Very Computer

Here is an interesting problem for some of you PC oldtimers ;-)

I need some help with an old PC I am upgrading to run Linux.  It used to
be a 286, now it is a 486.  I wanted to put in a decent sized IDE hard
drive in the machine to replace the 40 meg seagate ST-251 MFM hard drive
with an unknown model # Western Digital controller, but the market value of
small IDE's is outrageous for its true worth.  So I bought another ST-251 to
put side by side.  I tested it by pulling out the old hard drive and put
in the new hard drive.  Worked great.  Now putting them together is where
the problems come into play.

I connected the two drives with the wide ribbon cable containing the 2
slots.  I adjusted the jumpers in the back of the drives according to the
specs listed from Seagate, and with some help from the drive lights in
the front of the drive (Yes the LED's came built-in), got the correct
settings for the jumpers (I believe). The one connected to the middle of
the cable has drive select 1 jumber connected, the drive at the end has
the drive 2 jumper enabled.  From now on I will refer to each HD
as Drive A1 and Drive B1 so as not to be confused with the floppy drives.

I connected the smaller cable from drive A1 to the drive 0 slot in the
controller, and I did the same with the other cable on Drive B1 going to
the controller's Drive 1.  I set up everything in the Bios to recognize
both drives.  Booted up the computer -- error cannot find drive D:, hit F1
to continue.  When I hit F1, drive A1 boots up (and, of course cannot
find the other HD).  Hmm...

So I switched the smaller (data, I believe) cables -- Drive A1 is now
connected to drive 1 on the controller, and drive B1 to controller's Drive
0.  Same error, and B1 now boots up, with the other HD nowhere to be found.  
Thinking that maybe the data cable connected to the controller's drive 1
slot may be bad.  I switched the physical cables with no success, same
result.

I need some help here.  My first instinct is that the second drive needs
to be low-level formatted for the controller to recognize it, but it
doesn't make much sense, since it recognizes each drive if it is
connected to drive 0.  If it does need low-level formatting, will the
lowlevel formatter in the BIOS be OK for this, and what about
auto-interleave.  I've not worked with MFM much -- it's old and
complicated, and difficult to find documentation.  Any help would be greatly
appreciated.  Sorry about the length of the article, I wanted to present
as much info needed to be clear.

Thanks,

Mike

I just checked my Seagate docs... and it says that if you have a cable
with a twist in it (a section of 5 or 7 wires is saparated and
twisted. I forget exactly how many. It is different for floppy cables
than for dard drive cables) ... then Drive C should be jumpered as
Drive-Select 2. Drive D is always Drive Select 2.

(If the cable does not have a twist in it, jumper Drive C as
Drive-Select 1.)

Attach drive C to the end of the cable, drive D to the middle of the
cable.

Also check for terminating resistors... only the last drive should
have one ("LAST" means the drive at the end of the cable. In this
case, that is Drive C:)  Although I have seen setups where both drives
have them and they still work. Terminating resistors seem to affect
the drive-select jumpers.  The terminating resistor is on the bottom
of the drive, flat with about 7 pins on it. Remove it from drive D,
which is the drive connected to the middle of the cable.

(When/if re-installing the terminating resistor, note that pin-1 on
the resistor is denoted by a dot or numeral-1. A square pad on the
disc drive board indicates pin-1 at the socket.)

Quote:>I need some help here.  My first instinct is that the second drive needs
>to be low-level formatted for the controller to recognize it,

Quote:>connected to drive 0.  If it does need low-level formatting, will the
>lowlevel formatter in the BIOS be OK for this, and what about
>auto-interleave.  

Quote:> I need some help here.  My first instinct is that the second drive needs
> to be low-level formatted for the controller to recognize it, but it
> doesn't make much sense, since it recognizes each drive if it is
> connected to drive 0.

If the larger cable is twisted (like most floppy cables, except
that twisted pins are not the same), both drives should have the same
jumper settings, and it might be DS2 for both or DS1 for both.

Quote:> If it does need low-level formatting, will the
> lowlevel formatter in the BIOS be OK for this, and what about
> auto-interleave.  I've not worked with MFM much -- it's old and
> complicated, and difficult to find documentation.  Any help would be greatly
> appreciated.  Sorry about the length of the article, I wanted to present
> as much info needed to be clear.

 got the correct

Quote:> settings for the jumpers (I believe). The one connected to the middle of
> the cable has drive select 1 jumber connected, the drive at the end has
> the drive 2 jumper enabled.

