P4M800-M7A Non-Booting Frustration

[Alec]

Hello all,

This is a somewhat long post, and one which you might want to disregard
if you don't know much about assembling computers. If you do know lots
about assembling computers or have had a problem similar to the one
I'll describe in the next paragraph, your help would be very much
appreciated, however. Here's the story:

I just bought (yesterday) a mobo/CPU combo of the Biostar P4M800-M7A
and Intel Celeron D 3.06GHz, which worked beautifully up until about an
hour ago when I tried to install more hard drives. Now when I power
the system on, the power LED, which previously went on when the
computer was powered on, no longer lights, whereas the hard drive LED
is now constantly lit. The system will not boot (no video) when
powered on, nor will it give me any beep codes - it just sits there
spinning the fans and leaving the HD LED on. It will not power down by
a momentary push on the front power switch, but it will turn off after
a four-second push as stated in the manual.

To facilitate your understanding of the situation, I'll go through my
assembly process leading up until the error:
- purchased mobo/CPU combo, inspected, set up on non-static foam pad
- cleaned CPU surface and heatsink contact w/ isopropyl and cotton
swabs
- added heat-transfer paste to CPU surface, mounted adequate heatsink
with fan
- added 512MB pc2100 ddr stick and another 256MB pc2100 ddr stick,
seated correctly
- mounted motherboard on breakout chassis from computer case
- connected LED/switch headers from computer case, also connected
PCI-slot expansion USB jumper cable to USB header pins
- installed breakout chassis back into computer case
- connected ATX power (from a 350W power supply), connected VGA cable
leading to CRT monitor
- powered on computer, succesfully booted into BIOS, saved a couple
inconsequential settings
- added components one-by-one in the following order, powering on
system to test after each installation:
- - SoundBlaster Live! 5.1 PCI sound card with front breakout
expansion
- - quasi-generic Ultra ATA/133 RAID card (has a Silicon Image ASIC on
it)
- - two optical drives on secondary IDE channel, one master, one
slave, correctly configured
- - one Maxtor 9GB ATA hard drive as master on primary IDE channel

At this point the system worked correctly. I installed the new drivers
for the motherboard integrated peripherals (AC'97 audio, VIA IDE
controllers, etc) and rebooted. No problems, everything working as
intended. With the system in seemingly great condition, I decided to
put the finishing touch on it all: add four big ol' Western Digital
hard drives to the raid controller. I powered the computer down, shut
off the power supply, dutifully ensured that the drives were paired in
master-slave configuration, and then connected the secondary IDE
controller from the RAID card to the bottom two (120GB drives) and the
primary IDE controller from the RAID card to the top two (250GB
drives). I then powered up the computer, only to find that the top two
drives were not being recognized, but that the bottom two were. I
rebooted again to ensure this was not a one-time config mistake. I
then powered the computer down, shut off the power (I think) and
checked all the connections, finding that the top drive (the master) of
the primary IDE pair from the RAID card had its IDE cable only halfway
in (the bottom half of it in, top half out, resulting in about a 45
degree angle). "Ahhhh...", said I, and I reconnected the cable fully.
Expecting my system to finally be fully up and running, I pushed the
power button. The fans spun up, but the display did not go on, and
there was a general lack of usual noise. I power cycled the computer a
couple times -- still nothing. It was then that I noticed that the
power LED that had previously lit with every boot was no longer on, and
that the hard drive LED was now permanently on whenever the computer
powered up (ie, whenever the fans were spinning).

With my system now non-functional, I decided to take steps backward. I
removed each component one-by-one in roughly the same order that I
assembled it. Now I have only the CPU, RAM, and built-in VGA output
connected (well, also the ATX power and the headers to the case
LEDs/switches), but the system still displays the exact same behavior:
no video output, no beeps, no power LED, fans are spinning, and hard
drive LED is constantly on when the system is running. I have reset
the CMOS, but it had no effect on the system behavior.

