Boot won't start

I put together a machine from old spare parts. It boots to the MB spash
screen, but goes no further.

The MB is an Intel D850MV. This is a SCSI system with a couple
Adaptec adapters supporting several SCSI hard disks.

The power supply is one that was working not long ago, and so assume it
can pass the power-on test. The CPU should then execute BIOS code, and
so an issue might be RAM, but it also has been working OK. My memory
tester is on floppy, and unfortunately I don't have a floppy drive in
the machine at this point to run it.

Does the display of the MB spash screen suggest the CPU is already
reading ROM BIOS code? The first thing that the ROM BIOS code does is to
POST, but I get no beep, so I know I'm not getting that far.

If one starts with the assumption that MB, CPU, and RAM are all in good
shape, what would be the most likely area of trouble? I don't have a
power LED connected because the board presumes two pins and the case
offers a three pin connector.

--

Haines Brown, KB1GRM

Haines Brown wrote:
> I put together a machine from old spare parts. It boots to the MB spash
> screen, but goes no further.
>
> The MB is an Intel D850MV. This is a SCSI system with a couple
> Adaptec adapters supporting several SCSI hard disks.
>
> The power supply is one that was working not long ago, and so assume it
> can pass the power-on test. The CPU should then execute BIOS code, and
> so an issue might be RAM, but it also has been working OK. My memory
> tester is on floppy, and unfortunately I don't have a floppy drive in
> the machine at this point to run it.
>
> Does the display of the MB spash screen suggest the CPU is already
> reading ROM BIOS code? The first thing that the ROM BIOS code does is to
> POST, but I get no beep, so I know I'm not getting that far.
>
> If one starts with the assumption that MB, CPU, and RAM are all in good
> shape, what would be the most likely area of trouble? I don't have a
> power LED connected because the board presumes two pins and the case
> offers a three pin connector.

Since you get the splash screen the power's working and the CPU's
starting the POST. In my experience CPUs and BIOSen are all-or-
nothing: they either work or they don't, so the CPU and the BIOS
are okay.

Here's the Intel website for the board:
http://www.intel.com/design/motherbd/mv/index.htm

Do you have some other known-good memory of any size to try?
Note that RIMMs need not be installed in matched pairs, just
put CRIMM fillers in the blank slots. Your memory may have
developed an error in the upper area, higher than the low
memory the CPU uses for POST.

You _have_ removed all the cards except for the VGA and
disconnected any IDE drives? Your AGP card is 1.5V 4X?

Now, all that being said I have one board where the POST
starts, it seeks the floppy, and then it goes stupid.
Haven't found what's wrong. I may be incorrect and BIOSen
may develop bitrot. I snagged a fancy EPROM maker on ePay
that'll write the funny square ones so I'm gonna try
rewriting it in the near future.

Hope this helps, and the best of luck.

--
-- Marten Kemp
(Fix name and ISP to reply)

On Apr 5, 6:20*pm, Haines Brown <bro...@teufel.hartford-hwp.com>
wrote:
> Thepower supplyis one that was working not long ago, and so assume it
> can pass the power-on test. The CPU should then execute BIOS code, and
> so an issue might be RAM, but it also has been working OK. My memory
> tester is on floppy, and unfortunately I don't have a floppy drive in
> the machine at this point to run it.
>
> Does the display of the MB spash screen suggest the CPU is already
> reading ROM BIOS code?

The very first BIOS byte requires a CPU to execute. Chances are
that all CPU functions are working if it puts every little does on a
splash screen display. However, even a defective power supply can
permit the system to get that far.

Generally, a CPU that can get that far should also be able to beep a
speaker. Is that speaker properly connected?

Long before trying to check out anything else, are power supply
voltages still good even when system stops on a splash screen? IOW a
3.5 digit multimeter and 30 seconds can confirm any one of power
supply's orange, red, purple, and yellow wires. Those numbers must
exceed 3.23, 4.87, and 11.7. And its gray wire must remain above 2.4
volts.

Trying to boot without memory must definitively result in speaker
beeps. If not, find out why the speaker is not properly connected.
If speaker cannot beep, then other problems probably exist. CPU does
not need keyboard, mouse, sound card, video card, memory, disk drive,
or most of the motherboard to beep that speaker. But it does need
power supply voltages. A defective supply can boot or partially boot
a computer.

w_tom <> writes:

> On Apr 5, 6:20Â*pm, Haines Brown <bro...@teufel.hartford-hwp.com>
> wrote:
> The very first BIOS byte requires a CPU to execute. Chances are that
> all CPU functions are working if it puts every little does on a splash
> screen display. However, even a defective power supply can permit the
> system to get that far.

That I wouldn't have guessed. Thanks for the info.

However, water has gone over the dam. When I turned the machine on its
side, I ceased having anything at all (no fans, etc.). I suspected the
power switch, but eventually ruled that out. So I disconnected the power
supply from anything (except one drive moter for min load) and measured
voltages, and there were none.

A replacement power supply on its way. However, without any connections,
I wonder if the power supply can be expected to work. Since power-on
depends on a temporary closing of a circuit external to the PS (the
front panel on switch), can one in fact test PS-voltage outs with
nothing connected to it or without having to short something?

One curious thing. If I depress and hold the power-on switch five
seconds and release it for five seconds, and repeat this, about every
third cycle the CPU fan goes through a few rotations. Does this sound
like an overload, either inside or outside the PS, that is causing a PS
shutdown? At least I know I don't have a fuse or power switch problem.

--

Haines Brown, KB1GRM

On Apr 8, 9:09 am, Haines Brown <bro...@teufel.hartford-hwp.com>
wrote:
> However, water has gone over the dam. When I turned the machine on its
> side, I ceased having anything at all (no fans, etc.). I suspected the
> power switch, but eventually ruled that out. So I disconnected the power
> supply from anything (except one drive moter for min load) and measured
> voltages, and there were none.

Power supply 'system' includes something separate from the power
supply - a power supply controller. No power supply controller - then
power supply cannot operate.

First, only useful measurement of a power supply is with everything
connected. Measuring with only one load says little. More load means
a more useful number. A two minute procedure would have identified
which (if anything) in that power supply 'system' is defective AND
provided numbers so that we *know* the replacement part actually fixed
something. Only useful test of a supply is when connected to its
'controller'.

A computer can boot just fine - and the power supply 'system' is
still defective. But those multimeter numbers will even identify
failures sometimes months before those failures cause a computer
crash.

A two minute procedure is "When your computer dies without
warning....." starting 6 Feb 2007 in the newsgroup alt.windows-xp
at:
http://tinyurl.com/yvf9vh
Connector chart to locate each color:
http://www.hardwarebook.net/connecto.../atxpower.html
Numbers from those measurements posted here may elicit further useful
facts. Those same numbers that tell you little can sometimes tell me
something different.

If the power supply is replaced, a new supply is not known working
until again taking those four critical voltage numbers on orange, red,
purple, and yellow wires with a maximum load. Just another reason why
that two minute procedure is so informative.

Is that how it is done. I was thinking something to solve the problem. But i guess my thought was wrong. I'll check my books then.