M4A78-E quirky BIOS resets 1st HD
JClark
ASUS M4A78-E
Latest BIOS (1708)
AMD Phenom X4
2GB OCZ PCZ PC2 6400 800 MHZ (X2)
WDC Caviar SATA HD 1T (2)
RAID 1 Array (AMD BIOS)
Seagate External USB HD 1T
Problem:
About every fourth boot, the BIOS resets the first hard drive to the
external drive, which is not bootable, so I have to go back into the
BIOS and change the first hard drive to RAID ARY-1 to get it to boot.
I've been through the BIOS with a fine-tooth comb looking for
something to change this quirky behavior, but no success.
I also still get the funny lack of recognition of the USB mouse (on
bootup) frequently enough to be annoying, but it's improved after
increasing the Southbridge voltage to 1.35. (Discussed extensively in
another thread.) I only mention this in case the two situations are
somehow connected.
It isn't an insurmountable problem for me at home, but I planned to
put this system in my office, where it will be the "server" with two
workstations networked. My office girls will have no clue how to get
back into the BIOS and reset hard drive 1 to RAID-ARY-1. This will be
a real problem there.
Any thoughts or suggestions greatly appreciated.
Jack
Paul
BIOS designs have an amazing number of differences in behavior
between them. Even though the core BIOS design comes from companies
like Award (Phoenix) or AMI, there can still be policy differences
from one generation of BIOS to the next.
The first "feature" that could be responsible, is "overclocking recovery".
That feature is intended to detect an "unclean" session, meaning the BIOS
uses a flag of some sort, to determine whether everything completed
normally during the last session. My Asrock board, detects three unclean
sessions (three presses of reset in a row), as an indication the
CPU operating frequency may be too high. The BIOS design may respond
by resetting *all* settings to defaults, or it may just change the
CPU input frequency and multiplier back to nominal values. That way,
the user can get into the BIOS and try another overclocking value.
For users that don't overclock, they may be puzzled by seeing their
BIOS settings change on their own. But the feature isn't clever enough
to tell the difference between someone overclocking, and a system
running at nominal which isn't stable for some reason. A board
that doesn't start with full stability at T=0, can also trigger
the feature.
The second "feature" is the logic surrounding boot disk choice. I've
had BIOS, where *every* time you change even one disk, you have to
enter the BIOS and verify the boot choice. Otherwise, you'll end
up booting from the wrong disk. The logic on my current board
is much better, and is similar to the behavior of my older boards.
My current board will hold onto its boot choice, if I add or remove
an unrelated disk. So I can trust it enough, to not be entering
the BIOS all the time to fix it. If I unplug the actual boot
disk, then I can expect it to choose something else from
the BIOS list.
If the only BIOS setting that is flailing around right now is
the boot choice, then it could be the second "feature" doing it.
A bad BIOS design that gets upset when any hard drive goes
missing. If the BIOS doesn't detect some drive in time, it
may be determining a drive is removed and changing the boot
choice. If you find a number of other BIOS settings are
also getting changed, then you'd have to assume that
is an overclocking recovery event. So if your Vsb is
getting returned to 1.20V, that could be caused by
overclocking recovery.
Paul
JClark
>For users that don't overclock, they may be puzzled by seeing their
>BIOS settings change on their own.
Hi Paul,
Yes that describes me. All overclocking options are untouched from
default (AUTO) settings.
>If the only BIOS setting that is flailing around right now is
>the boot choice,
That seems to be the case. The previously noted USB mouse problem
(thanks again!) now seems to be solved by (sigh!) buying a newer mouse
to replace the old IBM USB mouse I was using. I should post that as a
solution in the other thread, and will do so). My embarrassment is
overcome by my need to inform.
>if the BIOS doesn't detect some drive in time, it
>may be determining a drive is removed and changing the boot
>choice.
I think that is what was happening. I've achieved some success, I hope
permanent, by disabling the second drive (which was the external USB
drive) in BIOS. This choice is found under: BOOT|HARD DISK DRIVES
1st Drive
2nd Drive
I selected RAID:RAID Ary 1 for 1st boot drive, and "disabled" for 2nd
drive.
Then you have to go back to the preceding selection, "Boot Device
Priority" and make sure that the 1st boot device is RAID:RAID Ary 1.
with this setup, so far it's working. Now I can dump this system on
the office girls. Hooray.
>if your Vsb is
>getting returned to 1.20V,
No, it's still set at 1.35, but now that I've hopefully discovered the
"answer" to both apparent BIOS quirks, I will reset it to the nominal
1.20 and see what happens.
Can't thank you enough for your interest and instruction. As with
Chaucer's clerk: "And gladly wolde he lerne and glady teche."
Jack