On 4 Jul 2005 07:56:18 -0700,
(E-Mail Removed) wrote:
>I've recently installed a Netgear DG834G for a relative. I chose this
>unit primarily because it got good reviews for coverage. I've found
>however that the coverage is awful. The unit is installed at desk
>height, about 20cm to the side of corner. Round that corner is along
>hallway and I'm finding that I lose signal (both with a laptop and
>standalone WIFI signal strength meter) after a mere 6 meters.
>
>I realise it's hard to diagnose an issue like this with just the
>above information, but has anyone got any ideas why the range is so
>poor? Are there any environmental that could reduce the range so much,
>does it sound like a faulty unit, may an external antenna help or would
>do I just need a more powerful router - six meters does seem pretty
>poor?
>
>Any advice very gratefully received.
Amazing. I recently ran into EXACTLY the same issue.
I recently installed a Netgear WGR614v5, which is the same as the
DG834G router without the DSL modem, in a medical office. I had
similar coverage problems only when turning the corner into offices.
However, I didn't have time to investigate much but did have a
WRT54Gv2.2 handy that was going to my next customer. Using the WRT54G
improved the range somewhat, but didn't really solve the basic
problem. Turning the corner into the offices tended to kill the
connection.
So, I installed Netstubler on the customers laptop and did some
comparison testing between the two wireless routers. I could put
either router in the window, and walk down the street, maintaining
line of sight about 50 meters before the signal would become totally
useless. Line of sight range was somewhat better with Linksys. I
would consider both to be operating normally as long as I had line of
sight. I did have to change channels to avoid interference from a
nearby access point.
However, inside the office, the difference was dramatic. Turn the
corner at about 10 meters, using the WFGR614v5, and the connection
would literally drop. Do the same with the Linksys, and it would slow
down dramatically, but continue to work. Apparently, something was
different in the way the chipsets handled reflections or signal loss.
It didn't take long to discover that the walls were full of foil
backed insulation. There was even aluminum foil duct tape to connect
the gap across the vertical studs. Even the doors had some kind of
foil or metal inside. This was undescored when I found that I could
get 3-4 bars on my phone call with the door open, but "no service"
when the door was closed. These offices were shielded. (I later
found out that this was intentional).
Since I didn't want to give them the WRT54G, I decided to see if I
could tweak the settings to improve the situation. The major change
was to fix the wireless speed at 6Mbits/sec which is the slowest OFDM
speed but is faster than their 1.5Mbit/sec DSL connection. That
worked with the Netgear WGR614 sufficiently to be considered
acceptable. They still had to leave the doors open to use the
wireless.
My guess(tm) is that the Netgear products emphasize speed over data
reliability and will aggressively attempt to use the fastest speed
possible. This is wonderful for getting impressive benchmarks, but
not very useful in a highly reflective environment.
My suggestion is to repeat my test. First, perform a line of sight
test. Use Netstumbler to record the signal levels. You should get at
least 25 meters range without any dropouts or tendencies to die. At
extreme range, 50 meters would be my guess for normal operation. If
the router can't do that, it's defective. Verify with a 2nd laptop if
possible.
Then use Netstumbler to do a "site survey" of the working area. Are
there any RF holes? Any places where the "noise" level goes
rediculously high, which might imply interference or reflections?
Does repositioning the DG834G help? Do this with the wireless speed
set to "auto". Then try it again with the speed fixed at 6Mbits/sec.
If the signal strengh is generally higher (it should be), and the
noise levels are generally lower (they might be), then you have a
winner. If you have a different model access point or router
available, it can also be tested.
Incidentally, I have a really bad attitude about all in one
modem/router/wireless boxes. Details if you want them.
--
Jeff Liebermann
(E-Mail Removed)
150 Felker St #D
http://www.LearnByDestroying.com
Santa Cruz CA 95060 AE6KS 831-336-2558
Similar Threads
Linear Mode
