Home All Groups Group Topic Archive Search About

Upgrading wireless protocol

Author
13 Mar 2005 4:05 PM
Jeff & Kate Wooddell
I have an HP laptop with built-in 802.11b wireless, but would like to
upgrade to G or Super-G.  Can I just buy a notebook adapter for the faster
protocol (along with a matching router)?  And would that override the
built-in "b"?

Jeff Wooddell

Author
13 Mar 2005 5:00 PM
Duane Arnold
"Jeff & Kate Wooddell" <jkwoodd***@yahoo.com> wrote in
news:0Iednf4tXPdw-anfRVn-2A@comcast.com:

> I have an HP laptop with built-in 802.11b wireless, but would like to
> upgrade to G or Super-G.  Can I just buy a notebook adapter for the
> faster protocol (along with a matching router)?  And would that
> override the built-in "b"?
>

I don't see why you couldn't go to the Device Manager on the O/S and
disable the wireless B device and go from there.

Duane :)
Author
13 Mar 2005 5:02 PM
Jeff Liebermann
On Sun, 13 Mar 2005 08:05:41 -0800, "Jeff & Kate Wooddell"
<jkwoodd***@yahoo.com> wrote:

>I have an HP laptop with built-in 802.11b wireless, but would like to
>upgrade to G or Super-G.  Can I just buy a notebook adapter for the faster
>protocol (along with a matching router)?  And would that override the
>built-in "b"?

Yes.  You can run two wireless cards in your unspecified model laptop
without difficulty.  Just disable the existing 802.11b internal card
in your unspecified operating system.  It doesn't "override" or
automatically disable the 802.11b card.  You can actually run them
both simultaneously, but on different channels.

If your unspecified HP laptop is fairly recent, it probably uses a
mini-PCI card as the existing internal 802.11b device.  It may be
possible to find a replacement mini-PCI card with an Atheros chipset
that supports Super-G, instead of adding a 2nd card.  Size and driver
issues will need to be checked, but I think it's possible.
  http://www.xserves.com/snsa/cm9.html
(There are probably others).

However, permit me to offer an alternative.  You'll see a rather large
increase in thruput going from 802.11b to 802.11g.  However, you will
not see anything as spectacular going from 802.11g to Super-G.  The
108mbit speeds are effective only at fairly close range (less than
6ft) and only under ideal circumstances (no 802.11b radios in sight,
no interference, minimum reflections, etc).  Therefore, methinks that
simply replacing the (assumed) mini-PCI 802.11b card with an
"ordinary" 802.11g mini-PCI card, would be adequate for whatever you
have planned. 

HP probably has a recommended 802.11g card as an upgrade.  Without a
model number, I can't offer a recommendation.  Selecting a random card
might be a problem as detailed on this page for wireless mini-PCI card
compatibility for IBM Thinkpads:
  http://www.srcf.ucam.org/~mjg59/thinkpad/tables.html



--
Jeff Liebermann    je***@comix.santa-cruz.ca.us
150 Felker St #D   http://www.LearnByDestroying.com
Santa Cruz CA 95060    AE6KS  831-336-2558
Author
14 Mar 2005 3:46 AM
JB
Show quote Hide quote
"Jeff Liebermann" <je***@comix.santa-cruz.ca.us> wrote in message
news:s4r831pr33k8kd27ss0nc67ffg5s5d5imf@4ax.com...
> On Sun, 13 Mar 2005 08:05:41 -0800, "Jeff & Kate Wooddell"
> <jkwoodd***@yahoo.com> wrote:
>
>>I have an HP laptop with built-in 802.11b wireless, but would like to
>>upgrade to G or Super-G.  Can I just buy a notebook adapter for the faster
>>protocol (along with a matching router)?  And would that override the
>>built-in "b"?

