Netgear WG511v2 doesn't measure up to WG511 v1

The original Netgear WG511 (v1) PC Card is one of the best 802.11g products
I've ever used or tested. It has remarkably good range (better than most
other 802.11b and 802.11g products I've used and tested), very good stability
(better able to hold a difficult connection than other products), and very
good speed (particularly in less than optimal conditions).

I recently needed an additional 802.11g PC Card, and when I couldn't readily
find an original WG511 (v1) PC Card, I went with the WG511v2 (which is a
different design). My expectation was that it would be at least as good as
the original WG511 (v1), and perhaps better. Unfortunately, it isn't even as
good:

* Range is significantly worse than the original WG511 (v1), on the order of
10-20% less. In some locations where the original WG511 (v1) can connect and
hold a weak signal, the WG511v2 is too unstable to be usable

* Stability of the WG511v2 is much worse than the original WG511 (v1), to the
point of losing connections even in strong signal conditions.

These observations are based on extensive careful testing of more than one
sample of both types of cards in multiple locations with the latest Netgear
drivers under Windows XP SP2 on an IBM ThinkPad T30.

Accordingly, my recommendation is for the original WG511 (v1) over the
WG511v2, which didn't even perform as well as other (non-Netgear) 802.11g
cards I've used and tested. Easy way to tell them apart:

* The original WG511 (v1) has a silver antenna bulge.

* The WG511v2 has a black antenna bulge.

See also <http://kbserver.netgear.com/products/WG511.asp>.

Although the original WG511 (v1) may be hard to find new, good used ones are
readily available on eBay (as of this writing) at bargain prices.

--
Best regards, HELP FOR CINGULAR GSM & SONY ERICSSON PHONES:
John Navas <http://navasgrp.home.att.net/#Cingular>

On Sat, 08 Oct 2005 11:37:10 GMT, John Navas
<(E-Mail Removed)> wrote:

So much for new and improved.

>* Range is significantly worse than the original WG511 (v1), on the order of
>10-20% less.

A bit of math in antenna gain, tx power, and rx sensitivity.
Rounding off a bit.
12dB is 25% of your range.
6dB is 50%
3dB is 75%
1.5dB is 87.5%
Therefore, for you 20% loss in range, there is about a 1dB difference
between the v1 and v2 incantation. I have no way to tell if it's in
receive sensitivity, transmit power, or antenna gain. My guess(tm)
would be antenna gain. 1-3 dB variations in product lots are all too
common.

>In some locations where the original WG511 (v1) can connect and
>hold a weak signal, the WG511v2 is too unstable to be usable

Different chipsets have different algorithms for dealing with
reflections. If the weak signal is the result of going through walls
or floors, you're really measuring how well the receiver can handle
multipath. How well does it work in a line of sight and minimal
reflective environment?

The v1 to v2 transition suggests a change of vendor so that two cards
may be completely different inside. Could I trouble you for the FCC
ID numbers so I can look at the inside photos on the FCC web pile?

>* Stability of the WG511v2 is much worse than the original WG511 (v1), to the
>point of losing connections even in strong signal conditions.

Ouch. Now, that's bad. When that happens, check whatever utility is
supplied with the card for the connection speed. If it has slowed
down to 1-2Mbit/sec, then the card and/or it's driver is doing a very
bad job of handling interference. They're suppose to slow down or
stop in the presence of interference, then speed back up when it goes
away. Some drivers do not recover very gracefully. If the speed
stays fairly high, then that's not the problem.

>These observations are based on extensive careful testing of more than one
>sample of both types of cards in multiple locations with the latest Netgear
>drivers under Windows XP SP2 on an IBM ThinkPad T30.

What access points or wireless routers were you using? Just curious
if you're mixing chipsets.

