Atari hard drives? The actual Atari Answer :) - test-drb@ccmp.vtda.org

3 Jun 1998

Okay gang. I just can't take it, I gotta help this guy become an atarian. :)
Okay, this is what; you need. You need an ATSCI to SCSI converter
controller. The most common for the ST series is made by ICD. They can still
be had from ICD but they are around 79.00.  There is an Atari dealer, yes,
dealer, amazing in my area. [KCMO] Their home page is
www.systemsfortomorrow.com.  They are a great bunch of guys and I got my ICD
SCSI controller for them for a 1040 Swap. [I had two]. But he has a few used
ones and can get you one if you want.  This machine is very worth spending
the money to get a HD for.  My 1040ste I have upgraded to 4 megs, it has a
14.4 modem and a quantum 105lps scsi HD on.  It will currently connect to
the internet, and i mean with PPP an PAP authentication, will do IRC, FTP,
Popmail and graphically surf all via freeware. Please let me know if I can
be of any assistance.

Bill Girnius

-----Original Message-----
From: Tony Duell &lt;ard(a)p850ug1.demon.co.uk&gt;
To: Discussion re-collecting of classic computers
&lt;classiccmp(a)u.washington.edu&gt;
Date: Wednesday, June 03, 1998 6:26 PM
Subject: Re: Atari hard drives?

...
 >
> > MFM and RLL are encoding schemes. They define how to turn a stream of
0's
...
    and 1's into the magnetic patterns on the disk.
They have nothing to do
 with how a drive talks to its controller 
 Of course, I was just using the vernacular... (About like "CMOS setup" -

I don't normally mind - in fact I often misuse the terms myself. But whem
we're talking about hard disk operation, it helps to be exact.

CMOS setup is less worrying. After all you are setting up parameters
which are stored in CMOS RAM (part of the real time clock chip). I do
object to people who don't know the difference between the CMOS RAM that
stores the BIOS setup parameters and the BIOS ROM itself.

  you didn't make any comments about that one
though.  :-))  But I didn't
 remember that it was called ST506.  Sounds like a Seagate part number.
 Was it a PC hard drive, or even earlier? 
The SR506 interface is named after the ST506 drive (5 Mbytes) which was
(AFAIK) the first drive to use this interface. Actually, the right name
should be ST412 (which was a 10 Mbyte drive that was used in the original
IBM XTs, etc), since there are minor differences. But if I called it an
ST412 interface few people would understand me.

> > You're probably thinking of the servo bursts - the signals that keep
the
...
 > > heads on a track. Some drives did use a
particular side of one of the
> > platters for these. Other drives 'embedded' them in the sector headers
on
...
    the data platters. The latter is not common on ST506
interfaced drives,
 but is common on SCSI/IDE drives 
 Hmmm, interesting.  So low-level formatting doesn't rewrite these? 
Absolutely not. That's one reason why bulk-erasing hard drives can ruin them.
...
 The only way to rewrite the servo data is to open up
the disk in a clean
room, afix a position transducer (laser interferometer !) to the head
arm, position the heads accurately where you want them using a special
electonic system linked to the positioner coil and the transducer and
write the servo bursts. I know of nobody who has the equipment to do that.

  That would imply that the servo bursts are not
involved in determining
 sectors at all.  I used to use an RLL card with a couple of formerly 
Correct. On embedded servo drives, the number of sectors/track is fixed,
since you want the servo bursts in the headers, and not in the middle of
user data. But on servo surface drives (with the 'wasted' surface), the
servo data is pretty much continuous and doesn't have anything to do with
the sectoring

  MFM drives, and it made more sectors per track,
so I thought the
 wasted platter had something to do with that. 
No, it's similar to the fact that I can take a 5.25" floppy and format it
in my TRS-80 model 1 with 10 sectors (88K) or in my Apple ][ with 16
sectors (143K). Both machines have 35 tracks, in the same position.

> > Are you sure: While almost all clone controllers have a formatter routine
...
    at C800:5, 
 Yep, that's the one. 
Well, unless you move the BIOS to CC00:0000 or somewhere...

>
> > I couldn't find it in the original IBM XT hard disk BIOS.
>
> Hmmm.  Well I think I remember using a full-length IBM controller with a
> 30 meg drive, and only being able to format 10 megs of it because it was
> the XT controller designed for the 10 meg hard drive.  But I don't
remember
...
 > if I used debug to format it, or something else. 
This is my mom's machine
...
 > which she never uses, so one of these days I can
replace it with something
...
   more useful
and refesh my memory about its contents.  I'm going to try
 to restore it to original condition since it is so close already. 
It's not in the BIOS source in my Techref. The first few lines of that
source are :

0000 55     DB 055H ; Generic BIOS header
0001 AA     DB 0AAH
0002 10     DB 16D  ; ROM length byte (ARD)
0003    DISK_SETUP  PROC FAR
0003 EB 1B  JMP SHORT L3
0005 3530.. DB '5000059 (C)COPYRIGHT IBM 1982' ; Copyright notice

So on the true-blue IBMs, there's a copyright notice there, not a jump
to the low-level formatter.

>   _______                      KB7PWD @ KC7Y.AZ.US.NOAM ecloud(a)goodnet.com
...
 >  (_  | |_)   Shawn T. Rutledge    on the web:
http://www.goodnet.com/~ecloud
...
 >  __) | |
\__________________________________________________________________
...
   * quantize the
universe * 808 State * virtual reality * mad science *

-tony 