10 Jan
2010
10 Jan
'10
8:08 p.m.
Hi!
I'd like to install on an old i386SX-16 mainboard. It comes along with
8MB RAM and 387-25 co-processor.
My problem is that none of the 42 predefined HDD types fits my HDD,
unfortunately not even the heads and sectors settings. (Missing some
of the HDD's space due to a too small cylinders value wouldn't be a
problem.)
So I've got this mainboard. I cannot find a model description
somewhere, only "REV 2B" and "COPYRIGHT 1989". Quality assurance
sticker
displays "HOWTEH Q.A", another sticker has "S/N: H 3204328" on it.
During boot-up, "386SX/82335 Modular BIOS Version 3.04 (P24)" and
"Copyright(c) 1984-88 Award Software Inc." are displayed, Ctrl+Alt+ESC
gives the BIOS settings screen. (A photo can be found at
http://lug-owl.de/~jbglaw/i386SX.jpeg .)
My problem is that HDD settings 1 to 42 don't fit my HDD in any way.
43 to 47 contain all-zeroes values, so I guess they're all used for
user-defined type. However, I cannot figure out any way to enter my
correct C/H/S values. The displayed help is correct, but not exactly
helpful to enter those numbers :) Maybe anybody recognizes this BIOS
and can give me a hint?
Thanks, JBG
--
Jan-Benedict Glaw jbglaw at lug-owl.de +49-172-7608481
Signature of: http://catb.org/~esr/faqs/smart-questions.html
the second :
11 Jan
11 Jan
1:47 p.m.
On 11 Jan 2010 at 2:08, Jan-Benedict Glaw wrote:
My problem is that none of the 42 predefined HDD types
fits my HDD,
unfortunately not even the heads and sectors settings. (Missing some
of the HDD's space due to a too small cylinders value wouldn't be a
problem.)
What kind of hard disk and controller are you using? If it's a
standard MFM, 17 sectors per track was pretty much standard across
all drives. (25, 26 or 32 for RLL controllers, but they have their
own setup).
So you should be able to at least find *something* that will halfway
work.
Of course, if this is an IDE or SCSI drive and controller, the
controller should have its own BIOS to work in a setup like this. If
it doesn't, you need to find one unless you're into BIOS hacking.
You can, in a pinch, select a value that's less in all dimensions.
For example, most IDE drives claim to have 63 sectors per track, but
using only 17 of those will work, even if a substantial part of your
drive will go unused.
Just don't try to low-level format an IDE drive--some drives will
actually support this and the result sometimes doesn't turn out well.
--Chuck
2:45 p.m.
-----Oorspronkelijk bericht-----
Van: cctalk-bounces at classiccmp.org
[mailto:cctalk-bounces at classiccmp.org] Namens Chuck Guzis
Verzonden: maandag 11 januari 2010 19:47
Aan: General Discussion: On-Topic and Off-Topic Posts
Onderwerp: Re: HDD settings on i386 AWARD BIOS
On 11 Jan 2010 at 2:08, Jan-Benedict Glaw wrote:
There is also the possibility to use a program like SpeedStor witch will
load it's own bios extension at boot time.
You can use it to format the drive and use a standard type 2 bios settings,
while you can use the whole disc capacity.
If you want a copy contact me off-list.
-Rik
My problem is that none of the 42 predefined HDD
types fits my HDD,
unfortunately not even the heads and sectors settings.
(Missing some
of the HDD's space due to a too small
cylinders value wouldn't be a
problem.)
What kind of hard disk and controller are you using? If it's
a standard MFM, 17 sectors per track was pretty much
standard across all drives. (25, 26 or 32 for RLL
controllers, but they have their own setup).
So you should be able to at least find *something* that will
halfway work.
Of course, if this is an IDE or SCSI drive and controller,
the controller should have its own BIOS to work in a setup
like this. If it doesn't, you need to find one unless you're
into BIOS hacking.
You can, in a pinch, select a value that's less in all dimensions.
For example, most IDE drives claim to have 63 sectors per
track, but using only 17 of those will work, even if a
substantial part of your drive will go unused.
Just don't try to low-level format an IDE drive--some drives
will actually support this and the result sometimes doesn't
turn out well.
--Chuck
5:07 p.m.
On Mon, 2010-01-11 10:47:10 -0800, Chuck Guzis <cclist at sydex.com> wrote:
On 11 Jan 2010 at 2:08, Jan-Benedict Glaw wrote:
It's one of these common 16bit multi-io ISA cards featuring two serial
ports, one IDE port, a floppy port and a parallel port.
My problem is that none of the 42 predefined HDD
types fits my HDD,
unfortunately not even the heads and sectors settings. (Missing some
of the HDD's space due to a too small cylinders value wouldn't be a
problem.)
What kind of hard disk and controller are you using? If it's a
standard MFM, 17 sectors per track was pretty much standard across
all drives. (25, 26 or 32 for RLL controllers, but they have their
own setup). Of course, if this is an IDE or SCSI drive and
controller, the
controller should have its own BIOS to work in a setup like this. If
it doesn't, you need to find one unless you're into BIOS hacking.
