What's the minimum required for a mbo to post?

What is required to make a beep?

Processor (and fan unless you're living dangerously)
Power supply
Stick of memory

Anything else? Is a video card required?

I'm still having my crazy problem in which my P5K Deluxe system won't
post. It's getting worse. I cannot believe the problem is my power
supply as I have two power supplies and when I switch them out the
problem continues unchanged. I even RMA'd one of them and it came back
untouched -- they could find no problem.

I want to strip the system down to the barest essentials and add things
one at a time to figure out what makes it fail. Problem is, failure is
intermittent so this problem is a tough one to pin down.

--
Bill Anderson

I am the Mighty Favog

Bill Anderson wrote:
What is required to make a beep?

Processor (and fan unless you're living dangerously)
Power supply
Stick of memory

Anything else? Is a video card required?

I'm still having my crazy problem in which my P5K Deluxe system won't
post. It's getting worse. I cannot believe the problem is my power
supply as I have two power supplies and when I switch them out the
problem continues unchanged. I even RMA'd one of them and it came back
untouched -- they could find no problem.

I want to strip the system down to the barest essentials and add things
one at a time to figure out what makes it fail. Problem is, failure is
intermittent so this problem is a tough one to pin down.


If you have a processor and a motherboard, the processor executes the
BIOS code on the motherboard. If the BIOS code does a check for memory
or a video card, and they're missing, then the BIOS will attempt to make
a beep code, using the system speaker.

If memory is present, and the locations below 640K are bad, in fact the
BIOS may not be able to beep in that case. Which is why you should
remove the memory, if attempting to verify it can beep at you.

So that means, processor+HS/fan, motherboard, power supply, case speaker,
front power switch to start the system, might be enough.

A bad motherboard can be intermittent. For example, if a large chip were to
have a cracked solder ball, that could make and break contact, when the
motherboard is flexed.

If, when installing the motherboard in the system, there was a metal standoff
present, where there is no hole for it to line up with, then that standoff
could short to something. On one particular Asus motherboard, when that happens,
an audio channel winks out. So something as trivial, as not checking standoffs,
before putting a new motherboard in an old system case, can cause problems like
that.

That kind of thing, is one reason why eventually a "cardboard test" might be
suggested to you. You can put the motherboard on your kitchen table, along
with the power supply. Place a telephone book, with cardboard cover, underneath
the motherboard as a base. You can then test the system there. I've built
up entire systems on a kitchen table, installed Windows and tested
components like that. The biggest risk by doing so, is the video card is
unsupported in the slot it is plugged into. If something tugs on the video cable,
it can be game over. That is not the kind of experiment to do, with kids around.
You need a quiet work area, where the gear won't get disturbed.

Paul

I dont know much about them, but on dealextreme are pci boards testers
with a numeric panel that shows you the error number wich (?) can tell
you what is failing

http://www.dealextreme.com/details.dx/sku.33305

Hope it helps.

On Sat, 20 Feb 2010 22:08:15 -0500, Paul wrote:

So that means, processor+HS/fan, motherboard, power supply, case speaker,
front power switch to start the system, might be enough.

A bad motherboard can be intermittent. For example, if a large chip were to
have a cracked solder ball, that could make and break contact, when the
motherboard is flexed.

snip good advice

I've just had that with a Windows Home Server I'm trying to build on a
K8V-XE mobo I was recently given (and asked about here).

It came with a 2600+ Semp and I stuck a G of RAM in it and an old PCI
video card and initially (days) it all seemed fine.

Then I added a Gb Ethernet card and noticed WHS wouldn't Hibernate.
This (apparently) can happen if there is a video issue and there (now)
was. The PCI video card was having an issue with the LAN card, then
the video went funny on it's own then the PC wouldn't boot? (following
the thread, I don't think this board issues any BOST beeps when there
is no RAM).

I tried a PCIe video card and thought that was it (as it all came up
again) and then with other plain PCI cards and it did not. I replaced
the PSU (just because) and then it came up again on a plain PCI card?

Every time it doesn't come up it doesn't complete the POST and I don't
know if it was a conflict between the LAN and Video cards, it's
screwed the mobo or it's just an intermittent problem (I don't know
the history of the board).