>Accordingly, my recommendation is for the original WG511 (v1) over the
>WG511v2, which didn't even perform as well as other (non-Netgear) 802.11g
>cards I've used and tested. Easy way to tell them apart:
>
> * The original WG511 (v1) has a silver antenna bulge.
>
> * The WG511v2 has a black antenna bulge.
>
>See also <http://kbserver.netgear.com/products/WG511.asp>.
>
>Although the original WG511 (v1) may be hard to find new, good used ones are
>readily available on eBay (as of this writing) at bargain prices.

--
Jeff Liebermann (E-Mail Removed)
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

[POSTED TO alt.internet.wireless - REPLY ON USENET PLEASE]

In <(E-Mail Removed)> on Sat, 08 Oct 2005 07:05:15
-0700, Jeff Liebermann <(E-Mail Removed)> wrote:

>On Sat, 08 Oct 2005 11:37:10 GMT, John Navas
><(E-Mail Removed)> wrote:
>
>So much for new and improved.

My guess is new and cheaper (in the negative sense of the word).

>>* Range is significantly worse than the original WG511 (v1), on the order of
>>10-20% less.
>
>A bit of math in antenna gain, tx power, and rx sensitivity.
>Rounding off a bit.
> 12dB is 25% of your range.
> 6dB is 50%
> 3dB is 75%
>1.5dB is 87.5%
>Therefore, for you 20% loss in range, there is about a 1dB difference
>between the v1 and v2 incantation. I have no way to tell if it's in
>receive sensitivity, transmit power, or antenna gain. My guess(tm)
>would be antenna gain. 1-3 dB variations in product lots are all too
>common.

These results were consistent with two different samples of each card.
(That's not statistically valid, but better than just one of each.)

>>In some locations where the original WG511 (v1) can connect and
>>hold a weak signal, the WG511v2 is too unstable to be usable
>
>Different chipsets have different algorithms for dealing with
>reflections. If the weak signal is the result of going through walls
>or floors, you're really measuring how well the receiver can handle
>multipath. How well does it work in a line of sight and minimal
>reflective environment?

These results are consistent over multiple environments, both with and without
walls and floors.

>The v1 to v2 transition suggests a change of vendor so that two cards
>may be completely different inside. Could I trouble you for the FCC
>ID numbers so I can look at the inside photos on the FCC web pile?

The cards I currently have:
v1: PY3WG511-F
v2: PY3WG511V2H1

Google suggests that there are at least 5 versions of this card:
PY3WG511 Intersil
PY3WG511-F Intersil
PY3WG511V2H1 Marvell
PY3WG511V3 Atheros
PY3WG511TV1H3 Atheros
That last is probably the WG511T, a distinctly different Netgear product, and
I'm thinking the prior one may be as well, leaving the first three as members
of the base WG511 family, with two chip vendors. My guess is that the Marvell
implementation is cheaper but not as good as the Intersil implementation.

>>* Stability of the WG511v2 is much worse than the original WG511 (v1), to the
>>point of losing connections even in strong signal conditions.
>
>Ouch. Now, that's bad. When that happens, check whatever utility is
>supplied with the card for the connection speed. If it has slowed
>down to 1-2Mbit/sec, then the card and/or it's driver is doing a very
>bad job of handling interference. They're suppose to slow down or
>stop in the presence of interference, then speed back up when it goes
>away. Some drivers do not recover very gracefully. If the speed
>stays fairly high, then that's not the problem.

The speed stays fairly high (36+). Signal strength at the PC Card is 6-8
balls out of 8 on the Netgear monitor.

There are two access points in the area (inside a public library) with
overlapping coverage, roughly the same signal strength, and the same SSID.
There are no other obvious sources of interference, not even any other
apparent wireless clients when I was testing. My guess is that the card is
sporadically trying unsuccessfully to switch access points (roam).

>>These observations are based on extensive careful testing of more than one
>>sample of both types of cards in multiple locations with the latest Netgear
>>drivers under Windows XP SP2 on an IBM ThinkPad T30.
>
>What access points or wireless routers were you using? Just curious
>if you're mixing chipsets.