What could this mean? I suppose it's possible that it's a power issue
of some kind, either coming directly from the power supply to the hard
drives, or a combination of the power coming through the power cables
and the general load brought on by running a PCI RAID card with two
additional IDE channels, now fully connected. I do have a lot of
equipment for one at-least-two-year-old 350W power supply to handle.
Assuming this is some kind of power problem (and not a
partially-connected-IDE-cable-related problem), in what form might it
have manifested? Did the power supply blow a fuse? Did the
motherboard? Or might the PCI bus have had too much power going
through it and melted a copper track? Or, if the system voltage
dropped low because of the initial current spike of turning on all the
components, might that reduced voltage have screwed up the BIOS or some
other component?

Why do computers have to be so damn complicated? Dammit!

If you have any advice, I'd love to hear it, even if it's "You're
screwed. Get a new motherboard."

Thanks in advance,
Alec

 

[Paul]

Hello all,

This is a somewhat long post, and one which you might want to disregard
if you don't know much about assembling computers. If you do know lots
about assembling computers or have had a problem similar to the one
I'll describe in the next paragraph, your help would be very much
appreciated, however. Here's the story:

I just bought (yesterday) a mobo/CPU combo of the Biostar P4M800-M7A
and Intel Celeron D 3.06GHz, which worked beautifully up until about an
hour ago when I tried to install more hard drives. Now when I power
the system on, the power LED, which previously went on when the
computer was powered on, no longer lights, whereas the hard drive LED
is now constantly lit. The system will not boot (no video) when
powered on, nor will it give me any beep codes - it just sits there
spinning the fans and leaving the HD LED on. It will not power down by
a momentary push on the front power switch, but it will turn off after
a four-second push as stated in the manual.

To facilitate your understanding of the situation, I'll go through my
assembly process leading up until the error:
- purchased mobo/CPU combo, inspected, set up on non-static foam pad
- cleaned CPU surface and heatsink contact w/ isopropyl and cotton
swabs
- added heat-transfer paste to CPU surface, mounted adequate heatsink
with fan
- added 512MB pc2100 ddr stick and another 256MB pc2100 ddr stick,
seated correctly
- mounted motherboard on breakout chassis from computer case
- connected LED/switch headers from computer case, also connected
PCI-slot expansion USB jumper cable to USB header pins
- installed breakout chassis back into computer case
- connected ATX power (from a 350W power supply), connected VGA cable
leading to CRT monitor
- powered on computer, succesfully booted into BIOS, saved a couple
inconsequential settings
- added components one-by-one in the following order, powering on
system to test after each installation:
- - SoundBlaster Live! 5.1 PCI sound card with front breakout
expansion
- - quasi-generic Ultra ATA/133 RAID card (has a Silicon Image ASIC on
it)
- - two optical drives on secondary IDE channel, one master, one
slave, correctly configured
- - one Maxtor 9GB ATA hard drive as master on primary IDE channel

At this point the system worked correctly. I installed the new drivers
for the motherboard integrated peripherals (AC'97 audio, VIA IDE
controllers, etc) and rebooted. No problems, everything working as
intended. With the system in seemingly great condition, I decided to
put the finishing touch on it all: add four big ol' Western Digital
hard drives to the raid controller. I powered the computer down, shut
off the power supply, dutifully ensured that the drives were paired in
master-slave configuration, and then connected the secondary IDE
controller from the RAID card to the bottom two (120GB drives) and the
primary IDE controller from the RAID card to the top two (250GB
drives). I then powered up the computer, only to find that the top two
drives were not being recognized, but that the bottom two were. I
rebooted again to ensure this was not a one-time config mistake. I
then powered the computer down, shut off the power (I think) and
checked all the connections, finding that the top drive (the master) of
the primary IDE pair from the RAID card had its IDE cable only halfway
in (the bottom half of it in, top half out, resulting in about a 45
degree angle). "Ahhhh...", said I, and I reconnected the cable fully.
Expecting my system to finally be fully up and running, I pushed the
power button. The fans spun up, but the display did not go on, and
there was a general lack of usual noise. I power cycled the computer a
couple times -- still nothing. It was then that I noticed that the
power LED that had previously lit with every boot was no longer on, and
that the hard drive LED was now permanently on whenever the computer
powered up (ie, whenever the fans were spinning).