>
> However, permit me to offer an alternative.  You'll see a rather large
> increase in thruput going from 802.11b to 802.11g.  However, you will
> not see anything as spectacular going from 802.11g to Super-G.  The
> 108mbit speeds are effective only at fairly close range (less than
> 6ft) and only under ideal circumstances (no 802.11b radios in sight,

Using your own averages from another thread, standard 802.11g runs 20 Mbps.
But Super-G runs easily 45 and in good conditions will top 60 Mbps. And by
good conditions, I mean no other wireless nets nearby and no channel
interference. I'm not sure where the under six feet comes from. It's the
exact *opposite with the D-Link 624, which actually starts losing throughput
too close to the router and runs nicely (maybe 30 Mbps) from 100 feet with
no interference on the same floor of my house.

So, if you get 5 Mbps on an older 802.11b routers, 5 to 60 is much more
impressive than 5 to 20.
Author
14 Mar 2005 3:57 AM
DLink Guru
> I'm not sure where the under six feet comes from. It's the exact *opposite
> with the D-Link 624, which actually starts losing throughput too close to
> the router and runs nicely (maybe 30 Mbps) from 100 feet with no
> interference on the same floor of my house.


Thats what I would expect, since most transmitters have a dead spot about 2
to 3 feet around it.

Robert...
Author
14 Mar 2005 4:03 AM
Brian K
interference. I'm not sure where the under six feet comes from. It's the
> exact *opposite with the D-Link 624, which actually starts losing
> throughput too close to the router and runs nicely (maybe 30 Mbps) from
> 100 feet with no interference on the same floor of my house.
>
>
>

Interesting. How did you measure your 30 Mbps?

Brian
Author
14 Mar 2005 6:46 AM
Jeff Liebermann
On Mon, 14 Mar 2005 04:03:54 GMT, "Brian K" <iibntgyea4
REMOVE_T***@hotmail.com> wrote:

>interference. I'm not sure where the under six feet comes from. It's the
>> exact *opposite with the D-Link 624, which actually starts losing
>> throughput too close to the router and runs nicely (maybe 30 Mbps) from
>> 100 feet with no interference on the same floor of my house.

>Interesting. How did you measure your 30 Mbps?

Assorted benchmarking software:

IPerf
  http://dast.nlanr.net/Projects/Iperf/

Chariot

http://www.ixiacom.com/products/performance_applications/pa_display.php?skey=pa_ixchariot&section=perftest

Netstat Live
  http://www.analogx.com/contents/download/network/nsl.htm

NetBench
  http://www.etestinglabs.com/benchmarks/netbench/default.asp

Unix
  time (cat big_file | rcmd remote_machine "dd of=/dev/null")

List of various benchmarking software from PC Magazine
  http://www.pcmag.com/category2/0,1738,1938,00.asp

Or, you can just copy a known large file and time how long it takes to
copy.


--
Jeff Liebermann    je***@comix.santa-cruz.ca.us
150 Felker St #D   http://www.LearnByDestroying.com
Santa Cruz CA 95060    AE6KS  831-336-2558
Author
14 Mar 2005 8:02 AM
Brian K
Thanks, but I was interested to know how JB did it. My transfer rate timing
large file copying, using FTP, DU meter etc is only 15 Mbps with Wireless G
Speedbooster. This is with two computers in the same room, a few metres from
the router. I can't get it any higher.

Brian



Show quoteHide quote
>
> Assorted benchmarking software:
>
> IPerf
http://dast.nlanr.net/Projects/Iperf/
>
> Chariot
>
> http://www.ixiacom.com/products/performance_applications/pa_display.php?skey=pa_ixchariot&section=perftest
>
> Netstat Live
http://www.analogx.com/contents/download/network/nsl.htm
>
> NetBench
http://www.etestinglabs.com/benchmarks/netbench/default.asp
>
> Unix
>  time (cat big_file | rcmd remote_machine "dd of=/dev/null")
>
> List of various benchmarking software from PC Magazine
http://www.pcmag.com/category2/0,1738,1938,00.asp
>
> Or, you can just copy a known large file and time how long it takes to
> copy.
>
>
> --
> Jeff Liebermann    je***@comix.santa-cruz.ca.us
> 150 Felker St #D   http://www.LearnByDestroying.com
> Santa Cruz CA 95060    AE6KS  831-336-2558
Author
14 Mar 2005 4:58 PM
Jeff Liebermann
On Mon, 14 Mar 2005 08:02:06 GMT, "Brian K" <iibntgyea4
REMOVE_T***@hotmail.com> wrote:

>Thanks, but I was interested to know how JB did it. My transfer rate timing
>large file copying, using FTP, DU meter etc is only 15 Mbps with Wireless G
>Speedbooster. This is with two computers in the same room, a few metres from
>the router. I can't get it any higher.