Uh? Had these multi-id controllers have their own BIOS extension? My
understanding is that the BIOS can handle these--all it needs to know
is the geometry. (And this is because the BIOS uses C/H/S addressing
instead of linear addresses.)
I know for sure that I once entered the values (well, actually
cyl=1024 (IIRC) and the "correct" numbers for heads and sectors.) The
drive attached to the box is a modern 20GB HDD, which worked
flawlessly in that box. There's even a Linux system installed on it,
which was installed while the HDD was in that box.
You can, in a pinch, select a value that's less in
all dimensions.
For example, most IDE drives claim to have 63 sectors per track, but
using only 17 of those will work, even if a substantial part of your
drive will go unused.
The trick was to get heads and sectors correct (wrt. the drive's
parameters, even if the drive was larger that some 540MB) and scale
cylinders to the maximum allowed by the BIOS. That way, using a
"modern" large IDE disk, one can boot off with an old BIOS by having a
/boot partition (containing everything that the boot loader (through
BIOS calls) needs to access.)
I even screwed off the VGA/EGA card, the network card as well as the
multi-io board to be able to have a view at the back of the
motherboard. Unfortunately, everything I found was a sticker
mentioning the end-of-guarantee date...
MfG, JBG
--
Jan-Benedict Glaw jbglaw at lug-owl.de +49-172-7608481
Signature of: Fortschritt bedeutet, einen Schritt so zu machen,
the second : da? man den n?chsten auch noch machen kann.
5:57 p.m.
On 11 Jan 2010 at 23:07, Jan-Benedict Glaw wrote:
Uh? Had these multi-id controllers have their own BIOS
extension? My
understanding is that the BIOS can handle these--all it needs to know
is the geometry. (And this is because the BIOS uses C/H/S addressing
instead of linear addresses.)
Some of the early ones did indeed have their own BIOSes.
Some of the early IDE drives even had a translation mode that gave
you 17 sectors per track (I don't know if your current one does).
Some motherboard BIOSes pretty much had a "take it or leave it"
table of values, but I don't know if that extends to yours (The IBM
5170, for example gave you a limited set of choices).
Later motherboards knew about IDE (and ESDI) drives and would issue
the IDENTIFY command and read the drive geometry directly. Earlier
boards made to work only with MFM controllers obviously couldn't do
this.
You might try going to type 47 and hitting a few keys to see if the
BIOS will allow you to enter values--one of the old Award BIOS
versions allowed that. One of my Award BIOSes has NONE, 1-46 and
USER, for example.
It's also possible that another maker's BIOS will work, but you
haven't identified the chipset for the board, so it's hard to say if
anything readily available can be used.
Cheers,
Chuck
12 Jan
12 Jan
4:04 p.m.
Of course, if this is an IDE or SCSI drive and
controller, the=20
controller should have its own BIOS to work in a setup like this. If=20
it doesn't, you need to find one unless you're into BIOS hacking.
Uh? Had these multi-id controllers have their own BIOS extension? My
5:47 p.m.
On 12 Jan 2010 at 21:04, Tony Duell wrote:
I was wondering about that too. The IDE interface is essentially the
same at the I/O port level as the original PC/AT hard disk controller,
so it whould work with the standard motherboard BIOS.
Which, if you'll re-read my message, is what I essentially said--
particularly the thing about "BIOS hacking". The issue here is that
this Award BIOS apparently doesn't implement dynamic configuration of
the drive table. Hence, BIOS "hacking" would be in order--or,
preferably, a controller with a BIOS extension (they were numerous).
But also, as I mentioned, many IDE drives implement a 17-sector
translation mode by default, so one of the standard BIOS definitions
could conceivably be used.
The OP didn't mention if he had access to an EPROM burner or if he
felt comfortable patching code. I assumed the he felt negatively
about both questions.
Another alternative is to completely replace the BIOS with say, an
AMI one with user drive configurations. But the OP didn't state what
chipset his system was using and with a 386SX vintage system, that
could make a difference.
--Chuck
Of course, if this is an IDE or SCSI drive and
controller, the=20
controller should have its own BIOS to work in a setup like this.
If=20 it doesn't, you need to find one unless you're into BIOS
hacking.
Uh? Had these multi-id controllers have their own BIOS extension? My 
11 Jan
11 Jan
1:57 p.m.
On Sun, Jan 10, 2010 at 7:08 PM, Jan-Benedict Glaw <jbglaw at lug-owl.de> wrote:
My problem is that none of the 42 predefined HDD types
fits my HDD,
unfortunately not even the heads and sectors settings. (Missing some
of the HDD's space due to a too small cylinders value wouldn't be a
problem.)
I was going to suggest what chuck said. Find a scsi controller (or
something) that has its own bios.
I'm at work, so no time to google at the moment. But isn't there an
open-source bios project out there? Not sure how far along it is.
But you might be able to get a copy of that on an eprom that would
work maybe?
brian
5305
days inactive
5306
days old
7 comments
5 participants
participants (5)
-
ard＠p850ug1.demon.co.uk -
brianlanning＠gmail.com -
cclist＠sydex.com -
hp-fix＠xs4all.nl -
jbglaw＠lug-owl.de