Newsgroups: comp.sys.dec.micro
Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!sol.ctr.columbia.edu!cunixf.cc.columbia.edu!shibuya.cc.columbia.edu!lasner
From: lasner@shibuya.cc.columbia.edu (Charles Lasner)
Subject: Re: Rainbow EchoMail Digest
Message-ID: <1991Apr21.191505.17129@cunixf.cc.columbia.edu>
Summary: RX-50s, the naked truth
Keywords: RX50,hub-rings
Sender: Charles Lasner
Nntp-Posting-Host: shibuya.cc.columbia.edu
Reply-To: lasner@shibuya.cc.columbia.edu (Charles Lasner)
Organization: Columbia University
References: <Added.Ac3tRZu00jZdExC08Q@andrew.cmu.edu>
Date: Sun, 21 Apr 1991 19:15:05 GMT
Re: the never-ending RX50 media story.

I was the one quoted here again, so I'll reply about the Maxell diskettes.

Anytime you use hub-ringed diskettes in a 96 tpi drive, you are taking a
high-risk chance with your data.  The condition of the drive (dirt, age) as
well as the condition of the diskette, as well as its manufacture, dictate
your (slim) chances of long-term reliability of the data.

The problem is that those diskettes were designed to deal with the problems
of the widespread DD-type drives used on a million poor xt clones.  Many of
these crappy drives chew up the hub areas of the diskettes.  Any of you out
there remember what RX01s do to the 8" diskettes?  Any of you remember the
third-party add-on user-installable hub-rings?

The 48 TPI 40 track drives used with those DD disks have "non-thin-film"
heads, to coin a retroactive phrase.  This means that each track, which owns
1/48 " uses about 50% of this space for data recording.  The rest is guard
zone and tunnel erase zone.  This is also why you can read the DD disks in the
HD drives so easily; basically you can't miss :-)

The 96 TPI disks, like the RX50, use thin-film heads.  Not as snazzy as the
HD disk drives like TEAC FD-55GFR, as used in good AT-clones, but 96 TPI
none the less.  This of course means that each track owns 1/2 of what the
corresponding track owns in the 48 TPI drive, but it doesn't end there.

The thin-film heads use a shorter magnetic gap, which translates into far less
use of the media than would otherwise be predicted, about 1/5 as much.  So,
this means that 96 TPI drives record their data on about 1/10 of the media as
do 48 TPI drives.  The chances that the media's hub-rings can be grabbed 
differently each time now greatly increases chances of data errors.  Each time
the diskette is passed through a hub-chewing drive, the greater the likelihood
that a 96 TPI drive can't reliably use that diskette.  Further, some vendors,
possibly Maxell, will use somewhat thicker hubrings, in an attempt to produce
better results with the intended use on DD drives, which only worsens prospects
for use on 96 TPI drives.

I myself have had problems using 3M diskettes intended for DD usage.  The DSDD
media were formatted on HD 96 TPI drives in DD format for possible reading
on DD drives (360K).  They were unreliable on other HD drives due to the
slight wear on the hub-rings.  Registration errors were high because the HD
drive couldn't reliably grab the media in the same place.  We resorted to
pulling the media out and reseating it a lot to retrieve the data with many
retries.  If the same media were reformatted on the target drive, and not
removed and reseated, it read back fine.  Removing and reseating the diskette
made it flakey even on the drive that formatted and wrote it moments ago.
These same diskettes worked fine on TEAC FD55B drives (DD 360K), because the
hub-rings are not really that worn, just beyond the tolerance limit for 96 TPI
mis-use.

cjl (Charles Lasner) (lasner@watsun.cc.columbia.edu)


