Question: singal strength, same each end?

Hi All,

I have an 18db parabolic antenna pointed at a Tin Cantenna about 800 metres
away line of sight.

From the netcard at the cantenna side, I'm getting 40% signal strength. My
access point rarely sees the cantenna however and generally not for very
long.

I've heard the antenna gain is same for send and receive so I take it that
my home made cantenna should be sending something out. Could it be that the
parabolic antenna sends in a wider arc which the cantenna can see but not
vice versa?

There are some tall trees to the immediate right of the line but I can see
the antenna clearly. I've read a little about the "Fersnal zone" and that
you shouldn't point the antennas directly at each other. Could this be
causing dramas?

Thanks for any input.

Jac

On Tue, 17 Aug 2004 09:37:02 +0800, "Amhir Sohail"
<(E-Mail Removed)> wrote:

>I have an 18db parabolic antenna pointed at a Tin Cantenna about 800 metres
>away line of sight.
>
>From the netcard at the cantenna side, I'm getting 40% signal strength. My
>access point rarely sees the cantenna however and generally not for very
>long.

Ok, something is wrong. If you move closer with the hardware, can the
access point eventually pickup the signal? Just checking if things
are properly configured.

>I've heard the antenna gain is same for send and receive so I take it that
>my home made cantenna should be sending something out. Could it be that the
>parabolic antenna sends in a wider arc which the cantenna can see but not
>vice versa?

Transmit and receive gains are identical. The antenna pattern doesn't
really matter with small aperature antennas. All that's inportant is
the forward antenna gain.

However, transmitter power output and receiver sensitivities are NOT
necessarily identical. Let's say you have a radical difference in
transmitter power output between ends of the link. Since the signal
strength the receiver displays is also a function of the other
transmitters power output, you'll get different signal levels. This
may be your problem.

However, unless you have something broken, the typical transmit signal
levels and receiver sensitivities are not sufficiently different to
have one end enormously different from the other. Perhaps with a
tower mounted power amplifier, but not with conventional radios.
Methinks something is broken.

The problem is that there's no way to tell which end is the problem.
That's were a 3rd radio is necessary for troubleshooting. Use this
3rd radio (preferably a laptop running Netstumbler) to connect to each
end, one at a time. By process of elimination, either the
transmitter, or the other receiver, on the comatose link path, with be
the culprit.

>There are some tall trees to the immediate right of the line but I can see
>the antenna clearly. I've read a little about the "Fersnal zone" and that
>you shouldn't point the antennas directly at each other. Could this be
>causing dramas?

It's Fresnel zone diffraction. For 800 meters range and 2.4Ghz, the
Fresnel zone is 5 meters radius at midpoint. If you have a clear line
of sight with a 10 meter diameter midpoint clearance, you should be
fine. If the tree is closer to one end, you'll need less clearance.
The Fresnel zone is also symmetrical, so that would not cause the
symmetry problem you're having.
http://gbppr.dyndns.org:8080/fresnel.main.cgi



--
Jeff Liebermann (E-Mail Removed)
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 AE6KS 831-336-2558

"Jeff Liebermann" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> On Tue, 17 Aug 2004 09:37:02 +0800, "Amhir Sohail"
> <(E-Mail Removed)> wrote:

> Ok, something is wrong. If you move closer with the hardware, can the
> access point eventually pickup the signal? Just checking if things
> are properly configured.

When I had both units in the same house they connected instantly with no
configuration.

Last night I swapped the access point (DLink 2100ap) and netcard (Netgear
WG311) around and I still got a 40% signal going the other way. I'm guessing
that rules out transmit and sensitivity issues?

The cable run to the netgear was pretty long (and I was using RG58 for half
of it) so I've ordered another 2100ap and I'll try bridge mode using Cat5 to
make most of the run to the PC.
I do have a laptop so I'll get netstumbler. How far does the signal spread?
Am I likely to see it from the ground if both antennas are pointing at each
other on roofs?

Trees are closer one end so maybe it will infract a little on the signal. Do
you think 40% signal is suffcient to get a connection?

Thanks for your response

Jac

On Wed, 18 Aug 2004 09:40:15 +0800, "Amhir Sohail"
<(E-Mail Removed)> wrote:

>
>"Jeff Liebermann" <(E-Mail Removed)> wrote in message
>news:(E-Mail Removed).. .
>> On Tue, 17 Aug 2004 09:37:02 +0800, "Amhir Sohail"
>> <(E-Mail Removed)> wrote:
>
>> Ok, something is wrong. If you move closer with the hardware, can the
>> access point eventually pickup the signal? Just checking if things
>> are properly configured.

>When I had both units in the same house they connected instantly with no
>configuration.

OK. The radio configuration (channel, WEP, etc) is correct.

>Last night I swapped the access point (DLink 2100ap) and netcard (Netgear
>WG311) around and I still got a 40% signal going the other way. I'm guessing
>that rules out transmit and sensitivity issues?

Yep. My 2nd best guess is a high VSWR antenna or coax connection.

>The cable run to the netgear was pretty long (and I was using RG58 for half
>of it)

What's that in meters (or feet). Coax losses at 2.4GHz are
horrendous. RG-58 absolutely sucks at 2.4GHz. I'm suprised it works
at all.

>so I've ordered another 2100ap and I'll try bridge mode using Cat5 to
>make most of the run to the PC.

Yep. Keep the access point as close as possible to the antenna. If
possible go for an outside radio mount with PoE (power over ethernet)
to supply power to the radio.