If you have more questions, email me. I have lots of useless knowledge about
mfm drives!
-Paul

[followup-to newsgroups trimmed a bit]

Quote:>I connected the smaller cable from drive A1 to the drive 0 slot in the
>controller, and I did the same with the other cable on Drive B1 going to
>the controller's Drive 1.  I set up everything in the Bios to recognize
>both drives.  Booted up the computer -- error cannot find drive D:, hit F1
>to continue.  When I hit F1, drive A1 boots up (and, of course cannot
>find the other HD).  Hmm...

>So I switched the smaller (data, I believe) cables -- Drive A1 is now
>connected to drive 1 on the controller, and drive B1 to controller's Drive
>0.  Same error, and B1 now boots up, with the other HD nowhere to be found.  

Quote:>Thinking that maybe the data cable connected to the controller's drive 1
>slot may be bad.  I switched the physical cables with no success, same
>result.

The length of the cables won't be the problem, I've once used MFM
cables 6 feet long :-)

It should not need low-level formatting. The drive light doesn't go
on before the BIOS says "drive D not found"? If not, then the formatting
is not (yet) the problem.

Paul Slootman
--
Murphy Software, Enschede, The Netherlands        | ... It beeped and said

Thanks to everyone for their help.  I was able to fix the
problem.  I had to remove the terminating resistor on the middle drive,
and I think I changed a jumper as well.  I was just happy it worked.  I
didn't get into PC's until the IDE drive era, so this MFM stuff was all a
mystery.  Your help was greatly appreciated.

Mike

Because there is a twist in the cable, the jumper settings on both drives
should be THE SAME.

The wider ribbon cable is virtually identical to a floppy controller cable.
The narrower cable is, as you say below, the (8-bit wide) data bus cable.

With both drives set to the same jumper position, you should get one selected
as "C" and one as "D" whichever common jumper setting you use.  You would have
to experiment with the data cables to get them the right way round.

If both disks have been recognised OK individually, you certainly do not need
to do a low-level format.  But you DO need to manually program the BIOS with
the disk parameters (tracks, cylinders, heads).

Incidentally, you _could_ easily replace the MFM drives with an IDE drive: you
have to either remove the MFM controller card (or disable it, if it's built onto
the motherboard), install an IDE controller card (not many pounds or dollars)
and an IDE drive up to 504MB (that's 504 x 1024 x 1024 bytes).  You could get
a larger IDE HD by using OnTrack DiskManager or similar.

--
-----------------------------------------------------------------
 Frank Burgum                                  Navigation Section
           Philips Semiconductors Systems Laboratory, Southampton


-----------------------------------------------------------------

Both drives have worked when plugged into the controller's drive 0
connector using either data cable.  This means that both drives are
working, both data cables are OK and the low level formatting seems
to be compatible with the controller.  

You didn't say whether the control cable has a twist or not.  If
there is NO twist, then the drive whose data cable goes to the
controller's drive 0 should be jumpered for drive select 1 and the
drive whose data cable goes to ctl drive 1 should be jumpered for
drive select 2.  In this configuration, either drive should work
with either connector on the control cable.  There could be a
problem with the cable's drive select 2 wire so you might try
another cable if you have one.

If the cable has a twist, then both drives should be jumpered for
drive select 2 and the connector with the twisted wires must go to
drive 0 and the connector with the untwisted wires must go to drive
1.

It sounds as if you have not low level formatted either drive and
you didn't say whether the controller is an 8 or 16 bit card or
whether it has an on board BIOS.  If the jumpers and cables all
appear to be OK, I think I would try to low level format drive 1
and see if the controller can talk to it.  If it's a 16 bit card,
it probably does not have a BIOS and you should use the formatter
in the system's BIOS.  If it's an 8 bit card, it probably has its
own BIOS and you should use that for low level formatting by running
DEBUG and typing 'g=c800:5'.  If the low level formatter can talk
to the drive, then try FDISK and FORMAT to complete the job.

NOTE: the Western Digital controller on my old XT included a low
level format option to create two logical disks on drive 0 (this
got around the old 32M byte limit that existed prior to DOS 3.3).  
If this option was used on drive 0, the controller would not let
the system see drive 1.  Is your drive 0 divided into two logical
drives?

Let me know if any of this helps.  If not, fill me in on the missing
data that I mentioned and I'll try to think of something else.

Tom
--
Thomas M. Root

McMinnville, OR