P4P v P4V v P4S?

  • Thread starter Thread starter PL
  • Start date Start date
P

PL

I've got a 3.0GHz pentium (prescott) cpu on a P4P800SE, which fails to
boot. With no other way of checking things out, I'm looking to buy a
cheap Celeron CPU and mainboard. I'm sorted with the CPU, but would be
interested for feedback on whether to consider a type P4V v P4S
mainboard if only to avoid the well-documented Intel latchup problem?

What do you reckon?
 
I have the same MB and just installed a 3.2 800fsb chip all is well for the
moment.
I did change my memory to 3200 for the 800fsb. I do have a good PS.
Also a 6600gt video card.
The reason I selected this board was because of the video problems with the
sis and via
chipset with high end video cards and games like Doom3 and HL2.
 
I have a P4P800 SE with P4 3.2 Ghz 2GB RAM.
It is normal, for me, to have boot problem.
I have always (90%) to press the HW reset button to boot.
It seems to be a PSU problem, I have changed my 300W PSU with a 550W PSU but
nothing is changed.
At the momentI have no idea, someone is advising me to change an other time
my PSU with an other kind of PSU but for now I'm waiting.

Good luck and let me know if you are able to solve the problem

Alberto
 
Alb said:
I have a P4P800 SE with P4 3.2 Ghz 2GB RAM.
It is normal, for me, to have boot problem.
I have always (90%) to press the HW reset button to boot.
It seems to be a PSU problem, I have changed my 300W PSU with a 550W PSU but
nothing is changed.
At the momentI have no idea, someone is advising me to change an other time
my PSU with an other kind of PSU but for now I'm waiting.

Good luck and let me know if you are able to solve the problem

Alberto

Today I set the PC up and after 5 (hard) re-boots, it finally got
working. There is one post on the ASUS forum about this 'problem', but
no explanation/solution.

I have a 500w psu that according to ASUS PCProbe outputs as follows:

+3.3v @ +3.44v
+12v @ 12.22v
+5v @ +5.16v
Vcore @ 1.404v
CPU temp 37 C

Does anyone see a problem here?
 
PL <"neatripple AT said:
Today I set the PC up and after 5 (hard) re-boots, it finally got
working. There is one post on the ASUS forum about this 'problem', but
no explanation/solution.

I have a 500w psu that according to ASUS PCProbe outputs as follows:

+3.3v @ +3.44v
+12v @ 12.22v
+5v @ +5.16v
Vcore @ 1.404v
CPU temp 37 C

Does anyone see a problem here?

Percentage wise, your voltages are still in the "normal" range,
although I'd want to verify the voltages with a multimeter.

A failure to POST could have any number of root causes. One
reason is if the Vcore regulator circuit abruptly shuts down
when the system is powered up. One model of motherboard had
issues with Antec Truepower supplies, and the problem was
corrected in subsequent production by changing some component
value. Otherwise, it could be just about any component sitting
on a system bus, that could throw a wrench in the works.

If you had just CPU, memory, video card, keyboard/mouse, does
it work any better ? Try testing with just one DIMM at a time.

If you switch power supplies, try using a different brand of
supply.

In terms of gathering information, if the motherboard had
Vocal POST, you could listen for any error message, as that
might hint at the source of the problem. There is also a
debugging card you can get, that plugs into a PCI slot, and
it intercepts writes to port 80, and displays them on a
two digit LED display. The BIOS companies have lists of
which BIOS routine is running, when a particular number
is being displayed. A port 80 POST card used to cost $100,
but some Ebay sellers have sold them for $20, as they really
don't have a lot of expensive chips on the card.

Here is an example of a diagnostic PCI POST card:
http://cgi.ebay.com/PC-PCI-Motherbo...6791398721QQcategoryZ1244QQrdZ1QQcmdZViewItem

The main value of a PCI POST card, is it can show you whether
the processor got to execute any code or not. Another device
that provides a little info like that, is the LED displays
on some expensive memories. A failure to see any acivity on
the DIMM's LED display can be used to imply that the BIOS code
never tried to initialize memory.

In terms of buying a good power supply for your computer, you
should be buying them based on the rating information printed
on the label on the side. Many of the cheaper supplies give
an exaggerated total power rating for the supply - if,
for example, the supply doesn't offer a minimum of 12V@15A
output, I would expect problems with a P4 based system. There
were some old 500W supplies (when 500 watters first came out),
that only offered 12V@10A. Such supplies were only usable on
AthlonXP systems, where the +5V was the main source of
motherboard power. On my P4C800-E Deluxe, using four DIMMs,
my measured consumption is [email protected] amps, [email protected] amps, and
12V@6amps. This does not include the video card, or the disk
drives. In my case, it is meant to show that not much +5V is
used, a moderate amount of 3.3V, and the +12V can be calculated
based on the model of processor being used, at a site like this:

http://takaman.jp/D/?english

I only have a 2.8GHz Northwood, so don't use that much +12V.

