NetNews Usenet Archive 1992 #30/NN_1992_30.iso/spool/comp/unix/bsd/10129

Text File | 1992-12-12 | 5.0 KB | 99 lines

Newsgroups: comp.unix.bsd
Path: sparky!uunet!usc!cs.utexas.edu!sdd.hp.com!caen!hellgate.utah.edu!fcom.cc.utah.edu!cs.weber.edu!terry
From: [e-mail] (A Wizard of Earth C)
Subject: Re: [386BSD] ARGH! 720K 3.5" floppy support, anyone?
Message-ID: <[e-mail]>
Sender: [e-mail]
Organization: Weber State University  (Ogden, UT)
References: <[e-mail]> <[e-mail]> <[e-mail]>
Date: Sat, 12 Dec 92 07:21:57 GMT
Lines: 87

In article <[e-mail]> [e-mail] (Dick Dunn) writes:
> [e-mail] (Jordan K. Hubbard) writes:
>>>...We need to assume that
>>>1.44MB also means 720K and that 1.2MB also equals 360K (though I'll
>>>admit to being a bit fuzzy on the later - do all 5.25" drives support
>>>360K?).
>
>5.25" HD drives all support *reading* 360k...but...
>
>[e-mail] (A Wizard of Earth C) writes:
>|A patch to "fix" this was posted some time ago.  It's questionable in it's
>|utility, especially with regards to writing the floppies; this is (I've
>|heard) because there is a difference in the rotational speed on high vs.
>|low density disks, which basically requires you to blow a register to
>|set the speed.  This makes it OK for reading, but unreliable at writing
>|and formatting ...
>
>Terry's got the wrong reason but definitely the write...er, right problem.
>1.2 MB drives can read 360 KB floppies, but cannot write them reliably.
>The problem is one of read/write head design--simplifying somewhat, a 1.2
>MB drive uses a narrower track than a 360 KB drive.  If you write on a 360
>drive, the 1.2 drive can read it just fine: it's got this narrow head
>cruising down the middle of a great wide data track.  But if you try to
>*write* a 360 floppy on a 1.2 drive, you end up writing new data down the
>middle of the track but leaving old data on either side of it.  A 1.2 drive
>*may* be able to recover this data--depends on drive design; I've had about
>half-and-half luck with the drives I've tested.  A 360 drive is quite
>unlikely to read it at all.  Remember, it's got a wider head, so it's
>getting a signal mixed from the old wide track and the new narrow track.

This is certainly a problem if you are mixing drives on reads an writes
on a single piece of media -- I certainly wasn't advocating that.  The
problem I was referring to was a spindle-speed problem in some drives.
TEAC drives (1.2M) tend to support this correctly.  With have a couple
of machines with 360K Mitsubishi drives that fail with the machine in
"turbo" mode because of the change in spindle speed (stupid drive gets
it's clock from the wrong place).

I remember an earlier patch (March?) that I unfortunately didn't save that
dealt with blowing the correct controller registers in the device driver
and formatting correctly.

The generally accepted procedure for disk production of 360K disks on a
1.2M drive on a Xenix box involves formatting and writing the disks on
that box with the 1.2M drive.  As long as the disk isn't read on a 360K
drive *after* being rewritten by a 360K drive and then rewritten *again*
by a 1.2M drive (generally not a problem on distribution disks), you
won't experience the hysteresis problems you describe.

The problem (as it was explained to me) is that on large reads, with the
write clock and the read clock not *exactly* matched, there is basically
the same effect the record companies have benn trying for by getting
the DAT record frequency to be anharmonic to the CD sampling frequency
-- signal drop out on destructive cariier interference.

If you blow the clock register, it should be safe.

The point is not to interchange disks between real 360K drives and 1.2M
drives, but rather to be able to master 360K disks using 1.2M drives,
and to enable use of lower cost (6-8 cent) low density disks, which one
might have lying around.  The other use is to import 360K disk contents
directly, which, as was pointed out, is not effected by the hysteresis.

I've actually mastered more that 5000 disks this way for small scale
distributions when the company I used to work for consisted of the owner,
me, and one sales type.

For disks formatted on 360K drives, and read/written or written/read on
360K and 1.2M drives running at 360K, this *is* a potential means of data
transfer, as long as you *guarantee* that there is *never* a 1.2M drive
write after a 360K drive write (including the writes during format).

For disks that are never rewritten (mastered distribution disks) this is
never a problem.


                    Terry Lambert
                    [e-mail]
                    [e-mail]
---
Any opinions in this posting are my own and not those of my present
or previous employers.
--
-------------------------------------------------------------------------------
                                        "I have an 8 user poetic license" - me
 Get the 386bsd FAQ from agate.berkeley.edu:/pub/386BSD/386bsd-0.1/unofficial
-------------------------------------------------------------------------------