With my system now non-functional, I decided to take steps backward. I
removed each component one-by-one in roughly the same order that I
assembled it. Now I have only the CPU, RAM, and built-in VGA output
connected (well, also the ATX power and the headers to the case
LEDs/switches), but the system still displays the exact same behavior:
no video output, no beeps, no power LED, fans are spinning, and hard
drive LED is constantly on when the system is running. I have reset
the CMOS, but it had no effect on the system behavior.

What could this mean? I suppose it's possible that it's a power issue
of some kind, either coming directly from the power supply to the hard
drives, or a combination of the power coming through the power cables
and the general load brought on by running a PCI RAID card with two
additional IDE channels, now fully connected. I do have a lot of
equipment for one at-least-two-year-old 350W power supply to handle.
Assuming this is some kind of power problem (and not a
partially-connected-IDE-cable-related problem), in what form might it
have manifested? Did the power supply blow a fuse? Did the
motherboard? Or might the PCI bus have had too much power going
through it and melted a copper track? Or, if the system voltage
dropped low because of the initial current spike of turning on all the
components, might that reduced voltage have screwed up the BIOS or some
other component?

Why do computers have to be so damn complicated? Dammit!

If you have any advice, I'd love to hear it, even if it's "You're
screwed. Get a new motherboard."

Thanks in advance,
Alec

Based on the description of the hard drive cable being half-inserted,
I'd say you blew an IDE port. To verify that, you'd want to check which
motherboard connector you had that half-inserted connector connected to.
If it happened to be a Southbridge port, then the Southbridge might
have been damaged. If the half-inserted connector had been on a
RAID controller of some sort (i.e. not a Southbridge port), then the
motherboard would still be able to boot, and only the RAID controller
would potentially be damaged.

Let's assume "You're screwed". If you do get another Biostar, one thing to
check on those, is whether the 12V signals on the ATX 2x2 connector
are a dead short to the single 12V pin on the main power connector.
I've read a couple reports, where users were able to boot and run
their computer, with the ATX12V 2x2 connector disconnected. You should
not be able to do that, and if the ATX12V is disconnected, the BIOS
screen should not appear. I cannot explain why Biostar would do that.
The implication of that design decision, is that an ATX supply with
12V1/12V2 (i.e. dual rail, the current standard for supplies), is not
compatible with that Biostar "feature". I would recommend to any potential
user of a Biostar board, to use their ohmmeter, to check for that faulty
design. The ATX12V connector 12V should be separate from the main connector
12V. If it did turn out you had that Biostar "feature", I'd look for an
older supply with a single rail (when you look on the power supply label,
there would be no mention of 12V1 and 12V2, but just a single
current rating for "12V").

This is an example of a supply you could use with a shorted Biostar:

ENERMAX EG365P-VE FMA 1.3 ATX 350W Power Supply
http://www.newegg.com/Product/Product.asp?Item=N82E16817103455

If you don't own an ohmmeter, there is another way to test it. Install
the processor in the Biostar motherboard. Connect the CPU fan. Connect
the motherboard speaker. Connect the power switch. Connect the main
power connector, but not the ATX12V 2x2. If the bad design decision is
present on the board, the motherboard will be able to start, and the
BIOS will beep the missing RAM/missing video card sound on the computer
speaker. That tells you that the processor is running, even though the
ATX12V connector is not present. At that point, you'd want to reconsider
your choice of power supply for the build. I would not connect a 12V1/12V2
type supply, if it was my build, if I did hear the BIOS beep the speaker,
when no ATX12V was connected.

HTH,
Paul

 

[Rod Speed]

This is a somewhat long post, and one which you might want to
disregard if you don't know much about assembling computers.
If you do know lots about assembling computers or have had a
problem similar to the one I'll describe in the next paragraph, your
help would be very much appreciated, however. Here's the story:

I just bought (yesterday) a mobo/CPU combo of the Biostar P4M800-M7A
and Intel Celeron D 3.06GHz, which worked beautifully up until about
an hour ago when I tried to install more hard drives. Now when I power
the system on, the power LED, which previously went on when the
computer was powered on, no longer lights, whereas the hard drive
LED is now constantly lit. The system will not boot (no video) when
powered on, nor will it give me any beep codes - it just sits there
spinning the fans and leaving the HD LED on.

