--On Tuesday, December 09, 2003 20:45:45 -0800 "Peter C. Wallace"
<pcw_at_mesanet.com> wrote:
> On Tue, 9 Dec 2003, Eric Smith wrote:
>
> > The part of the ST-506 disk emulation that's of most concern to me
> > is whether anything needs to be done about write precompensation.
[...]
> > I'm guessing that it won't be
> > enough to cause reads to fail MOST of the time, but it will reduce
> > the timing margins enough to that it could potentially cause
> > occasional soft errors.
>
> Right, one option would be for the emulator to have a FM, MFM or RLL
> front end and actually decode the bitstream. Then the write-precomp
> could be undone...
But then you need to know what the bitstream is, which goes against
what we've been saying. And what about different encoding methods:
what happens when someone wants to emulate a drive on a system you've
not thought of?
Actually, I don't think you need to know the encoding scheme at all. I
believe the precomp (if/when applied) depends only on the sequence of
transitions in the data.
--
Pete Turnbull
Network Manager
University of York, UK
Received on Wed Dec 10 2003 - 06:23:52 GMT