Other than the D-Link home network, all the others are unknown hotspot
hardware.

>>Accordingly, my recommendation is for the original WG511 (v1) over the
>>WG511v2, which didn't even perform as well as other (non-Netgear) 802.11g
>>cards I've used and tested. Easy way to tell them apart:
>>
>> * The original WG511 (v1) has a silver antenna bulge.
>>
>> * The WG511v2 has a black antenna bulge.
>>
>>See also <http://kbserver.netgear.com/products/WG511.asp>.
>>
>>Although the original WG511 (v1) may be hard to find new, good used ones are
>>readily available on eBay (as of this writing) at bargain prices.

--
Best regards, HELP FOR CINGULAR GSM & SONY ERICSSON PHONES:
John Navas <http://navasgrp.home.att.net/#Cingular>

John Navas <(E-Mail Removed)> wrote:
> These results were consistent with two different samples of each card.
> (That's not statistically valid, but better than just one of each.)

That sounds like a small anecdotal observation.

---
Clarence A Dold - Hidden Valley (Lake County) CA USA 38.8,-122.5

[POSTED TO alt.internet.wireless - REPLY ON USENET PLEASE]

In <di9hab$32l$(E-Mail Removed)> on Sat, 8 Oct 2005 22:30:35 +0000 (UTC),
(E-Mail Removed) wrote:

>John Navas <(E-Mail Removed)> wrote:
>> These results were consistent with two different samples of each card.
>> (That's not statistically valid, but better than just one of each.)
>
>That sounds like a small anecdotal observation.

As I wrote. Notwithstanding a great deal of testing and documentation. So do
you have anything substantive to add (e.g., evidence one way or the other), or
(with apologies to Peggy Lee) is that all there is?

--
Best regards, HELP FOR CINGULAR GSM & SONY ERICSSON PHONES:
John Navas <http://navasgrp.home.att.net/#Cingular>

John Navas <(E-Mail Removed)> wrote:

> As I wrote. Notwithstanding a great deal of testing and documentation. So do
> you have anything substantive to add (e.g., evidence one way or the other), or
> (with apologies to Peggy Lee) is that all there is?

I just wonder why your commentary is substantial, and other sources are
considered anecdotal.

--
---
Clarence A Dold - Hidden Valley (Lake County) CA USA 38.8,-122.5

John Navas <(E-Mail Removed)> wrote:
> The original Netgear WG511 (v1) PC Card is one of the best 802.11g products
> I've ever used or tested. It has remarkably good range (better than most

I've had a couple of the WG511, I presume they are V1, since they don't say
v2. These are very good cards, and are often very cheap on Amazon. I
bought the first because it was only $19 after rebates. I bought the
second after I found out how good the first was. I think it is a bit
better than an Orinoco Silver Classic.

> * The original WG511 (v1) has a silver antenna bulge.
> * The WG511v2 has a black antenna bulge.

Netgear uses their own FCC-ID range, so it's difficult to pinpoint the
original manufacturer, but could you post the new FCCID?

Mine is FCC ID PY3WG511-F. I identified it as an Intersil chipset.

--
---
Clarence A Dold - Hidden Valley (Lake County) CA USA 38.8,-122.5

John Navas <(E-Mail Removed)> wrote:
> The original Netgear WG511 (v1) PC Card is one of the best 802.11g
> products I've ever used or tested.

> * Stability of the WG511v2 is much worse than the original WG511 (v1), to
> the point of losing connections even in strong signal conditions.

> Although the original WG511 (v1) may be hard to find new, good used ones
> are readily available on eBay (as of this writing) at bargain prices.

Amazon (where I bought mine), still shows the gray antenna bulge.
Today it's $29.99 after rebates.
Fry's Outpost.com shows the gray bulge for $49.99 after rebates.

Actually, I can't find an online source that shows the black bulge. Maybe
the photos aren't updated when the version changes.

