Path: utzoo!utgpu!jarvis.csri.toronto.edu!clyde.concordia.ca!uunet!tut.cis.ohio-state.edu!rutgers!cmcl2!acf5!mitsolid
From: mitsolid@acf5.NYU.EDU (Thanasis Mitsolides)
Newsgroups: comp.sys.ibm.pc
Subject: Re: MFM as an RLL drive?
Message-ID: <11330008@acf5.NYU.EDU>
Date: 23 Dec 89 19:18:00 GMT
References: <MLM.89Dec11110300@hector.cs.brown.edu>
Organization: New York University
Lines: 47

/* acf5:comp.sys.ibm.pc / phil@diablo.amd.com (Phil Ngai) /  5:08 pm  Dec 22, 1989 */
In article <423@hitech.ht.oz> clyde@hitech.ht.oz (Clyde Smith-Stubbs) writes:
|Would someone who knows what they are talking about please stand up
|and clear this up: to my knowledge the difference between MFM and RLL is NOT
|that the clock rate is different, but that RLL has less redundant information,
|i.e. to record one MFM bit takes two transitions while to record one
|RLL bit takes only 1 point something transitions. If this is the case, then
|the drive electronics surely are unaware of the difference? The only thing
|that should matter is the stability of the drive (i.e. motor flutter etc.)
|because of the reduced redundancy.

I think this is not exactly accurate.
From what I understand, MFM coding needs 0.75 pulses to encode a bit.
2.7RLL takes .5 pulses for a bit.
(on the average, since a bit is recorded in 2 clock tiks always)

Now, the less the number of pulses needed, the more the number of "silences".
2.7RLL means that the minimum number of silences between successive pulsses
is 2, while the maximum is 7.
MFM is equivalent to 1.3RLL and FM to 0.1RLL.

The trick is to keep the minimum physical distance between pulses constant
no matter what the encoding.
That decreases the length of a clock tik.
That means that the size of silences becomes smaller.
Now the controller must be able to recognise smaller silences.

Also, the clock (synchronization) is reset on pulses.
When the distance of pulses increases, the controller must be able
to count more and more silences accurately.

Once the controller is RLL and able to perform the task, the question is:
Is the HD able to record accurately the smaller silences or
in other words, the now finely varying distance of pulses?

I don't see that anything would be damaged if the HD cannot keep up to the task.

The above is what I understand about RLL encoding.
I hope it will clarify some issues.

Thanasis


-------------------------------------------------------------------------------
Internet: mitsolid@csd2.nyu.edu	         (mitsolid%csd2.nyu.edu@relay.cs.net)
UUCP    : ...!uunet!cmcl2!csd2!mitsolid
-------------------------------------------------------------------------------