That likely indicates that it cant read the hard drive anymore for some reason.

It will not power down by a momentary push on the front power switch,
but it will turn off after a four-second push as stated in the manual.

To facilitate your understanding of the situation, I'll go through my
assembly process leading up until the error:
- purchased mobo/CPU combo, inspected, set up on non-static foam pad
- cleaned CPU surface and heatsink contact w/ isopropyl and cotton
swabs
- added heat-transfer paste to CPU surface, mounted adequate heatsink
with fan
- added 512MB pc2100 ddr stick and another 256MB pc2100 ddr stick,
seated correctly
- mounted motherboard on breakout chassis from computer case
- connected LED/switch headers from computer case, also connected
PCI-slot expansion USB jumper cable to USB header pins
- installed breakout chassis back into computer case
- connected ATX power (from a 350W power supply), connected VGA cable
leading to CRT monitor
- powered on computer, succesfully booted into BIOS, saved a couple
inconsequential settings
- added components one-by-one in the following order, powering on
system to test after each installation:
- - SoundBlaster Live! 5.1 PCI sound card with front breakout
expansion
- - quasi-generic Ultra ATA/133 RAID card (has a Silicon Image ASIC on
it)
- - two optical drives on secondary IDE channel, one master, one
slave, correctly configured
- - one Maxtor 9GB ATA hard drive as master on primary IDE channel

At this point the system worked correctly. I installed the new
drivers for the motherboard integrated peripherals (AC'97 audio, VIA
IDE controllers, etc) and rebooted. No problems, everything working
as intended. With the system in seemingly great condition, I decided
to put the finishing touch on it all: add four big ol' Western Digital
hard drives to the raid controller. I powered the computer down, shut
off the power supply, dutifully ensured that the drives were paired in
master-slave configuration, and then connected the secondary IDE
controller from the RAID card to the bottom two (120GB drives) and the
primary IDE controller from the RAID card to the top two (250GB drives).
I then powered up the computer, only to find that the top two drives were
not being recognized, but that the bottom two were. I
rebooted again to ensure this was not a one-time config mistake. I
then powered the computer down, shut off the power (I think) and
checked all the connections, finding that the top drive (the master)
of the primary IDE pair from the RAID card had its IDE cable only
halfway in (the bottom half of it in, top half out, resulting in
about a 45 degree angle). "Ahhhh...", said I, and I reconnected the
cable fully. Expecting my system to finally be fully up and running,
I pushed the power button. The fans spun up, but the display did not
go on, and there was a general lack of usual noise. I power cycled the
computer a couple times -- still nothing. It was then that I noticed that
the power LED that had previously lit with every boot was no longer on,
and that the hard drive LED was now permanently on whenever the
computer powered up (ie, whenever the fans were spinning).

With my system now non-functional, I decided to take steps backward.
I removed each component one-by-one in roughly the same order that I
assembled it. Now I have only the CPU, RAM, and built-in VGA output
connected (well, also the ATX power and the headers to the case
LEDs/switches), but the system still displays the exact same behavior:
no video output, no beeps, no power LED, fans are spinning, and hard
drive LED is constantly on when the system is running. I have reset
the CMOS, but it had no effect on the system behavior.

What could this mean? I suppose it's possible that it's a power
issue of some kind, either coming directly from the power supply
to the hard drives, or a combination of the power coming through
the power cables and the general load brought on by running a PCI
RAID card with two additional IDE channels, now fully connected.
Yes.

I do have a lot of equipment for one at-least-two-year-old 350W power supply to handle.
Yes.

Assuming this is some kind of power problem (and not a
partially-connected-IDE-cable-related problem), in what form
might it have manifested? Did the power supply blow a fuse?

Nope, because the fans still come on. It isnt feasible
to fuse the output rails, just the mains input.

The obvious thing to try is to check the rails with a multimeter if you
have one or try the current minimal config with a new power supply.

Did the motherboard?

Possible, but unlikely. They normally only have fuses
on the external stuff like the PS/2 and USB ports etc.

Or might the PCI bus have had too much power
going through it and melted a copper track?

Unlikely and that would normally be on the raid
card anyway and that is no longer in the system.