Paul
 
PL <"neatripple AT said:
I've got a 3.0GHz pentium (prescott) cpu on a P4P800SE, which fails to
boot. With no other way of checking things out, I'm looking to buy a
cheap Celeron CPU and mainboard. I'm sorted with the CPU, but would be
interested for feedback on whether to consider a type P4V v P4S
mainboard if only to avoid the well-documented Intel latchup problem?

What do you reckon?

All motherboards have some weakness or unsavory behavior. Until
the latchup problem appeared, I would recommend the Intel chipsets.
If Intel was out of the question, I might try an SIS board next.
Newegg has a couple models in the $50 range, which might be suitable
as a testing platform. As always, use Google and some other search
engine, and search based on the motherboard model number, or the
part number of the chipset chips (Northbridge and Southbridge), to
see if there are any issues with the board you have in mind.

An alternative solution, is buy an Intel chipset based board,
a separate PCI USB card for your USB peripherals, and a PS/2
keyboard (so you can access the BIOS screens). That will remove
most of the apparent risk from using an ICH4 or ICH5 southbridge.
If you don't use the motherboard USB ports or the motherboard USB
headers, that could reduce the latchup risk by a factor of 10 or
so.

Paul
 
Paul said:
PSU but



Percentage wise, your voltages are still in the "normal" range,
although I'd want to verify the voltages with a multimeter.

A failure to POST could have any number of root causes. One
reason is if the Vcore regulator circuit abruptly shuts down
when the system is powered up. One model of motherboard had
issues with Antec Truepower supplies, and the problem was
corrected in subsequent production by changing some component
value. Otherwise, it could be just about any component sitting
on a system bus, that could throw a wrench in the works.

If you had just CPU, memory, video card, keyboard/mouse, does
it work any better ? Try testing with just one DIMM at a time.

If you switch power supplies, try using a different brand of
supply.

In terms of gathering information, if the motherboard had
Vocal POST, you could listen for any error message, as that
might hint at the source of the problem. There is also a
debugging card you can get, that plugs into a PCI slot, and
it intercepts writes to port 80, and displays them on a
two digit LED display. The BIOS companies have lists of
which BIOS routine is running, when a particular number
is being displayed. A port 80 POST card used to cost $100,
but some Ebay sellers have sold them for $20, as they really
don't have a lot of expensive chips on the card.

Here is an example of a diagnostic PCI POST card:
http://cgi.ebay.com/PC-PCI-Motherbo...6791398721QQcategoryZ1244QQrdZ1QQcmdZViewItem

The main value of a PCI POST card, is it can show you whether
the processor got to execute any code or not. Another device
that provides a little info like that, is the LED displays
on some expensive memories. A failure to see any acivity on
the DIMM's LED display can be used to imply that the BIOS code
never tried to initialize memory.

In terms of buying a good power supply for your computer, you
should be buying them based on the rating information printed
on the label on the side. Many of the cheaper supplies give
an exaggerated total power rating for the supply - if,
for example, the supply doesn't offer a minimum of 12V@15A
output, I would expect problems with a P4 based system. There
were some old 500W supplies (when 500 watters first came out),
that only offered 12V@10A. Such supplies were only usable on
AthlonXP systems, where the +5V was the main source of
motherboard power. On my P4C800-E Deluxe, using four DIMMs,
my measured consumption is [email protected] amps, [email protected] amps, and
12V@6amps. This does not include the video card, or the disk
drives. In my case, it is meant to show that not much +5V is
used, a moderate amount of 3.3V, and the +12V can be calculated
based on the model of processor being used, at a site like this:

http://takaman.jp/D/?english

I only have a 2.8GHz Northwood, so don't use that much +12V.

Paul

The PC is already stripped down to a minimum with just CPU, one stick of
memory, video card, HDD, floppy, mouse and keyboard.

The power supply is not a 'recognised' brand, but delivers 17A@12V on
the printed label.

The PC will finally boot but only after several 'on-off' attempts. When
it's in it's 'tempermental mode' the mobo green light illuminates, all
fans spin, but there's no HD activity and no monitor signal. Once it's
does start up though (gets to the OS), it will happily shut down and
restart normally.

This gives me the impression of working properly once it's 'warmed up'.
Is this likely?

I've seen a few posts about this on the ASUS forum and the suggestions
are always the same - strip down to a minimum configuration and start
re-building. One poster suggested that on a certain board, there was a
blob of excessive solder around the CPU, which was shorting out and
causing intermitant start up problems.

I'll wait to see what happens tomorrow. If it's the same trouble to
start up, I'll strip out completely (including board from case) and
re-build. Having said that I've already swapped out just about
everything (case, psu, video, HDD, memory) and so far it hasn't sorted
what has now become a real PITA.
 
