Re: destorying the hard drive
"Franc Zabkar" <fzabkar@iinternode.on.net> wrote in message
news
p5ju3hvdtjsic3icon7vnoncaqht3o3on@4ax.com...
> On Mon, 24 Mar 2008 19:25:51 -0500, "Mike Y" <joe@user.com> put finger
> to keyboard and composed:
>
> >Hmm, just think back to HiFi tape days... I know, it's not directly
> >applicable to this, but from a purely analog point of view, did you
> >ever make a tape, then tape over and (on some equipment) if you
> >cranked up things hear a faint remnant? Usually that was just due to
> >cheap equipment that didn't fully erase, but in a 'master studio' if the
> >tape wasn't 'brand new' out of the box, it just wasn't used to master.
> >Sure, equipment evolved and it quickly became a non-issue. But
> >that was analog in an analog world. Residual data in magnetic drives
> >wasn't an issue, the issue was just making sure the latest data was
> >the data it extracted. Yeah, they did improve the recording technology
> >a bit, but almost all the 'improvements' were in the reading and
> >data separation areas.
>
> As you have already said, in the analogue domain you have degrees of
> magnetisation whereas in the digital domain you have flux reversals. I
> notice that analogue recording devices such as VCRs and compact audio
> cassette recorders have an erase head before the read/write head. When
> an erase circuit fails in a VCR, a ghost of the previous audio signal
> remains after new audio information is recorded. IIRC, such a failure
> does not significantly impact on the video signal. You should be able
> to see this for yourself if you hold the erase head away from the tape
> path during recording. I think the bias oscillator (via the
> record/play head) tries its best to randomise the magnetic domains but
> there is always a small audible remnant of the previous recording.
> Therefore the erase head is necessary to completely wipe the tape. In
> the digital domain, however, there is no erase head, only a read/write
> head. AIUI, a separate erase head is not required because the
> magnetisation is always "full on", ie either positive or negative.
> Zeros and ones are encoded as flux transitions, not flux levels, so
> there is no portion of the medium that is not fully magnetised. Since
> the circuitry is designed to detect these flux reversals, then it does
> not discriminate between a 100% one bit or a "noisy" 70% one bit. Both
> are seen as a logic 1.
>
> - Franc Zabkar
You hit it pretty square on the head. Mostly. In writing, the current is
not 'full on'. It's balanced with where the head is positioned on the
media.
Early drives had reduced write current controls that the controller was
responsible for triggering when the head was positioned halfway towards
the center of the spinning media. That means there were only two levels,
and in each 'zone' there would be areas where the current was reduced
under optimum. And always remember, the circuitry was designed to
give 'reliable readback' of data, it was NOT necessarily designed to
mandate magnetic saturation at all points on the media. If two much
current caused distortions and errors, and too little meant reduced S/N
ratios that could be compensated for with PLL recovery schemes, then
you can bet the MOST the current would ever be set for was before it
saturated and caused data errors and let the data separators do their
job on read.
And your statement about the 100% or 70% both being seen as a logic
one is dead on. Well, in FM recording. In MFM and later formats, the
bits are NOT 1 or 0, they are time marks, and the time marks determine
0 or 1. But the same idea applies.
Anyway, drives went from FM to MFM to RLL to ESDI (5 different
ESDI speeds if I remember right) to finally IDE, where the drive and
the controller are one. Why separate them? Once the drive and
controller are one, does it matter what's inside the bubble? Nowadays,
most drives lie worse than the last car salesman I DIDN'T deal with.
But it doesn't matter, all that matters to the user is the 'data packets'
that come out the IDE bus (IF it's an IDE drive, the SATA interface
if it's a SATA drive, or SCSI if it's SCSI) are correct and contain
the information desired. With the new drives the media is now
referred to, if I remember right, ZBR, or something like Zone Bit
Recorded. Basically it means the controller in the drive breaks the
media up into 'Zones' and changes things as it feels necessary. That
is, the number of sectors on a track changes, depending on the
velocity under the head (near the rim is faster) and even the RATE
of flux changes is adjusted for best performance and capacity.
The end result is that with modern drives, the user at the 'interface
point' to a drive is even FURTHER isolated from what goes on in
the sealed box call a hard disk drive, and reading anything from the
drive other than what the drive is designed to read and report would
be quite impossible.