dslreports logo
 
    All Forums Hot Topics Gallery
spc
uniqs
30

shdesigns
Powered By Infinite Improbabilty Drive
Premium Member
join:2000-12-01
Stone Mountain, GA
(Software) pfSense
ARRIS SB6121

shdesigns to disconnected

Premium Member

to disconnected

Re: Signal Strength Meter for PCS Cellular with -100dBm Sens.

I'm in a similar situation here. Tower is less than a mile away but too many trees and hills. Almost no signal.

Same for UHF TV. No direct line of sight. I get a stronger signal with the antenna pointing any direction other than toward the tower.

The problem is multipath. I may get a decent signal and usually do but a slight wind and the level starts oscillating. Rain, temperature, aircraft, trucks driving by all cause a momentary cancellation.

I have an amplified antenna but it only helps a tiny bit. I see little difference between a highly directional antenna and a bow-tie antenna. Both get agood signal but intermittent nulls.

Unless you can get a LOS setup, I doubt any amp/antenna combination will be reliable. Only a multi-antenna setup with some discriminator would be effective.

disconnected
@108.247.170.x

disconnected

Anon

Multipath is a killer. And at nearly 2GHz, individual LEAVES become reflectors. Millions of them, fluttering in the breeze.
I noticed yesterday that my signal was above normal for about half a minute. I noticed one of those big army helicopters flying south of me. Must have been reflecting some signal from the tower.
I actually get almost as good a signal from the tower 18 miles away as from the one 3.5 miles away, because the father tower is not so shadowed by the slope of the mountain. I'm on the level plateau, quite a ways from the dropoff, so everything in the valley is shadowed. When the trees are bare, I can see the mountain range on the other side of the valley in the next town over.
Last night, I had -102dBm in the studio, but today, we have very heavy rain and signal dropped to -125dBm. I could strangle the idiots who thought it would be a good idea to rollout PCS in a hilly, forested part of the country.
My idea of a LOS setup would be a multi-stacked J-pole array. But alas, I tried something similar with a sector panel antenna with an 90° beamwidth and an 8° elevation pattern, but it didn't work well compared to the log periodic Wilson antenna (the one in the white radome).
There are still wild swings in signal. This afternoon, standing under the inside panel antenna, I observed the signal dropping to -125dBm for a minute, then it came back up to -93 or so. Either the tower shut off the transmitter momentarily, or that was one heck of a deviation in signal.
TheMG
Premium Member
join:2007-09-04
Canada
MikroTik RB450G
Cisco DPC3008
Cisco SPA112

TheMG

Premium Member

said by disconnected :

I could strangle the idiots who thought it would be a good idea to rollout PCS in a hilly, forested part of the country.

I'm surprised they decided to put the towers down below. Usually carriers like to stick their towers on the tallest hill they can find (within reason).
lutful
... of ideas
Premium Member
join:2005-06-16
Ottawa, ON

lutful to disconnected

Premium Member

to disconnected
said by disconnected :

Multipath is a killer. And at nearly 2GHz, individual LEAVES become reflectors. Millions of them, fluttering in the breeze.

Notice the huge difference between windy and calm days: »Re: RSSI fluctuations

Curiously my very first post in DSLR was about making 2.4Ghz links through foliage ( »Re: Country House to City House (2 mile NLOS) link ) back in 2004. I posted example paths through foliage later on (
»Best way to deal with wet foliage? ) and also provided suggestions to other WISPs to make such links.

The obscure "multi-path" antennas we used in those links had wider beam (because of moderate gain) and was somewhat insensitive to polarization. That explained their seemingly better performance than higher gain fixed polarization antennas.

I did some research in 2006/7 using arrays of single patch antennas (nominally 6.5dBi each) oriented at random angles plus with a low insertion loss combiner and LNA. That worked even better through foliage.


disconnected
@108.247.170.x

disconnected

Anon

That's exactly the concept that I was trying to go after with my 90° beamwidth sector antenna. But then I discovered that the technical advisor from Wilson was right: use the highest gain parabolic dish that you can get. That improve the signal but of course an improvement over the next zilch is not much of a signal still. Right now we're having a band of nasty thunderstorms moving through the northeast it stretches from Canada to Delaware. I saw the signal completely wink out at several points where the phone displayed searching for service. Rain is a real problem when it comes down this hard.