>I do have a laptop so I'll get netstumbler. How far does the signal spread?

How high is your antenna? An 18dBi dish will have about a 10 degree
-3dB beamwidth. Try:
distance = height / tan (5degrees) = height / 0.087

>Am I likely to see it from the ground if both antennas are pointing at each
>other on roofs?

How high are your antennas? Most antennas have "side lobes" that will
allow for a connection nearby that are out of the beamwidth.

>Trees are closer one end so maybe it will infract a little on the signal. Do
>you think 40% signal is suffcient to get a connection?

Yes. 20% should yield a flakey, but useable connection.
--
# Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
# 831.336.2558 voice http://www.LearnByDestroying.com
# (E-Mail Removed)
# 831.421.6491 digital_pager (E-Mail Removed) AE6KS

"Jeff Liebermann" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> On Wed, 18 Aug 2004 09:40:15 +0800, "Amhir Sohail"
> <(E-Mail Removed)> wrote:
> Yep. My 2nd best guess is a high VSWR antenna or coax connection.
>
> >The cable run to the netgear was pretty long (and I was using RG58 for
half
> >of it)
>
> What's that in meters (or feet). Coax losses at 2.4GHz are
> horrendous. RG-58 absolutely sucks at 2.4GHz. I'm suprised it works
> at all.

I had about 9 metres (flame away :-)) of RG58 on the cantenna and I've
dropped that to three now. Signal strength didn't seem to change too much
though, maybe something up there?

The parabolic has 6 metres of LMR400 and a half metre pigtail RG58.

I'm currently borrowing the dish but on the weekend I'm getting another one
(we had a paytv company that went bust and left these up on subscribers'
roofs) which I'll trim that cable down to 3 metres as well.

>>Trees are closer one end so maybe it will infract a little on the signal.
Do
>>you think 40% signal is suffcient to get a connection?

>Yes. 20% should yield a flakey, but useable connection.

Well that's strange then, for the 2 hours I was mucking around with this
last night that signal strength never went below 40% and never lost
connection.

I'll try with the bridge connection tonight and let the group know.

Thanks
Jac

"Amhir Sohail" <(E-Mail Removed)> wrote in message
news:4122d32e$0$27241$(E-Mail Removed) om.au...
> Well that's strange then, for the 2 hours I was mucking around with this
> last night that signal strength never went below 40% and never lost
> connection.
>
> I'll try with the bridge connection tonight and let the group know.
>
> Thanks
> Jac

I bought another Dwl-2100AP and tried in bridge mode last ngiht. Couldn't
get it working so I'll get up and adjust my antenna on my roof.

I do have a complaint about the Dlink 2100AP though. There is no site
survey! I couldn't check signal strength to verify I still had some
connection..

I really need an external antenna connection on the laptop to hook straight
into the antennas to aim. I'll have to make do with using netstumbler from a
halfway point and doign it that way. I'll update later

Jac

"Amhir Sohail" <(E-Mail Removed)> wrote in message
news:4123fcc4$0$27219$(E-Mail Removed) om.au...

> I really need an external antenna connection on the laptop to hook
straight
> into the antennas to aim. I'll have to make do with using netstumbler from
a
> halfway point and doign it that way. I'll update later
>
> Jac

I've tested the link using netstumbler, with the two parabolic dishes over
800 metre I'm getting an RSSI figure of around 20-24. Is this a percentage
value? The light is green in netsumbler so I guess that means that its
enough signal. It does go grey every so often (10 secs or so)which seams to
be a loss of signal. Is this normal or is my link still flakey?

Thanks
Jac

"Jac" <(E-Mail Removed)> wrote in message
news:4128803e$0$27219$(E-Mail Removed) om.au...
>
> "Amhir Sohail" <(E-Mail Removed)> wrote in message
> news:4123fcc4$0$27219$(E-Mail Removed) om.au...

> I've tested the link using netstumbler, with the two parabolic dishes over
> 800 metre I'm getting an RSSI figure of around 20-24. Is this a percentage
> value? The light is green in netsumbler so I guess that means that its
> enough signal. It does go grey every so often (10 secs or so)which seams
to
> be a loss of signal. Is this normal or is my link still flakey?
>
> Thanks
> Jac

I've just hook my laptop up next to my pc and run netstumbler again. I get
the greyed connection every so often to I guess thats not an issue with my
main link. The RSSI is on about 50 and 60 seems to be the max. I guess 20
RSSI is a little low then. Signal strenth as measured by the netgear
software however is generally around 50%.

Jac

Ok, the problem was the antenna setting on the Dlink 2100AP access point. It
was set to "best" as soon as I changed it to "1" which I assume is external
antenna I had immediate connection.

Thanks for your help on this one Jeff

Jac



>
> I have an 18db parabolic antenna pointed at a Tin Cantenna about 800
metres
> away line of sight.
>
> From the netcard at the cantenna side, I'm getting 40% signal strength. My
> access point rarely sees the cantenna however and generally not for very
> long.
>
> I've heard the antenna gain is same for send and receive so I take it that
> my home made cantenna should be sending something out. Could it be that
the
> parabolic antenna sends in a wider arc which the cantenna can see but not
> vice versa?
>
> There are some tall trees to the immediate right of the line but I can see
> the antenna clearly. I've read a little about the "Fersnal zone" and that
> you shouldn't point the antennas directly at each other. Could this be
> causing dramas?
>
> Thanks for any input.
>
> Jac
>
>
>