The WHS is currently running on an Atom 330 board and is ok but I was
enjoying playing with different CPU's (both the Turion and 3000+ Semp
support Cool n Quite) and I'd got a near silent PSU and passive CPU
sink for it. ;-(

Incidentally with the Sempron 2600+ the system was pulling about 63W
and with the 3000+ it went up more but also settled at just under 60
when idle. I didn't jet chance to actually measure it with the Turion.
;-(

Cheers, T i m

A video card is not required to "make it beep"; indeed, the lack of a
video card will actually cause a beep code error. As a practical
matter, however, a video card is normally considered to be part of the
minimum system.

I have seen instances in which a motherboard would not post because the
ps/2 mouse and keyboard were plugged in backwards (mouse in ps/2 KB port
and vice versa).

The wrong kind of memory can cause failure; some motherboards may
require a full bank of memory (not one module, but a set of multiple
modules).

You seem to be in denial that the most likely problem other than a user
error IS a bad motherboard.

Bill Anderson wrote:
What is required to make a beep?

Processor (and fan unless you're living dangerously)
Power supply
Stick of memory

Anything else? Is a video card required?

I'm still having my crazy problem in which my P5K Deluxe system won't
post. It's getting worse. I cannot believe the problem is my power
supply as I have two power supplies and when I switch them out the
problem continues unchanged. I even RMA'd one of them and it came back
untouched -- they could find no problem.

I want to strip the system down to the barest essentials and add things
one at a time to figure out what makes it fail. Problem is, failure is
intermittent so this problem is a tough one to pin down.

Some memory is normally required. Without memory (RAM) the CPU cannot
establish a stack and cannot execute subroutine calls (or, more
correctly, the returns from those calls). This will usually prevent
even the POST single beep.


Paul wrote:
Bill Anderson wrote:
What is required to make a beep?

Processor (and fan unless you're living dangerously)
Power supply
Stick of memory

Anything else? Is a video card required?

I'm still having my crazy problem in which my P5K Deluxe system won't
post. It's getting worse. I cannot believe the problem is my power
supply as I have two power supplies and when I switch them out the
problem continues unchanged. I even RMA'd one of them and it came
back untouched -- they could find no problem.

I want to strip the system down to the barest essentials and add
things one at a time to figure out what makes it fail. Problem is,
failure is intermittent so this problem is a tough one to pin down.


If you have a processor and a motherboard, the processor executes the
BIOS code on the motherboard. If the BIOS code does a check for memory
or a video card, and they're missing, then the BIOS will attempt to make
a beep code, using the system speaker.

If memory is present, and the locations below 640K are bad, in fact the
BIOS may not be able to beep in that case. Which is why you should
remove the memory, if attempting to verify it can beep at you.

So that means, processor+HS/fan, motherboard, power supply, case speaker,
front power switch to start the system, might be enough.

A bad motherboard can be intermittent. For example, if a large chip were to
have a cracked solder ball, that could make and break contact, when the
motherboard is flexed.

If, when installing the motherboard in the system, there was a metal
standoff
present, where there is no hole for it to line up with, then that standoff
could short to something. On one particular Asus motherboard, when that
happens,
an audio channel winks out. So something as trivial, as not checking
standoffs,
before putting a new motherboard in an old system case, can cause
problems like
that.

That kind of thing, is one reason why eventually a "cardboard test"
might be
suggested to you. You can put the motherboard on your kitchen table, along
with the power supply. Place a telephone book, with cardboard cover,
underneath
the motherboard as a base. You can then test the system there. I've built
up entire systems on a kitchen table, installed Windows and tested
components like that. The biggest risk by doing so, is the video card is
unsupported in the slot it is plugged into. If something tugs on the
video cable,
it can be game over. That is not the kind of experiment to do, with kids
around.
You need a quiet work area, where the gear won't get disturbed.

Paul

Barry Watzman wrote:
A video card is not required to "make it beep"; indeed, the lack of a
video card will actually cause a beep code error. As a practical
matter, however, a video card is normally considered to be part of the
minimum system.

