Xref: utzoo comp.periphs:3442 alt.folklore.computers:9441
Path: utzoo!utgpu!cs.utexas.edu!sun-barr!olivea!samsung!uunet!virtech!jje
From: jje@virtech.uucp (Jeremy J. Epstein)
Newsgroups: comp.periphs,alt.folklore.computers
Subject: Re: info on Ampex SMD drive wanted
Summary: Variable # of heads
Message-ID: <1991Feb08.182858.5289@virtech.uucp>
Date: 8 Feb 91 18:28:58 GMT
References: <1117@idcapd.idca.tds.philips.nl> <1991Feb7.202812.1086@spcvxb.spc.edu>
Organization: Virtual Technologies Inc.
Lines: 30

In article <1117@idcapd.idca.tds.philips.nl>, wilko@idca.tds.PHILIPS.nl (W.C. Bulte) writes:
> I am looking for information on an Ampex SMD diskdrive type# DFR-996,
> model# MKM0800C006. This is an SMD drive, with (it seems) 80Mbyte fixed and
> 16Mbyte removable 14" platters.

This sounds like a unit we used at Perkin-Elmer.  It was sold in three
configurations: 13/13, 13/39, and 13/67 MB (removable/fixed).  [13MB
is the formatted capacity; 16MB unformatted.]  As is so
often the case, they were almost identical inside.  There were always
three platters in the fixed portion, but with 1, 3, or 5 read/write heads
(the sixth side was for timing).  Upgrading consisted of getting additional
heads installed.

And now for a bit of folklore.  The fixed heads were addressed as 0, 1,
2, 3, and 4.  From the SMD standpoint, it was one drive (not two), and
the removable unit used head 16 (0x10).  If you screwed it up (as many
disk drivers did) and tried to treat the removable drive separately
from the fixed drive, you ended up scribbling on the hard drive.  This
made for a very hard problem to fix...everything would be normal, and
then suddenly your file systems were horribly corrupted.  I fixed this
bug in the disk driver once, only to have someone break it again in
the next release...

These drives are also about the slowest I've ever seen to spin up...
it seemed to take minutes!
-- 
Jeremy Epstein
TRW Systems Division
703-876-8776
jje%virtech@uunet.uu.net