POST PCI board?

  • Thread starter Thread starter Davej
  • Start date Start date
Davej said:
Are these still needed to get POST codes?
Click to expand...

Depends. If your MB has a speaker attached or built-in, and you have a POST problem, you'll get a series of beeps. The
meaning depends on your BIOS. I used to have a link to what they all meant, but seem to have deleted it. Google will
help you :-)
 
Davej said:
Are these still needed to get POST codes?
Click to expand...

They still work.

There are even PCI Express versions.

The only problem with POST codes, is getting an
accurate table of values. On several occasions,
someone will write in, that they have a particular
code, I look it up, and it is "reserved". That
means the BIOS designers used the code, but
did not document it.

As a result, the main value of POST codes, is as
a "go, no-go" indicator. If the POST hex display
stays at 0x00 or 0xFF, then you know the processor
is not executing BIOS code, and either the BIOS
chip is bad, or the processor is bad and failing
soon after the reset is deasserted.

As "SC Tom" mentioned, you can use "beep testing" as
an alternative. If you turn off the power and
pull the RAM sticks, then if the motherboard is
otherwise functional, the processor will beep "RAM
failure" in response. But to do that, the processor
must read BIOS code from the flash chip. So if
it beeps, you know it is not completely dead and can
read some BIOS code. And then, you've got about as much
info as you might have obtained from a POST card.

There are a few retail motherboards, that come with
a two digit hex display right on the motherboard. That
is a built-in POST code display. In such a case, you
save the money on buying a POST card to use for testing.

The cheapest POST cards, come from Ebay via Hong Kong.
You get the cards much cheaper that way, than spending
$100 at your local computer store.

So yes, if you want, you can still buy them, and
the BIOS will still be writing out the codes. It's all
a matter of whether the codes are documented properly,
and actually lead you in a useful direction in terms
of hardware diagnosis.

Since they're "progress codes" and not "error codes",
they have limited utility. The timing of the values
seen, tells you about as much, as the actual value displayed.
For example, if the last code change seen, is around
the 30 second mark, then you're probably failing a
keyboard check.

Paul
 
They still work.

There are even PCI Express versions.

The only problem with POST codes, is getting an
accurate table of values. On several occasions,
someone will write in, that they have a particular
code, I look it up, and it is "reserved". That
means the BIOS designers used the code, but
did not document it.

As a result, the main value of POST codes, is as
a "go, no-go" indicator. If the POST hex display
stays at 0x00 or 0xFF, then you know the processor
is not executing BIOS code, and either the BIOS
chip is bad, or the processor is bad and failing
soon after the reset is deasserted.

As "SC Tom" mentioned, you can use "beep testing" as
an alternative. If you turn off the power and
pull the RAM sticks, then if the motherboard is
otherwise functional, the processor will beep "RAM
failure" in response. But to do that, the processor
must read BIOS code from the flash chip. So if
it beeps, you know it is not completely dead and can
read some BIOS code. And then, you've got about as much
info as you might have obtained from a POST card.

There are a few retail motherboards, that come with
a two digit hex display right on the motherboard. That
is a built-in POST code display. In such a case, you
save the money on buying a POST card to use for testing.

The cheapest POST cards, come from Ebay via Hong Kong.
You get the cards much cheaper that way, than spending
$100 at your local computer store.

So yes, if you want, you can still buy them, and
the BIOS will still be writing out the codes. It's all
a matter of whether the codes are documented properly,
and actually lead you in a useful direction in terms
of hardware diagnosis.

Since they're "progress codes" and not "error codes",
they have limited utility. The timing of the values
seen, tells you about as much, as the actual value displayed.
For example, if the last code change seen, is around
the 30 second mark, then you're probably failing a
keyboard check.

    Paul
Click to expand...


Thanks for the insight. Sounds like I might be better off with some
spare parts. I have a system built on a P5B-E that has never, since
day one, booted cleanly in a cold room. It needs to warm up or it will
hang before anything appears on the screen. I will have to look and
see if it has a speaker connected directly to the MB.

Thanks again.
 
Davej said:
Thanks for the insight. Sounds like I might be better off with some
spare parts. I have a system built on a P5B-E that has never, since
day one, booted cleanly in a cold room. It needs to warm up or it will
hang before anything appears on the screen. I will have to look and
see if it has a speaker connected directly to the MB.

Thanks again.
Click to expand...

First thing I would check on that machine would be the CPU, power supply, and (if present) any case fans that are
connected to the MB. Years ago, we had a number of new Compaq workstations, and two of them displayed the same behavior.
Brand new, straight from Compaq to us.
Turned out that on one of them, the PSU fan was very slow to spin up, and would halt the system because it didn't send
the proper signal (correct RPM) so that the system would know it was OK to continue on. Power it off/on a couple of
times, then it would start fine. Two or three drops of oil cured it.
On the other one, we never did figure out exactly what was causing it (didn't want to waste much time on a warrantied
item), so Compaq sent us a new one that worked as it should.
 
Back
Top