joecool <joecool.2v63av@no-mx.wirelessforums.org> hath wroth:
>Have a relative with a cottage on a campground that recently put in free
>wifi for its members. But of course in her cottage she can't pick up the
>signal inside. Outside on the porch, she might grab signal at 20% or
>less, and hops around a lot. So I have been assigned with the task of
>"fixing her problem".
>
>Here is the info I gathered doing a site survey...
>First of all, I do not administer the local signal being put out, so I
>have no specific details or control over it.
Really? How about a photograph of the equipment or the antennas?
>Looks like they have an outside panel antenna mounted to a tower next
>to the main office. That's where the source is originating from. Her
>cottage is about 600 - 700 feet or so away from the originating antenna.
At 700ft, you're going to need a directional antenna at her end.
>Also through reasonable tree coverage, and over a small hill.
1. There is no such thing as a reasonable tree. Every time I've
tried to convince a tree to let my RF pass, the tree rejects my pleas.
I've tried various offerings and incantations, but nothing will make
the tree pass my signals.
2. Hills, whether small or large are even worse. Every time I try to
invoke the hill god to let my signal pass, the hill responds with
something like "eat dirt you worthless human". I suspect they are in
colusion with the trees.
>I used a
>laptop with a usb wireless adapter and an active usb extension mounted
>to a pole. Walked around trying various spots and heights with Cirond
>WinC software to measure strength.
Good. That's the way I do it but I use an ethernet wireless bridge
instead of USB. Actually, the last time I did it, I hung a wireless
router running in client mode on the back of a dish. Whatever works.
>What I found was at a certain
>location towards the rear of the cottage and about 10ft up, I would pick
>up signal at about 50% (I could never get more than that due to the
>distance I am guessing). By looking at the signal path visually, this
>was going through multiple trees, and one old building made of wood. by
>going up a minimum of 10ft, I was clearing the small hill. I am guessing
>they must be pumping out some decent power to be picking up 50% signal
>strength at a distance of 600ft or so and the trees & building.
I predict inevitable doom. It's easy enough to get a marginal
connection almost anywhere you have signal. However, keeping that
connection among the swaying trees and branches is going to be a
problem. In addition, you're apparently skirting the top of a small
hill, which probably puts it in the Fresnel Zone, which creates more
loss and more weirdness. I think you could probably make it work, but
I doubt if you can keep it working.
Try this simple test. Take you test pole and plant it firmly into the
ground with some kind of temporary tripod support. Aim and position
it as best you can. Then run Netstumbler for a while and watch the
signal levels. My guess is you'll see huge variations as things move
around (which includes you moving around).
My guess is that your unspecified model USB adapter was probably using
the stock ceramic or PCB antenna found in most such bottom of the line
devices. I'd give it an optimistic 0dBi gain. 6dBi is double your
range so I think you have a chance if you get a halfway decent
directional antenna. Notice that I said directional, not
omnidirectional. I suggest you repeat the test using a decent
directional antenna and perhaps 15 to 25 ft of the best coax cable you
can borrow. The coax will present some loss, but the antenna gain
should more than compensate for the loss.
>So here is the plan I have come up with...
>On the cottage, install a outside yagi antenna (+13.9 db) at 10ft up,
>run LMR-400 cable (15ft) into the cabin, connect to a linksys AP
>(WAP56G) in bridge mode, connect that to a WRT54G and broad cast to
>cover her cottage (on a different channel).
Barf. That's the way we did it in the stone age of wireless. It will
work but causes all manner or aesthetic and cable routing problems. I
have one installation in an airport hanger where the router is hanging
from the ceiling where the coax enters to keep things short.
Permit me to offer an alternative. Get a CPE (customer premisis
equipment) client bridge what does PoE (power over ethernet) and has
an external antenna connector. The antenna should be as high, big,
and ugly as possible. However, if it does not clear the hill, don't
bother, it's not going to work no matter how big and ugly. A small
antenna up very high is much better than an a big antenna down low.
Concentrate on the antenna as everything else is secondary.
Run outdoor (gel filled and UV proof) CAT5 from the CPE bridge radio
to the local router or computer. Some client bridges can handle more
than one MAC address so you may not need a router.