Some hints:

- How many meters?  Speeds change drastically below about 1 meter and
  after about 3 meters.
- What kind of hardware are you using? 
- Are both computers wireless or just one?  15Mbit/sec is about what
  you'll get if you go from wireless to wireless computah through an
  access point.  You should only be testing one wireless link at a
  time.
- What make and muddle of computah(s)? 
- Which OS?  Which OS mutation?
- Do you have another wireless computers available to act as a tie
  breaker? 
- Are you getting a connection speed of at least 54Mbit/sec?
- Have you tried "optimizing" or "tweaking" the TCP/IP stack? 
  Check your work at:
    http://www.dslreports.com/tweaks
  and follow their recommendations.
- Any sources of interference in the area running 802.11b?
- What speed do you get if you disable the Speedbeast feature?
- Have you tried the same benchmarks with a wired connection?
- Any MAC address filters or ACL (access control list) type of filters
  in the wireless router or access point?  If so, they can slow things
  down a bit.
- If WRT54GS, are you using the stock firmware or an alternative?
- WEP64, WEP128, WPA-PSK, or no encrytion?

I forgot about DuMeter.  Add to list:
  http://www.dumeter.com

Benchmarks for WRT54GS from Tom's Hardware:
  http://www.tomsnetworking.com/Reviews-142-ProdID-WRT54GS-8.php
Average speed at close range was 33Mbits/sec.  Top speed (for one
minute file transfers) was 38Mbits/sec.  (first graph)

For comparison, the non-SpeedBoost WRT54G yielded 21Mbit/sec average
and 25Mbits/sec peak in the same test.  (2nd graph)

The 3rd graph shows the effects of encryption type:
  "If you open up the full-sized view and eyeball the steady portions
  of each plot, I think you'll see that WPA-PSK with AES encryption
  imposes virtually no throughput penalty. Using WEP will cost you
  about 10% (33 vs 36 Mbps) and enabling WPA-PSK with TKIP extracts
  the worst penalty at about 17% (30 vs. 36 Mbps)."





--
Jeff Liebermann    je***@comix.santa-cruz.ca.us
150 Felker St #D   http://www.LearnByDestroying.com
Santa Cruz CA 95060    AE6KS  831-336-2558
Author
14 Mar 2005 9:32 PM
Brian K
..
>
> Some hints:
>
> - How many meters?  Speeds change drastically below about 1 meter and
>  after about 3 meters.

The computers were in the same room when I first set up the system. One
computer was 2 metres from the router, the other 3 metres. I can carry my
laptop about 35 metres from the router before losing the connection and
there are several plaster walls in between.



> - What kind of hardware are you using?

Linksys WRT54GS router, WMP54GS cards. Three computers have these wireless
cards, now each is in a different room. There are two brick walls between
the router and two of these computers and the transfer rate for large files
is around 8 Mbps. .


> - Are both computers wireless or just one?  15Mbit/sec is about what
>  you'll get if you go from wireless to wireless computah through an
>  access point.  You should only be testing one wireless link at a
>  time.

You are correct, I have been testing wireless to wireless through a router.
So 15 Mbps is all I can expect? There is no facility to turn off
Speedbooster, it's always on. After the first month my router was replaced
(by Linksys) but there was no speed change

> - What make and muddle of computah(s)?

Dell 8400, Dell 450, Gateway 1300(?)

> - Which OS?  Which OS mutation?

WinXP SP2

> - Do you have another wireless computers available to act as a tie
>  breaker?

Three desktops. Also have a Dell laptop with wireless G. When it is turned
on, the transfer rate between the other computers actually increases by 5 to
10% and the DU Meter graph looks more stable. The link speed naturally drops
from 125 to 54 Mbps.

> - Are you getting a connection speed of at least 54Mbit/sec?

Link speed 125 Mbps.

> - Have you tried "optimizing" or "tweaking" the TCP/IP stack?

