
assume artistic license in setting up for the photo shoot. I'll assume he
was standing with the laptop in his hands, pointing the cantenna by hand,
and then rearranged the equipment for the photo.

> Is the problem simply not a good enough connection? If so, why does the
> linksys show 3/5 signal and transmit fine? Or do I have another problem
> going on where I can't get an IP address?

Would it be possible to temporarily locate the PC closer? If not, bring
the WAP to the PC, it shouldn't need an internet connection to give you a
WiFi connection and an IP address. Troubleshooting the IP assignment
separately from the RF signal would be a good thing.

--
---
Clarence A Dold - Hidden Valley (Lake County) CA USA 38.8,-122.5

(E-Mail Removed) wrote:
>a Linksys WRT54G. According to the linksys software on the recieving
>end, I have a 51% signal, 3 out of 5 bars, and -77 dB with -79 dB of
>noise.

Only 2dB signal to noise? That can't be good...

William P. N. Smith <> wrote:
> (E-Mail Removed) wrote:
>>a Linksys WRT54G. According to the linksys software on the recieving
>>end, I have a 51% signal, 3 out of 5 bars, and -77 dB with -79 dB of
>>noise.

> Only 2dB signal to noise? That can't be good...

What do you expect from a can?

> > Only 2dB signal to noise? That can't be good...
>
> What do you expect from a can?

That's not because of the can!