Hi Bryan

I'd try and model this for you but the s/w I use is broken at the moment.

You can calculate path loss over an LOS path fairly easily. If you start
with the RF output in dBm, add antenna gains, subtract the path and
cable losses you'll get a close result. This can then be applied against
the receiver sensitivity (I use to use -87dBm for 802.11 links) If you
want to model in the existence of a reflector or (say) back to back yagi
you do it is two paths using the received power at the reflecting device
as the source power for the next minus about 3dB. The results tend to be
less than encouraging in most places but it depends on the application.
One study I did (for example) was to get a 410MHz signal into a railway
cutting on the edge of a mountain. (ie no source of reflection from the
other side) The 25W TX was about 2km away and we modeled two back to
back 13dBi antennas. With -82dBm as the lowest specd RX signal we only
got about 300m range from the reflecting array.

Freds suggestion of using a billboard is however a good one and simple
to experiment with. I also agree that the powerline wont be much help.

Pls tell us how it works out!

Cheers Bob VK2YQA

Bryan Martin wrote:
I am attempting to setup a 802.11b bridge between my location (siteA) down
the powerlines but off to the right NLOS (siteB) roughly 1 mile away.

Well the downside is that I dont think I can pull off the reflector senerio
or complete line of site going over the tree's. At the end of this
powerline is someone else's lot so I can't place anything on or around that.
I only have permission to put the unit in the building located behind
someone's house. Ideally I would rather not have any external antenna's on
the building if at all possible to make the intrusion on property as light
as possible. Thats what brought me to my question. The building is just an
old cheap yard shed more than anything. It has power and 2 windows on the
back side facing the back of the lot. I hope to be able to place the
wireless unit close to the window which should help push the signal. If the
building was another 25-30 foot more to the right I might would have LOS to
it but that is also when you cross over onto someone elses land which is not
an option. As for seeing the signal I am almost completely positive the
grid antenna would be able to intersect the wireless signal at the end of
the powerline but I am unsure if that is enough to get the signal back?

To see a very rough layout of what I am refering to see this
http://home.triad.rr.com/tempdir/images/layout.jpg


"Bob Bob" wrote in message
...
Hi Bryan

I'd try and model this for you but the s/w I use is broken at the moment.

You can calculate path loss over an LOS path fairly easily. If you start
with the RF output in dBm, add antenna gains, subtract the path and cable
losses you'll get a close result. This can then be applied against the
receiver sensitivity (I use to use -87dBm for 802.11 links) If you want to
model in the existence of a reflector or (say) back to back yagi you do it
is two paths using the received power at the reflecting device as the
source power for the next minus about 3dB. The results tend to be less
than encouraging in most places but it depends on the application. One
study I did (for example) was to get a 410MHz signal into a railway
cutting on the edge of a mountain. (ie no source of reflection from the
other side) The 25W TX was about 2km away and we modeled two back to back
13dBi antennas. With -82dBm as the lowest specd RX signal we only got
about 300m range from the reflecting array.

Freds suggestion of using a billboard is however a good one and simple to
experiment with. I also agree that the powerline wont be much help.

Pls tell us how it works out!

Cheers Bob VK2YQA

Bryan Martin wrote:
I am attempting to setup a 802.11b bridge between my location (siteA)
down
the powerlines but off to the right NLOS (siteB) roughly 1 mile away.

Hi Bryan

Am a little confused. Probably from working nights and not getting
enough sleep!

From your diagram you are trying to get a signal between the site that
has the 24dB gridpack and the "Outside building". You have done some
tests from the road and it seems to work.

I'd expect you biggest prohblem is the pine forest. They are excellent
for attenuating microwave signals! My gut feel is that you'll need a
gain antenna at the "outside building" as well as the existing one. This
very much depends on the pine tree density that I dont have any actual
attenuation figures on. I actually modelled a 2.4GHz link to my mother
in laws place about half a mile away through trees and with gridpacks at
each end. I had about 50dB of margin but was still worried about 100-200
yards of pine trees. If nothing else you'll get multipath interference
and may need to play with polarization.

If all else fails try bouncing the signal off the house. You might be
surprised!

Sorry I cant be more helpful. Pine needles and tree trunks I dont have
any figures on!

Cheers Bob

Bryan Martin wrote:
Well the downside is that I dont think I can pull off the reflector senerio
or complete line of site going over the tree's. At the end of this
powerline is someone else's lot so I can't place anything on or around that.
I only have permission to put the unit in the building located behind
someone's house. Ideally I would rather not have any external antenna's on
the building if at all possible to make the intrusion on property as light
as possible. Thats what brought me to my question. The building is just an
old cheap yard shed more than anything. It has power and 2 windows on the
back side facing the back of the lot. I hope to be able to place the
wireless unit close to the window which should help push the signal. If the
building was another 25-30 foot more to the right I might would have LOS to
it but that is also when you cross over onto someone elses land which is not
an option. As for seeing the signal I am almost completely positive the
grid antenna would be able to intersect the wireless signal at the end of
the powerline but I am unsure if that is enough to get the signal back?

You are right in that I am attempting to pass the signal from the building
to the 24db antenna down the powerline NLOS. Distance wise I have passed
the signal even further than I need NLOS but it was a strait shot NLOS not
off to the right like I am going to attempt. I will be able to actually try
this tomorrow and may post my results.


"Bob Bob" wrote in message
...
Hi Bryan

Am a little confused. Probably from working nights and not getting enough
sleep!

From your diagram you are trying to get a signal between the site that has
the 24dB gridpack and the "Outside building". You have done some tests
from the road and it seems to work.

I'd expect you biggest prohblem is the pine forest. They are excellent for
attenuating microwave signals! My gut feel is that you'll need a gain
antenna at the "outside building" as well as the existing one. This very
much depends on the pine tree density that I dont have any actual
attenuation figures on. I actually modelled a 2.4GHz link to my mother in
laws place about half a mile away through trees and with gridpacks at each
end. I had about 50dB of margin but was still worried about 100-200 yards
of pine trees. If nothing else you'll get multipath interference and may
need to play with polarization.

If all else fails try bouncing the signal off the house. You might be
surprised!

Sorry I cant be more helpful. Pine needles and tree trunks I dont have any
figures on!

Cheers Bob

Bryan Martin wrote:
Well the downside is that I dont think I can pull off the reflector
senerio or complete line of site going over the tree's. At the end of
this powerline is someone else's lot so I can't place anything on or
around that. I only have permission to put the unit in the building
located behind someone's house. Ideally I would rather not have any
external antenna's on the building if at all possible to make the
intrusion on property as light as possible. Thats what brought me to my
question. The building is just an old cheap yard shed more than
anything. It has power and 2 windows on the back side facing the back of
the lot. I hope to be able to place the wireless unit close to the
window which should help push the signal. If the building was another
25-30 foot more to the right I might would have LOS to it but that is
also when you cross over onto someone elses land which is not an option.
As for seeing the signal I am almost completely positive the grid antenna
would be able to intersect the wireless signal at the end of the
powerline but I am unsure if that is enough to get the signal back?