That leaves the selection of the mysterious radio. Anything that
calls itself a CPE, client bridge, customer radio, ethernet client
bridge, or similar buzzwords will work. Lots of indoor devices, such
as "game adapters" will also work, but will require waterproofing.
Some possibles:
<http://www.wlanparts.com/c=uBGqjloc2otmlUgD5dXlc2Vw4/category/client_systems/>
There are lots of others, but I'm too lazy to dig them out with
Google.
There are also versions that integrate the radio into the antenna:
<http://www.mikrotik.com/pricelist.php?sect=3#group7>
<http://www.tranzeo.com/products/radios/TR-CPQ-Series>
<http://www.tranzeo.com/products/radios/TR-6000-Series>
<http://www.tranzeo.com/products/radios/TR-CPE90-Series>
Make sure they'll do PoE or make sure you can get a PoE adapter (that
is really a 48VDC 802.3af adapter and some some simulated kludge).
>I am estimating I should be getting about 26 db of signal out.
26dBm(?) out of what?
>I have set up several wireless systems, but not much to do with
>rebroadcasting applications.
Consider yourself fortunate. Methinks repeaters, WDS bridges, and
mesh networks to be a nightmare. Avoid if possible though they all
are tempting. Incidentally, thank you for not suggesting putting a
repeater on top of the pole.
>If I was picking up 50% signal strength
>with my little usb adapter, would that be enough to run the setup I am
>proposing?
Dunno. Run this simple test. Setup your USB gizmo as a wireless
client. ping the access point continuously with something like:
ping 192.168.1.1 -t
If you get a consistent 2-6msec latency for the entire test duration,
I'm sure it can be made to work. If the latency is all over the map
plus some timeouts (which is what I would expect), you have some chain
saw or tower work to do.
>And to connect back at that distance? That should be about twice the
>power my usb adapter has.
Sigh. Double the "power" which I presume means the transmit RF
output, and you increase your range about 1.4 times. You'll barely
notice the increase in signal strength. It will also be in one
direction only as the other end will need to increase its power
identically in order to get a symmetrical improvement. That's another
reason why I suggest you concentrate on the antenna and not the radio.
>I am interested in hearing feedback from anyone who might have
>experience or thoughts in a setup like this.
Lots of experience. I live in a dense redwood, fir, and oak forest.
We're lucky if we can see the sun, much less get any RF through this
mess. I've done quite a bit of experimenting, which includes
repeaters, passive reflectors, brute force power, antenna games, and
beam forming. They all work to some degree, but there's no
replacement for real line of sight and a clear shot.
>Is this overkill, or do I
>need to use an additional amplifier?
Argh. You said a dirty word. Amplifiers are nothing more than
expensive jammers. You do NOT get much of range increase if you can't
hear the response from the other direction. On the other hand,
antennas increase the signal in both directions simultaneously. Think
antennas, not radios, amplfiers, boosters, and range extenders.
>I want to make a good educated
>guess on what I need before I start buying stuff to just try.
Educated guess is an oxymoron. A calculation would be more
interesting. However, I would need to know something about what
equipment is at the access point, the antenna pattern, exact
locations, line of sight, tree type, tree density, hill location, etc.
It's too much for doing this via usenet. You might have to settle for
an uniformed wild guess, which is acceptable if the store has a good
return policy.
>And no I don't think I can go high enough to clear the trees and
>buildings to get LOS. That would take a full blown tower.
Why would you want to blow up a perfectly good tower? How many feet
high would it need to be? Up to about 30ft, a pole will work just
fine. Radio Shock sells a usable 30ft push up pole.
You might also consider going underground to some place with a better
line of sight. I have quite a bit of buried cable running around the
forest and under the road. It's a bit of work, but not impossible.
Local forest soil is mulch, which is quite soft and easy to work. I've
used black flex sprinkler pipe, with CAT5, coax, fiber, etc shoved
inside.
--
Jeff Liebermann
jeffl@cruzio.com
150 Felker St #D
http://www.LearnByDestroying.com
Santa Cruz CA 95060
http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
Re: wireless application 
Linear Mode