The commentary in some different postings about the v2 speaks of a Marvel
chip or a Broadcom chip. One even laments that the v2 is being sold
without distinction from the v1, but doesn't say where that purchase was
made.


Where did you get yours? Was it obvious it was a v2 when you bought it?

--
---
Clarence A Dold - Hidden Valley (Lake County) CA USA 38.8,-122.5

On Sat, 08 Oct 2005 20:43:49 GMT, John Navas
<(E-Mail Removed)> wrote:

[Netgear WG511 v1 and v2]

>The cards I currently have:
>v1: PY3WG511-F
| https://gullfoss2.fcc.gov/prod/oet/cf/eas/reports/ViewExhibitReport.cfm?mode=Exhibits&RequestTimeout =500&calledFromFrame=N&application_id=984128&fcc_i d='PY3WG511-F'
| https://gullfoss2.fcc.gov/prod/oet/f...ive_or_pdf=pdf

Looks like a clone of the Intersil/Harris ISL38001C Prism GT Cardbus32
reference design. It was formerly type certified by
Rockwell/Intersil/Conexant as OSZ38001C at:
| https://gullfoss2.fcc.gov/prod/oet/cf/eas/reports/ViewExhibitReport.cfm?mode=Exhibits&RequestTimeout =500&calledFromFrame=N&application_id=655680&fcc_i d='OSZ38001C'
Looks exactly the same as the Netgear WG511v1. I can't tell who
manufactured it.

>v2: PY3WG511V2H1

| https://gullfoss2.fcc.gov/prod/oet/cf/eas/reports/ViewExhibitReport.cfm?mode=Exhibits&RequestTimeout =500&calledFromFrame=N&application_id=189923&fcc_i d='PY3WG511V2H1'
| https://gullfoss2.fcc.gov/prod/oet/f...ive_or_pdf=pdf

Marvell chipset. Made by Cameo Communications Inc in Taiwan.
| http://www.cameo.com.tw/products/htm...s/wlg-1100.htm

Note the color of the circuit board materials. The yellow color of
the Intersil version is low loss polysulfone. The green of the Cameo
version is G10/FR4 which has considerably higher loss. There's where
the RF got lost in the v2 version.

>Google suggests that there are at least 5 versions of this card:
> PY3WG511 Intersil
> PY3WG511-F Intersil
> PY3WG511V2H1 Marvell
> PY3WG511V3 Atheros
> PY3WG511TV1H3 Atheros
>That last is probably the WG511T, a distinctly different Netgear product, and
>I'm thinking the prior one may be as well, leaving the first three as members
>of the base WG511 family, with two chip vendors.

Incidentally, the reason they don't change the part number is to avoid
having massive dealer returns for obsolete stock. If the new version
is "functionally identical" to the original, with the same part
number, then the dealer is stuck with the "old" stock. Usually, the
new version is better in some way, but this time, it's the other way
around.

>My guess is that the Marvell
>implementation is cheaper but not as good as the Intersil implementation.

I agree. The v1 Intersil looks like a much better and substantially
more expensive to produce. Polysulfone falls apart when overheated
and is a pain to solder.

--
Jeff Liebermann (E-Mail Removed)
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

[POSTED TO alt.internet.wireless - REPLY ON USENET PLEASE]

In <di9siv$fel$(E-Mail Removed)> on Sun, 9 Oct 2005 01:42:55 +0000 (UTC),
(E-Mail Removed) wrote:

>John Navas <(E-Mail Removed)> wrote:
>
>> As I wrote. Notwithstanding a great deal of testing and documentation. So do
>> you have anything substantive to add (e.g., evidence one way or the other), or
>> (with apologies to Peggy Lee) is that all there is?
>
>I just wonder why your commentary is substantial, and other sources are
>considered anecdotal.

I think that's obvious, but then again, YMMV.

--
Best regards, HELP FOR CINGULAR GSM & SONY ERICSSON PHONES:
John Navas <http://navasgrp.home.att.net/#Cingular>