Xref: utzoo comp.sys.ibm.pc:15651 comp.sys.misc:1436
Path: utzoo!attcan!uunet!husc6!cfa!ward
From: ward@cfa.harvard.EDU (Steve Ward)
Newsgroups: comp.sys.ibm.pc,comp.sys.misc
Subject: Re: RLL Technical Details (long) (was Re: RLL- why it is hard on drives)
Summary: there may be an overlooked fact...
Message-ID: <1026@cfa.cfa.harvard.EDU>
Date: 17 May 88 19:58:23 GMT
References: <1255@kodak.UUCP> <638@mccc.UUCP> <216@octopus.UUCP> <650@mccc.UUCP> <661@mccc.UUCP>
Organization: Harvard-Smithsonian Ctr. for Astrophysics
Lines: 42

In article <661@mccc.UUCP>, pjh@mccc.UUCP (Pete Holsberg) writes:
> In article <218@octopus.UUCP> pete@octopus.UUCP (Pete Holzmann) writes:
> ...If you read all the way through this, you will (hopefully) understand WHY
> ...RLL works/doesn't work depending on the configuration you set up. You will
>....
> 
> Misled?  You think that the guys who told me that they were unable to
> MFM-format a plated media drive that they had run under RLL were
> kidding?  Frankly, I didn't initially believe them either, but enough of
> them said the same thing that I got rid of my ST-225 asap.  Maybe it's
> another hummingbird case: theoretically, there's nothing that an RLL
> controller can do to prevent a disk from being reformatted, but in these
> cited cases, it happened.  Probably in conjunction with a powersupply
> problem or some such thing, eh?
> 

MFM controllers may not work with RLL drives and vica versa, at least
with ST506-type disk drive interfaces.  This may account for some of
above commentary and problems.

This is so because the ST506 drive requires the data separator to be
on the disk drive controller - the data separator is not on the disk
drive.drive itself.  The controller interfaces to the data read/write
electronics in a fairly direct fashion.  The controller has to generate
an encoded data+clock signal and making assumptions (write
precompensation, read/write waveform characteristics, esp. pulse discr.
behavior) along the way about a drive they didn't manufacture.  This is
further complicated by the fact that the RLL drive is optimized for a
read/write pulse rate correlated to a bit rate of 7.5MBPS while the
MFM drive is 5.0MBPS.  It becomes a serious problem to match the
on-controller data separator to the on-disk read/write electronics
in any case, but especially in any universal MFM+RLL method.  It is not
impossible, just very hard.  This means you might see combinations
of controllers and disk drives that work better/worse together and
a drive that might work RLL and MFM with particular controllers but
not with others.  For best results, talk to the drive manufacturers
and use controllers they recommend.  The real solution is to place the
data separator on the disk drive so that the disk drive read/write
electronics can be optimized and matched with the data separator.
A SCSI I/O disk drive presumably has the electronics matched since
a single vendor makes the drive and electronics.  ESDI drives place
the data separator on (usually inside) the disk drive.