Warning: Netgear 802.11g can fry your machine

I recently purchased some NetGear equipment including an 802.11g card
WG311 which has specified on the box that it requires a minimum P3 and
PCI 2.2 slot. They're not kidding. I put it in my low powered Celeron
766MHz system and after about 5 minutes of booting up with the card
installed the whole box froze. Odd, because this machine has been
exceedingly reliable. I reseated everything, removed everything to see
if it would come back but alas no. Chip / RAM / Motherboard has died.

I then tried my faithful Pentium II 350MHz machine. Also a rock solid
machine that has faithfully run for years. It froze on initialising
PnP. I was very worried. I removed the WG311 and everything was fine.
My C766 is still dead.

Er. Yuk.

Spoke to the 24x7 help desk support (in India) and was told that they
did specify Pentium 3 or better, with PCI 2.2.

Of course that is on the box, not in the manual, nor is it in the
software when you pre-install the driver as indicated in the manual.

Not Happy

On Fri, 12 Dec 2003 14:14:05 GMT, (E-Mail Removed) (Simon) wrote:

>I recently purchased some NetGear equipment including an 802.11g card
>WG311 which has specified on the box that it requires a minimum P3 and
>PCI 2.2 slot. They're not kidding. I put it in my low powered Celeron
>766MHz system and after about 5 minutes of booting up with the card
>installed the whole box froze. Odd, because this machine has been
>exceedingly reliable. I reseated everything, removed everything to see
>if it would come back but alas no. Chip / RAM / Motherboard has died.
>
>I then tried my faithful Pentium II 350MHz machine. Also a rock solid
>machine that has faithfully run for years. It froze on initialising
>PnP. I was very worried. I removed the WG311 and everything was fine.
>My C766 is still dead.
>
>Er. Yuk.
>
>Spoke to the 24x7 help desk support (in India) and was told that they
>did specify Pentium 3 or better, with PCI 2.2.
>
>Of course that is on the box, not in the manual, nor is it in the
>software when you pre-install the driver as indicated in the manual.
>
>Not Happy

You broke something putting the card in and out.
There's nothing in PCI 2.2 that would cause a PCI 2.1 or even 2.0 to "fry".

A card requiring 2.2 capabilities may never work on PCI 2.1 or 2.0 system, but
it wouldn't cause a fatal *permanent* injury to the system.

A couple of possibilities: if your system supports standby, and you forgot to
either unplug the power cord completely, or flip the AC switch on the back of
the power supply (if so equipped), with live power still on the system board
and you plugging things in and out, perhaps something went to Heaven; or you
dislodged something and have yet to lodge it again ;-)

/daytripper

/daytripper

Daytripper, I'd like to learn the different PCI specs in order
to be able to address questions like this. Where did you learn
the information used in your reply?

Ton.


daytripper <(E-Mail Removed)> wrote in message news:<(E-Mail Removed)>. ..
>
> There's nothing in PCI 2.2 that would cause a PCI 2.1 or even 2.0 to "fry".
>
> A card requiring 2.2 capabilities may never work on PCI 2.1 or 2.0 system, but
> it wouldn't cause a fatal *permanent* injury to the system.
>

On 20 Dec 2003 21:46:45 -0800, (E-Mail Removed) wrote:

>
>daytripper <(E-Mail Removed)> wrote in message news:<(E-Mail Removed)>. ..
>>
>> There's nothing in PCI 2.2 that would cause a PCI 2.1 or even 2.0 to "fry".
>>
>> A card requiring 2.2 capabilities may never work on PCI 2.1 or 2.0 system, but
>> it wouldn't cause a fatal *permanent* injury to the system.
>>

>Daytripper, I'd like to learn the different PCI specs in order
>to be able to address questions like this. Where did you learn
>the information used in your reply?
>
>Ton.

I've been a PCI-SIG participant and reviewer for two different companies over
the last 13 years, and have done numerous PCI designs, including host (aka
"north") bridges, agents, and boards.

A quicker and likely easier way would be to buy a book (though not Solari's -
an unmitigated soup sandwich)...

cheers

/daytripper

daytripper <(E-Mail Removed)> wrote in message news:<(E-Mail Removed)>. ..
>
> ...buy a book...
>
> cheers
>
> /daytripper

Thanks daytripper, but I'm looking for a less indepth
level of understanding.
I'm a PC tech ( Wintel build and fix ).
I suppose that all I really need to know is
"what does the motherboard support" and
"what does the adapter require".
But I'm curious to know a bit more.
Do you know of any free web-based sources?

A question: What would be the reason for a PCI 2.2
adapter to require a PCI 2.2 mobo.That is, what prevents
some adapters from being backwards compatible with 2.1 or 2.0?

Ton.

On 21 Dec 2003 16:11:46 -0800, (E-Mail Removed) wrote:

>daytripper <(E-Mail Removed)> wrote in message news:<(E-Mail Removed)>. ..
>>
>> ...buy a book...
>>
>> cheers
>>
>> /daytripper
>
>Thanks daytripper, but I'm looking for a less indepth
>level of understanding.
>I'm a PC tech ( Wintel build and fix ).
>I suppose that all I really need to know is
>"what does the motherboard support" and
>"what does the adapter require".
>But I'm curious to know a bit more.
>Do you know of any free web-based sources?

Not really, no. Never had the need, you understand, I started with the first
draft PCI spec, there was nothing else.

>A question: What would be the reason for a PCI 2.2
>adapter to require a PCI 2.2 mobo.That is, what prevents
>some adapters from being backwards compatible with 2.1 or 2.0?

Delayed Transaction was the biggy with 2.2, followed by refinement of
autoconfiguration rules (for 64b support, etc). There were other less
interesting changes, like redefining latency from RST# deassertion,
restricting the use of LOCK# to bridges, clarification of master behavior
during a target disconnect, and a lot of clarifications of 2.1 changes that
were botched during editing...

/daytripper