Hardware Reference
In-Depth Information
would check the CMOS to make sure that the IDE controller(s) were enabled. Sometimes you
can boot from the secondary IDE/EIDE controller, so I would try that too. If it boots, great!
Time for a new motherboard or perhaps just purchase a new controller and disabling the
onboard controller. But I would seriously consider getting a new motherboard when budget
allows.
4. If I only had the one computer, then I would have to search for a known good hard drive
(and cable) that the BIOS supports. Then if it too does not boot, then I would have to guess
it's something with the controller/motherboard. If it does boot, then I would have lean towards
a bad drive.
5. Sometimes with an "Invalid Drive Type", you can actually boot with a floppy (assuming that
the drive is not an NTFS, HPFS, LINUX, Novell NetWare or some other format....) and then
access the hard drive. If this can be done, this might be one way to back up any data. You can
set up the "bad" drive as "slave" and then with a new drive formatted with whatever format is
needed, copy over whatever data that can be read on the "bad" drive. Back in the good old
days, when we had a drive that went beyond the 1024 cylinders (which is 99.99 percent of all
the IDE/EDIE drives made since 1992 and on), we had to "trick" the BIOS/CMOS. This was
done by taking the cylinders, dividing the number in half, and then doubling the heads:
Example: 1138 cylinders, 8 heads, 63 sectors this would translate to 569 cylinders, 16 heads
and 63 sectors. I would try this trick as perhaps my last resort. But this was used/done on
386/486 machines back in the late 1980s to early 1990s. One could look into a sector-by-
sector copying tool. Gibson research, the makers of
SpinWrite (
http://grc.com/spinrite.htm)
,
have an excellent utility. If the data is that important (mission critical), a drive recovery center
would have to deal with it. But be prepared to pay for it!. Not cheap but would have cost us
more had we not been able to recover the data. Some other things to consider include that a
bad power supply can also cause a hard drive not to boot (not allowing it to spin up to full
RPM), the amperage required to spin the motor is more than what the motherboard
draws/needs even if it has a full bus. Also, I've even seen some ISA, PCI, and AGP cards cause
conflicts with onboard IDE/EIDE controllers (usually in the form of IRQ and/or memory
address). Though these are usually funky-specialized boards, I have seen it happen. Again, I
am assuming that the drive is an IDE/EIDE. If it's an MFM, RLL, ESDI or SCSI, then the tactics
would differ slightly as each are set up and controlled differently. But since IDE/EIDE is
perhaps the popular and most used drive, I am going to assume that is the drive.
From: Jim Davison
Since you did not state IDE/SCSI, I will assume IDE. I will also assume that drive is not using a
bios modifier like those used to make older motherboards support larger drives. (I have seen
situations where users tried to enter drive specks in setup for SCSI drives). I would use the
following steps even considering that you state the user had "Tried" to manually enter the
settings in setup and also tried auto. The user may not know what they are doing. I also would
not trust the error codes. I would assume the error codes are only letting us know there is a
problem but would not trust the codes to give an accurate description of what the problem is.
1. In setup, Try IDE Auto Detect to see if the bios can even see the drive.
If yes, then I would use that setting and everything should be OK.
If yes, and the drive still does not boot, I would use fdisk/mbr in case the Master boot
record was destroyed.
If no, then I would go to step 2.
