P2B-S (going back a bit mow)

[Gary Ridgway]

Has anybody heard of the following problem that I have found with my
P2B-s motherboard.

Tried to install a Belkin 54g PCI card into my PC, but it is not seen as
a plug and play card, in fact it isn't seen at all, not even in the PCI
listing on boot up.
The card is OK as far as I can see, as my dad's PC sees it straight away
and tries to find the drivers.

I have tried the motherboard with the P2B-S BIOS 1012 and the later beta
version.

Any ideas?

 

[Stephan Grossklass]

Has anybody heard of the following problem that I have found with my
P2B-s motherboard.

Tried to install a Belkin 54g PCI card into my PC, but it is not seen as
a plug and play card, in fact it isn't seen at all, not even in the PCI
listing on boot up.
The card is OK as far as I can see, as my dad's PC sees it straight away
and tries to find the drivers.

I have tried the motherboard with the P2B-S BIOS 1012 and the later beta
version.

Any ideas?

This card is apparently not compatible with older PCI 2.1 (and not 2.2
up) compliant boards.

Stephan

 

[Legend]

This card is apparently not compatible with older PCI 2.1 (and not 2.2
up) compliant boards.

Stephan

move it to another slot... maybe it demands a bus mastering slot and
so far has not had one

 

[Stephan Grossklass]

move it to another slot... maybe it demands a bus mastering slot and
so far has not had one

Any PCI 2.1 compliant board has *only* bus mastering capable slots! This
has not been an issue since the i430FX.

Stephan

 

[Nikolaos Tampakis]

Any PCI 2.1 compliant board has *only* bus mastering capable slots! This
has not been an issue since the i430FX.

Stephan

There were some 5-slot boards with onboard audio which when enabled
renders one slot (usually the fifth) as slave-only.
Examples that come to mind are the Asus P5A (Ali Aladin V) and the MSI
MS-6321 (Intel BX) rev. 2.0 IIRC.
Of course this doesn't make what you said wrong - the slots are still
perfectly bus-master capable, just not under all circumstances.
I think both examples use some ESS/ensoniq chip (1371 or 1373?) so this
might actually have to do with the audio chip selection, I don't really
know (but considering there exist BX boards with 6 PCI slots and other
onboard PCI devices too without such restrictions, even MSI's 3.0
version of the board for example, this ought not be a limitation of the
chipset).

Regards
Nikos

 

[Gary Ridgway]

There were some 5-slot boards with onboard audio which when enabled
renders one slot (usually the fifth) as slave-only.
Examples that come to mind are the Asus P5A (Ali Aladin V) and the MSI
MS-6321 (Intel BX) rev. 2.0 IIRC.
Of course this doesn't make what you said wrong - the slots are still
perfectly bus-master capable, just not under all circumstances.
I think both examples use some ESS/ensoniq chip (1371 or 1373?) so this
might actually have to do with the audio chip selection, I don't really
know (but considering there exist BX boards with 6 PCI slots and other
onboard PCI devices too without such restrictions, even MSI's 3.0
version of the board for example, this ought not be a limitation of the
chipset).

Regards
Nikos

So is it a PCI compliance issue then? I have tried all of the slots.

 

[Nikolaos Tampakis]

Gary Ridgway wrote:
....

So is it a PCI compliance issue then? I have tried all of the slots.

Could be. It gets identified as a network adapter I presume?
Those are nowadays conforming to PCI 2.2 in order to support PCI-based
wake up without the need for external WOL headers and such.
A PCI 2.2 slot will have a pin dedicated to this wake-up functionality,
a PCI 2.1 slot won't. If the card for some reason expects a certain
behaviour on this pin during reset I can see the possibility of a failed
initialisation. But this would most likely be a design flaw in the card.
If you feel adventurous dig through the 2.2 PCI pinout, find this
suspicious pin and insulate it with a tiny bit of appropriate tape
(well, insulate the card's contact to it anyway). Needless to say this
is just a blind shot with little reasoning beyond the 'it could have to
do with that pin'.

Regards
Nikos

 

[Gary Ridgway]

Gary Ridgway wrote:
...

Could be. It gets identified as a network adapter I presume?
Those are nowadays conforming to PCI 2.2 in order to support PCI-based
wake up without the need for external WOL headers and such.
A PCI 2.2 slot will have a pin dedicated to this wake-up functionality,
a PCI 2.1 slot won't. If the card for some reason expects a certain
behaviour on this pin during reset I can see the possibility of a
failed initialisation. But this would most likely be a design flaw in
the card.
If you feel adventurous dig through the 2.2 PCI pinout, find this
suspicious pin and insulate it with a tiny bit of appropriate tape
(well, insulate the card's contact to it anyway). Needless to say this
is just a blind shot with little reasoning beyond the 'it could have to
do with that pin'.

Regards
Nikos

There again, might just get another mb and chip, probably AMD for this
one. One question, do you need the p4 type power supply unit for the
AMD's, or will the std ATX PII be OK?

Thanks

 

[Nikolaos Tampakis]

Gary Ridgway wrote:
....

There again, might just get another mb and chip, probably AMD for this
one. One question, do you need the p4 type power supply unit for the
AMD's, or will the std ATX PII be OK?

Thanks

As for the plug pinout, most Athlon boards will use the standard ATX
20-pin connector (I assume by saying AMD you're not referring to
Athlon64/Opteron).
But there are more serious requirements on the PSU capacity than for a
PII board.
AMD used to recommmend a PSU that could deliver at least 20 Amperes on
the +3,3 range IIRC (back in the hot and burning 1+ GHz Thunderbird
days, but even with cooler more recent Athlons the requirements for the
PSU wouldn't be much lower, let alone if you opt for the fastest chips).
So the plug will likely be OK but double check the PSU requirements to
see if yours fits.
Also you should more carefully check the 5VSB ratings if you plan to use
stuff such as suspend to RAM, wake-on-ring/lan/modem, wake on USB, wake
on PCI card, wake on PS2 (keyboard/mouse) etc. Last time I checked the
requirement for complete compliance to all those was 2 Amperes on the
5VSB line (versus the absolute minimal 0,01 Amperes required for the
motherboard be able to power up).

Regards
Nikos