I have seen instances in which a motherboard would not post because the
ps/2 mouse and keyboard were plugged in backwards (mouse in ps/2 KB port
and vice versa).

The wrong kind of memory can cause failure; some motherboards may
require a full bank of memory (not one module, but a set of multiple
modules).

You seem to be in denial that the most likely problem other than a user
error IS a bad motherboard.


I bought a new motherboard -- a P5Q Pro Turbo -- and removed all parts
from the P5K Deluxe and installed them on the P5Q. Guess what: it
wouldn't post.

So I began removing parts from the P5Q and discovered that one of my
four 1G sticks of memory seemed to be the problem. When it was
installed, no post. When it was removed, the P5Q worked. Install it;
no post. Remove it; beep.

OK, I figured I had isolated the problem. And if the trouble has been
bad memory, why torture myself reinstalling Windows on a new motherboard
when the old one seemed not to be the culprit?

So out came the P5Q and back went the P5K. And with just the three
sticks of memory installed, it was working fine. For a few days. Then
it was up to its old tricks.

So there's the reason I'm in denial about the motherboard being the culprit.

As for being in denial about user error, here's what happened this
morning when I turned on the computer.

1) I pressed the spacebar and the computer began powering up. Fans were
running for maybe 3-4 seconds. Then everything shut down. I watched
and waited for about 2-3 seconds. Then the fans powered up again all
by themselves, the hard drives began running, the optical drive light
came on, but no post. The blue light on the power button of my monitor
was blinking, which is the normal indication that it's powered up but
getting no video signal.

2) I held the case power button for about four seconds and the system
shut down. I wanted a few seconds and pressed it again. The fans came
on for no more than a second, maybe less, and then everything shut down.
I watched for 2-3 seconds and the fans came back on all by themselves,
the hard drives began running, the optical drive light came on, but no
post. Again, the monitor was receiving no video signal.

3) I held the case power button for about four seconds and the system
shut down. I wanted a few seconds and pressed it again. The fans cam
on, the hard drives began running, the optical drive light came on, but
no post, no video signal.

4) I held the case power button for about four seconds and the system
shut down. I wanted a few seconds and pressed it again. The fans came
on, I heard a beep, and I was in business.

The system has been running fine this morning for the past three hours
and I am confident it will continue to run normally for the rest of the
day. It always does. I won't have a problem until I shut it down
tonight and try to restart it tomorrow, when I'm pretty sure I'll
experience the same behavior.

Where in all that is there room for user error? Seriously, what do you
think I could be doing wrong? I'll be delighted to take the blame if
only I can figure out what's going wrong.

--
Bill Anderson

I am the Mighty Favog

Barry Watzman wrote:
Some memory is normally required. Without memory (RAM) the CPU cannot
establish a stack and cannot execute subroutine calls (or, more
correctly, the returns from those calls). This will usually prevent
even the POST single beep.


The code is register based, until memory is commissioned. The
memory is not running, at startup. There is setup to be done
first. And it can beep, with the memory missing.

Paul

"The code is register based, until ...."

That is an assumption on your part. And in some cases it may be true.

But if you are familiar with writing assembly language code, and you
know what goes into a BIOS, you would also know that on a modern
motherboard, it is impossible to get to the point where the single
success "beep" is issued without having a stack and the ability to call
subroutines.

Modern motherboards require FAR too much initialization before memory
can be sized or tested (even superficially) to allow you to get very far
using entirely register based code. Can you get some kind of error beep
out while still in entirely register based code? Depends on how much
priority the code writers put on that particular objective. Yes, I have
seen boards that would issue a beep with no memory at all. But in my
own [fairly extensive] experience, they are the exception, not the rule.
Most commonly, with no memory at all, you get absolutely nothing.


Paul wrote:
Barry Watzman wrote:
Some memory is normally required. Without memory (RAM) the CPU cannot
establish a stack and cannot execute subroutine calls (or, more
correctly, the returns from those calls). This will usually prevent
even the POST single beep.


The code is register based, until memory is commissioned. The
memory is not running, at startup. There is setup to be done
first. And it can beep, with the memory missing.

Paul

maybe a faulty PSU? did you tried another one?