No.

>  Check your work at:
>    http://www.dslreports.com/tweaks
>  and follow their recommendations.

Server busy.

> - Any sources of interference in the area running 802.11b?

No

> - What speed do you get if you disable the Speedbeast feature?

Can't be disabled except perhaps by turning on laptop.

> - Have you tried the same benchmarks with a wired connection?

Wired gives 65 Mbps.


> - Any MAC address filters or ACL (access control list) type of filters
>  in the wireless router or access point?  If so, they can slow things
>  down a bit.

No

> - If WRT54GS, are you using the stock firmware or an alternative?

Stock. Latest firmware.


> - WEP64, WEP128, WPA-PSK, or no encrytion?

WPA-PSK. Doesn't slow it down.
>
> I forgot about DuMeter.  Add to list:
http://www.dumeter.com


Thanks for your interest Jeff. So you feel 15 Mbps, the way I have been
measuring it, is OK? For example if a 1000 MB file takes 500 seconds to be
copied to another computer, this is 2 MB/sec (16 Mbps).


Brian
Author
14 Mar 2005 11:43 PM
Jeff Liebermann
On Mon, 14 Mar 2005 21:32:55 GMT, "Brian K" <iibntgyea4
REMOVE_T***@hotmail.com> wrote:

>> - Are both computers wireless or just one?  15Mbit/sec is about what
>>  you'll get if you go from wireless to wireless computah through an
>>  access point.  You should only be testing one wireless link at a
>>  time.

>You are correct, I have been testing wireless to wireless through a router.
>So 15 Mbps is all I can expect?

I think so but I'm not 100.0% sure.  Easy way to be sure is to connect
*ONE* of the test computahs into the wired ethernet LAN port on the
WRT54GS and repeat the test with only one wireless link.  I can easily
try it in my office, but not now.  Gotta run... more later.


--
# Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
# 831.336.2558 voice  http://www.LearnByDestroying.com
#                         je***@comix.santa-cruz.ca.us
#                           je***@cruzio.com     AE6KS
Author
15 Mar 2005 5:40 AM
Brian K
>
> I think so but I'm not 100.0% sure.  Easy way to be sure is to connect
> *ONE* of the test computahs into the wired ethernet LAN port on the
> WRT54GS and repeat the test with only one wireless link.  I can easily
> try it in my office, but not now.  Gotta run... more later.
>
>


I disabled the wireless card on one of the downstairs computers and ran a
cable from that computer to the router. My current wireless computer is 7
metres from the router with a clear line of sight.

I copied a single file, 1042 MB. Took 407 seconds which works out to be 2.56
MB/sec or 20.5 Mbps. The DU Meter showed an average rate of 22.1 Mbps. I
repeated the file copy with the laptop turned on in the background and the
figures were 430 seconds, 2.42 MB/sec, 19.4 Mbps and 21.0 Mbps on the DU
Meter. Link speed was 125 Mbps before the laptop was turned on and 54 Mbps
with it on.

What do you think about this? Still seems slow to me but maybe I have high
expectations.

Brian
Author
15 Mar 2005 5:56 AM
Brian K
After looking at Tom's graphs again I see I need to repeat my test with the
wireless computer 2 metres from the router.

Brian
Author
15 Mar 2005 6:33 AM
Brian K
Done at 2 metres. Same file, 323 seconds, calculates to 3.22 MB/sec, 25.8
Mbps. DU Meter average 27.9, peak 33.8 Mbps. (The DU Meter sat on 30-31 Mbps
most of the time but occasional gaps dropped the average) The transfer rate
on eyeballing DU Meter was the same whether the router was 1 up to 4 metres
from the card antenna.

I'ts looking better.

Brian
Author
14 Mar 2005 5:35 PM
JB
I assume you mean Super-G. There's Super-G and there's Afterburner, but not
Speedbooster. What model of router? What PC Card?

As for what I use, it's QCheck most of the time. See:

http://www.ixia.com

But the quick answer is - you have to enable Super-G mode in the router and
the compatible Super-G card, and if you don't, you will get 15.

