At 06:11 1/22/2000 , you wrote:
>David G wrote:
>>Just to confirm, you understand that on most SCSI drives the three
>>ID jumpers are numbered 0/1/2, but their *values* are 1/2/4 -- the
>>actual ID being the sum of the values of the jumpers installed.
>>Could this be why your IBM drive is addressed this way?
>No I didn't know that...Now I seem to have developed a possible bad
>Cheetah. After leaving things on to run Seti in the background
>overnight, I went to wake up the monitor this morning. Nothing
>happened. The monitor lamp changed to green as it should, but I got
>no display at all.


Maybe a power saving setting? Do you have the drive set to spin down?
(If you have EZSCSI4 you can do that. If it's a Win9X power
setting, I cannot help since I always disable this in both
the motherboard BIOS and Win9X Device Manager. Run SCSI Explorer
(installs with Adaptec's full EZSCSI) and check power settings for
your drives. Maybe the monitor failing to return was just a Win9X power
saving glitch and the DMI pool settings changed at that point.

>After a hard reboot, I got the ominous message after "Verifying DMI Pool
>Data", "Boot disk failure, insert system disk and press enter". This
>doesn't sound good. Upon another reboot, I see that the drive shows up
>in the Adaptec BIOS scan but with no ID#.

The Boot disk failure error message is standard if no boot disk
is detected. If you put a non-bootable floppy in the floppy drive and
try to boot, you will get this message. Of if the first hard drive does
not have an active partition or if the active partition does not
include the necessary boot files, you get this message. It does not
necessarily mean you have a bad hard drive.

Please take out that IDE drive while you try to trouble shoot these SCSI
problems. (It will make things simpler.) When you take it out, you should
see "Verifying DMI Pool Data" and then the boot disk error again.

Check to see if you are still set to boot from SCSI ID 0 in the
SCSI BIOS.

Your SCSI adaptor is not identifying your SCSI drives the way you want.
(Say you are set to boot from ID 0 but the drives are "1 and 2 and
3" or "1 and 1 and 2"...or "0 and 0 and 2" or "0 and 0 and 1" etc.

Did you change the SCSI IDs after reading David's message?  Did you
change one but not the other...so that they are both 0 or both 1? Check
the jumpers for SCSI ID on both Cheetahs. And check again! Look for
missing or miss-set jumpers.

By the way, if you have two drives set for 0 (or both 1), your Adaptec
might see them in the BIOS scan due to SCAM. Try turning off SCAM
in the Adaptec BIOS and see what happens. Did you lose one Cheetah
in the BIOS scan? SCAM may not work all the time...or correctly.

You want no ID jumper on the boot drive (all 4 pairs of ID pins open)
and a jumper across the ID 0 pins on the second Cheetah and a jumper
accross the ID 1 pins on the IBM.

First look to see if your boot drive has a jumper present. If it does,
it's not set for ID 0.

Was the SCSI BIOS installed? (You are notified after SCSI BIOS scans
the devices.) It should be installed, but if it wasn't you wouldn't
see the hard drives in DOS...and hence in FDISK (unless you used FDISK
from within the full Win9X GUI).

Boot from floppy disk and see what disk DOS sees as drive 1. The other
disks would be drives 2 and 3. But you might not see all three of them.
Which is missing? Is it the one that contains the operating system?
(Remember that the 2940UW calls the drives 0, 1, and 2 whereas FDISK
calls the drives 1, 2, and 3.) The knowledge of which drive DOS thinks
is drive C: may help. (Look at the files that are present on this disk.)

>In the Adaptec set up, the drive won't verify, with
>the message in Sense key 02h, which translates to "Not ready- The media is
>not ready to format. Be sure that media is inserted in the drive and that
>the media is spun up". I am unable to see the drive in Fdisk from a Win98
>boot disk. So I guess that means that the drive is kaput. I haven't
>changed
>anything, things were working before. Everything is properly plugged in,
>terminated etc.

Something is wrong with the way the 2940UW adaptor sees this drive.
I assume that this problem drive is the drive that you want to boot from.

>Could a failing even though new drive have been the possible cause of
>earlier problems, like being forced to go to safe mode upon any kind of
>software update?

No. You had two other major problems before. (Cable length and passive
termination.) You did something since those were fixed...maybe changing
ID jumpers.

>Is this a failed drive?
>I reinstalled my old WD IDE that was still loaded with Windows to
>access my
>mail etc. I did have it removed until the failure above and had
>changed the
>BIOS accordingly.

As a general rule, I would go into SCSI Select and set things so that
only IDs 0, 1, and 2 are scanned at boot. (If your IBM drive is ID 2.)
Also, set these IDs for Wide, Synchronous, and Send Unit Start Command
...and set all other IDs except 7 (that's 3 through 6 and 8 through 15)
to the opposites of these three settings. Watch out for 7 since this is
the SCSI controller itself. Set 7 for Wide and Synchronous enabled...
and 40 MB/sec (that's 20 MHz).

Disable "treat removable drive as fixed drive at boot" in SCSISelect
and disable "boot if bootable CDROM detected"... just for general
principles. Do you have the SCSI BIOS set for "more than 2 drives?"
(There's a setting for that. Enable it. It may help in DOS or with
Partition Magic someday.) Tell the SCSI BIOS to reset on errors.

When you get a SCSI problem, you have to check and recheck and
re-recheck all the settings and jumpers and terminations. Make sure
your ribbon cable is tightly plugged into the problem drive. Of course
you could fry a drive if you plug a cable into the wrong pins...or bend
some pins so that they short. (I'd think that having pins bent would
screw up the whole SCSI bus however...and cause problems with other
devices...maybe fry the controller so it wouldn't even report in at
boot.) But maybe the drive would still work after fixing the connection.
So unplug the problem drive and look at the pins on the cable to see if
any are bent. Then carefully plug the drive back in.

Also, think about power. Make sure the power connector is tightly
attached to the problem drive. (You might want to pull the power plug
from the problem drive and plug it back in.) Try switching power
connectors with a different drive. See if the problem shifts to the
other drive.

Check your spin up delays. If you have spin up delays set, you'll
see drive 0 appear in the BIOS scan right away, then 10 seconds
later drive 1 will start to spin up and then appear. Then after
another 10 seconds, drive 2 will start to spin up and then appear.
(Can you hear them spin up?) Could you hear drive 0 spin up when
the machine was doing the POST memory test? This is an important
question. Can you set drive activity LEDs for your drives? These
are helpful for diagnostic reasons...but you need the LEDs and
your Bay Coolers must have a place for the LEDs.

It does trouble me that you apparently booted okay at first...
and then the problem occurred overnight as the computer was running.
Are you sure you didn't make any changes in between?

Regards,
Bill

         PCBUILD maintains hundreds of useful files for download
                     visit our download web page at:
                     http://nospin.com/pc/files.html