PL <"neatripple AT said:
The PC is already stripped down to a minimum with just CPU, one stick of
memory, video card, HDD, floppy, mouse and keyboard.

The power supply is not a 'recognised' brand, but delivers 17A@12V on
the printed label.

The PC will finally boot but only after several 'on-off' attempts. When
it's in it's 'tempermental mode' the mobo green light illuminates, all
fans spin, but there's no HD activity and no monitor signal. Once it's
does start up though (gets to the OS), it will happily shut down and
restart normally.

This gives me the impression of working properly once it's 'warmed up'.
Is this likely?

I've seen a few posts about this on the ASUS forum and the suggestions
are always the same - strip down to a minimum configuration and start
re-building. One poster suggested that on a certain board, there was a
blob of excessive solder around the CPU, which was shorting out and
causing intermitant start up problems.

I'll wait to see what happens tomorrow. If it's the same trouble to
start up, I'll strip out completely (including board from case) and
re-build. Having said that I've already swapped out just about
everything (case, psu, video, HDD, memory) and so far it hasn't sorted
what has now become a real PITA.

Most likely the fault is either with the motherboard, or with the
power supply. Based on my limited observation of failure rates in
this group, I would replace the supply first, as motherboard
failures don't happen that often. (The latchup problem being an
exception, as that is a known weakness. I have no way of knowing
what percentage of boards are affected, as there is little in the
way of official recognition of that problem.)

I think the "solder blob under the processor socket" was an early
P4C800 Deluxe problem. Only the initial batch of boards were affected,
which could be a few thousand. In this picture, you can see how the
lower right hand lump of solder is bridging to an adjacent conductor.
This fault was causes by the order of assembly of the motherboard,
and the manufacturing process would have to be altered to fix this.

http://web.archive.org/web/20040227200251/http://koti.mbnet.fi/~nightops/eki/DSC00249.JPG

Here is a sample thread, where somebody goes through the same
debugging process you are. The only step they missed, was testing
the motherboard outside the computer case (the so-called "cardboard
test"). The reason for doing a cardboard test, is to eliminate
an accidental short to the bottom of the motherboard, as the source
of the problem. There are a few motherboards that don't have the
complete hole pattern, and if you remove the old motherboard from
your computer case, then install the new one, one of the old
brass standoffs no longer lines up with a motherboard hole. The
brass standoff can short to the bottom of the motherboard. On
an early A7N8X, this causes one channel of audio to fail.

http://www.sharkyforums.com/showthread.php?p=2160258

Paul
 
I'll bet if you disconnect your hard drive and boot from a floppy or cdrom
your problem may go away and if it does then it is your power supply
 
BigJim said:
I'll bet if you disconnect your hard drive and boot from a floppy or cdrom
your problem may go away and if it does then it is your power supply
I disconnected the hard drive and no floppy seek on boot-up, nothing,
nadda...
 
Paul said:
Most likely the fault is either with the motherboard, or with the
power supply. Based on my limited observation of failure rates in
this group, I would replace the supply first, as motherboard
failures don't happen that often. (The latchup problem being an
exception, as that is a known weakness. I have no way of knowing
what percentage of boards are affected, as there is little in the
way of official recognition of that problem.)

I think the "solder blob under the processor socket" was an early
P4C800 Deluxe problem. Only the initial batch of boards were affected,
which could be a few thousand. In this picture, you can see how the
lower right hand lump of solder is bridging to an adjacent conductor.
This fault was causes by the order of assembly of the motherboard,
and the manufacturing process would have to be altered to fix this.

http://web.archive.org/web/20040227200251/http://koti.mbnet.fi/~nightops/eki/DSC00249.JPG

Here is a sample thread, where somebody goes through the same
debugging process you are. The only step they missed, was testing
the motherboard outside the computer case (the so-called "cardboard
test"). The reason for doing a cardboard test, is to eliminate
an accidental short to the bottom of the motherboard, as the source
of the problem. There are a few motherboards that don't have the
complete hole pattern, and if you remove the old motherboard from
your computer case, then install the new one, one of the old
brass standoffs no longer lines up with a motherboard hole. The
brass standoff can short to the bottom of the motherboard. On
an early A7N8X, this causes one channel of audio to fail.

http://www.sharkyforums.com/showthread.php?p=2160258

Paul

Switched on this morning and... nothing. I'm going to RMA both the
mainboard and the PSU.
 
Well it is 4 times (a lot...) that the boot is ok.
The tricks was : press and release quickly the start button.
I tried it because once I keept the start button down for a while and I have
had an overclockig error (??!!??) (I'm not using overclock)
It is strange .... but seems to work

Alb
 
Back
Top