- JB


Show quoteHide quote
"Brian K" <iibntgyea4 REMOVE_T***@hotmail.com> wrote in message
news:2ybZd.197337$K7.48131@news-server.bigpond.net.au...
> Thanks, but I was interested to know how JB did it. My transfer rate
> timing large file copying, using FTP, DU meter etc is only 15 Mbps with
> Wireless G Speedbooster. This is with two computers in the same room, a
> few metres from the router. I can't get it any higher.
>
> Brian
>
>
>
>>
>> Assorted benchmarking software:
>>
>> IPerf
>>  http://dast.nlanr.net/Projects/Iperf/
>>
>> Chariot
>>
>> http://www.ixiacom.com/products/performance_applications/pa_display.php?skey=pa_ixchariot&section=perftest
>>
>> Netstat Live
>>  http://www.analogx.com/contents/download/network/nsl.htm
>>
>> NetBench
>>  http://www.etestinglabs.com/benchmarks/netbench/default.asp
>>
>> Unix
>>  time (cat big_file | rcmd remote_machine "dd of=/dev/null")
>>
>> List of various benchmarking software from PC Magazine
>>  http://www.pcmag.com/category2/0,1738,1938,00.asp
>>
>> Or, you can just copy a known large file and time how long it takes to
>> copy.
>>
>>
>> --
>> Jeff Liebermann    je***@comix.santa-cruz.ca.us
>> 150 Felker St #D   http://www.LearnByDestroying.com
>> Santa Cruz CA 95060    AE6KS  831-336-2558
>
>
Author
14 Mar 2005 5:58 PM
Jeff Liebermann
On Mon, 14 Mar 2005 11:35:11 -0600, "JB"
<jbrandonbbrem***@hotmail.com> wrote:

>As for what I use, it's QCheck most of the time. See:
>http://www.ixia.com

Try:
http://www.ixiacom.com
instead.


--
Jeff Liebermann    je***@comix.santa-cruz.ca.us
150 Felker St #D   http://www.LearnByDestroying.com
Santa Cruz CA 95060    AE6KS  831-336-2558
Author
13 Mar 2005 5:03 PM
Peter Pan
Jeff & Kate Wooddell wrote:
> I have an HP laptop with built-in 802.11b wireless, but would like to
> upgrade to G or Super-G.  Can I just buy a notebook adapter for the
> faster protocol (along with a matching router)?  And would that
> override the built-in "b"?
>
> Jeff Wooddell

Absolutely yes you can, at least with the USB stuff. I have a fuji laptop
with built in wireless, and sometimes use the external USB WiFi device
(Linksys WUSB54GP). By doing the USB I have Regular Ethernet, built in wifi,
and the External USB all installed/active/bridged.
With some manufacturers models, if you install a cardbus card, it may/may
not superscede the built in stuff, so rather than say maybe, though I'd give
you an alternative that I can say yes to, and tell you how.
Author
13 Mar 2005 5:03 PM
Jonathan
Jeff & Kate Wooddell wrote:
> I have an HP laptop with built-in 802.11b wireless, but would like to
> upgrade to G or Super-G.  Can I just buy a notebook adapter for the faster
> protocol (along with a matching router)?  And would that override the
> built-in "b"?
>
> Jeff Wooddell
>
>

Notebook wireless cards are implemented as MiniPCI cards.  MiniPCI cards
have the same or similar electrical specifications as PCI cards, but the
formfactor, power requirements, and connectors differ.  They have
contacts along one edge and snap into a connector similar to a laptop
DIMM connector.
In most cases, these cards are fairly upgradeable.  Since your laptop
already has "b", it must have the built-in antenna for 2.4Ghz operation.
  So there's a good chance that a "b/g" MiniPCI card would pop right in,
replacing your "b" card, and all you'd have to do is VERY VERY CAREFULLY
disconnect the delicate antenna connector from the old card and connect
it to the new one.

Note that this will most likely violate your notebook's warranty, so
it's not recommended until the warranty expires.  Or, have it
professionally done.

Look around eBay for MiniPCI wireless cards, they're usually not too
hard to find.  If your warranty has expired, you can probably find the
card for < US$50 and would only take 15min to install.

good luck,
Jonathan