Or, if the system voltage dropped low because of the initial
current spike of turning on all the components, might that reduced
voltage have screwed up the BIOS or some other component?
Unlikely.

Why do computers have to be so damn complicated?

Because life wasnt meant to be easy.

Dammit!

That's likely why its curled up and died, you swore at it.

If you have any advice, I'd love to hear it, even
if it's "You're screwed. Get a new motherboard."

I'd try the power supply first, essentially because that could
produce the symptoms you are seeing now and its cheaper to try.

 

[Rod Speed]

Based on the description of the hard drive cable being half-inserted, I'd say you blew an IDE
port.

He appears to be saying that that happened with one of the ports on
the raid card, so that shouldnt stop it working without the raid card now.

To verify that, you'd want to check which motherboard connector you had that half-inserted
connector connected to. If it happened to be a Southbridge port, then the Southbridge might have
been damaged. If the half-inserted connector had been on a RAID controller of some sort (i.e. not
a Southbridge port),

That appears to be what he is saying.

then the motherboard would still be able to boot, and only the RAID controller would potentially
be damaged.

So its unlikely that that is what stops it starting now.

Let's assume "You're screwed". If you do get another Biostar, one thing to check on those, is
whether the 12V signals on the ATX 2x2 connector are a dead short to the single 12V pin on the
main power connector.
I've read a couple reports, where users were able to boot and run
their computer, with the ATX12V 2x2 connector disconnected. You should
not be able to do that, and if the ATX12V is disconnected, the BIOS
screen should not appear. I cannot explain why Biostar would do that.
The implication of that design decision, is that an ATX supply with
12V1/12V2 (i.e. dual rail, the current standard for supplies), is not
compatible with that Biostar "feature". I would recommend to any
potential user of a Biostar board, to use their ohmmeter, to check
for that faulty design. The ATX12V connector 12V should be separate
from the main connector 12V. If it did turn out you had that Biostar
"feature", I'd look for an older supply with a single rail (when you
look on the power supply label, there would be no mention of 12V1 and 12V2, but just a single
current rating for "12V").

This is an example of a supply you could use with a shorted Biostar:

ENERMAX EG365P-VE FMA 1.3 ATX 350W Power Supply
http://www.newegg.com/Product/Product.asp?Item=N82E16817103455

Its likely his older power supply is like that already.

 

[Alec]

Thanks for the help, guys. I'm going to test out the computer today
with another 350W power supply that I have lying around from an old
computer.

To confirm Rod Speed's post, yes, the half-plugged-in connection was
leading to the primary IDE controller _of the RAID card_ on the PCI
bus. However, I don't believe that the half-connection killed the RAID
card. I was able to boot successfully when the plug was half in, and
the RAID card simply did not recognize either drive on the primary IDE
port, but did recognize both on the secondary port. I'm not terribly
experienced with hard drive controllers, so I have a question: if the
IDE connector is not fully plugged in, does the drive still spin up?
That is, will the drive still draw significant current even when not
connected to an IDE cable? If it would only spin up when connected,
then that might confirm that this is in fact a power issue, as the
problem only occurred after I correctly plugged in all IDE cables.

Of course, if they draw current regardless of IDE cable status, it
could have just been a statistical occurance - that is, it may have
been drawing a lot of current such that it was just below the PSU's
supply threshold, and when I booted a second time, a random current
spike timing overlap may have caused the downfall of the system.

Also, the issue should not be with the ATX 2x2 plug, as neither my
power supply nor my motherboard have one, although I do have a
converter if I need one (as I did on a different computer).

Fingers crossed that I've blown the power supply, and thanks for your
help.

 

[Alec]

Also, is it possible for me to verify the functioning of each pin on
the ATX power connnector? I downloaded a pinout, and I've verified
that in its "off" state (that is, when the main switch of the power
supply is on but the computer is not powered on), the 5V_SB and PWR_ON#
(is that the inverted indicator of PWR_ON, or is that a connection to
be made that would tell the power supply to spin up its fan etc?) are
both hovering around 5V.

I'd like to be able to have the power supply running at normal
operation (ie, all power on, fan spinning) without having it plugged
into the mobo so I can test for bad pins. I already checked the
continuity of supposedly connected rails on the pinout of the ATX
female plug on the motherboard, and no problems there.

So, the main point is: how do I make the power supply power up so I can
check the pins?

 

[Alec]

So, the main point is: how do I make the power supply power up so I can
check the pins?

While I'd still like to know this, I have a happy announcement: you all
were right about the power supply. Plugged in the spare, everything
booted up correctly. So I'm off to buy another one with greater load
capacity... maybe 500W, just so I'm sure this won't happen again.

Thanks Paul and Rod for your help. You're lifesavers.

Alec

 

[Rod Speed]

Thanks for the help, guys. I'm going to test out the
computer today with another 350W power supply
that I have lying around from an old computer.

To confirm Rod Speed's post, yes, the half-plugged-in connection
was leading to the primary IDE controller _of the RAID card_ on
the PCI bus. However, I don't believe that the half-connection
killed the RAID card. I was able to boot successfully when the
plug was half in, and the RAID card simply did not recognize either
drive on the primary IDE port, but did recognize both on the secondary
port. I'm not terribly experienced with hard drive controllers, so I
have a question: if the IDE connector is not fully plugged in, does
the drive still spin up? That is, will the drive still draw significant
current even when not connected to an IDE cable?

Normally, yes. There are a few drives that will shut down when
they dont like what they see with the ribbon cable, but not many.

If it would only spin up when connected, then that might
confirm that this is in fact a power issue, as the problem
only occurred after I correctly plugged in all IDE cables.

Yeah, certainly some cheap power supplys dont handle an overload
properly and can be killed by that. Some of those can overvoltage
other rails when dying and so kill what is plugged into them too.

If that happened to the +5V rail, that would explain why the fans still run.

Of course, if they draw current regardless of IDE cable status, it
could have just been a statistical occurance - that is, it may have
been drawing a lot of current such that it was just below the PSU's
supply threshold, and when I booted a second time, a random current
spike timing overlap may have caused the downfall of the system.

Also, the issue should not be with the ATX 2x2 plug, as neither
my power supply nor my motherboard have one, although I do
have a converter if I need one (as I did on a different computer).

Fingers crossed that I've blown the power supply, and thanks for your help.

Be interesting to see what it turns out to be, please post the result.

 

[Rod Speed]

Also, is it possible for me to verify the functioning
of each pin on the ATX power connnector?

Yes, on the basic stuff, anyway.

I downloaded a pinout, and I've verified that in its "off" state
(that is, when the main switch of the power supply is on but
the computer is not powered on), the 5V_SB and PWR_ON#
(is that the inverted indicator of PWR_ON, or is that a connection
to be made that would tell the power supply to spin up its fan etc?)

The # in the name just indicates that its active low. In other words
its supposed to be high and when the motherboard tells the power
supply to turn on, the motherboard pulls it down to roughly ground.

are both hovering around 5V.

That is as it should be.

I'd like to be able to have the power supply running at
normal operation (ie, all power on, fan spinning) without
having it plugged into the mobo so I can test for bad pins.

You can ground the PS_ON# line to do that.

Quite a few power supplys wont start with no load tho,
you should plug at least an older surplus hard drive into
one of the hard drive power plugs to give it some load.

I already checked the continuity of supposedly connected rails
on the pinout of the ATX female plug on the motherboard,

Not clear what you mean by that.

and no problems there.

Presumably you did check the main rails voltages
with the motherboard connected and with the front
panel power switch pressed momentarily ?

So, the main point is: how do I make the power
supply power up so I can check the pins?

Just ground the PS_ON# line and provide some load.

 

[Rod Speed]

While I'd still like to know this, I have a happy announcement:
you all were right about the power supply. Plugged in the
spare, everything booted up correctly. So I'm off to buy
another one with greater load capacity... maybe 500W,
just so I'm sure this won't happen again.

Thanks Paul and Rod for your help. You're lifesavers.

Thanks for the feedback, too rare in my opinion.

 

[DaveW]

You apparently at least fried the power supply unit by drastically
overloading a mere 350 Watt PSU with far too many drives. If it was an
inexpensive (cheaply made) PSU, then it is somewhat likely that it destroyed
the motherboard too.