11 Oct
1999
11 Oct
'99
10:35 p.m.
Okay, one final question before I drop this topic
forever (or until tomorrow, whichever comes first).
How is it possible that DD media could be of such
poor quality that it can't (reliably?) do 96 TPI,
while still being just fine at 48 TPI?
I mean, it is recording something like 3000 bits per
inch along each track, right? How could that same
media not have enough resolution to keep the bits of
adjacent tracks separate at less than 100 per inch?
Can somebody who believes this can happen give me a
mental model of what is going on there? I mean, in
terms of physics or geometry or mechanics or anything
measurable and specific.
It sounds to me like it must be the drive, and not
the media, that limits the number of tracks to
anything below 1500 TPI.
Thanks,
Bill.
12 Oct
12 Oct
4:25 a.m.
Hi Bill,
First off, 3:30 in the morning is too early to be reading mail. Go
to bed :)
Most of my knowledge is from hard disks, but should apply OK.
Basically, magnetic media is a tradeoff between coercivity (how
easy it is to flip the bits) and write current (how hard you have
to try and flip the bits). A material with high coercivity resists
being re-magnetized in a different direction. The more write current
you push through the head increases the electro-magnets power, but
also increases its SIZE. More write current means you are writing
bigger transistions. Remember, you can't read a 'north' as a one,
you can only detect when the 'north' changes to a 'south' (a
transistion).
Ok, so 48 TPI media has low coercivity, so a 96TPI head with higher
write current will cause the adjacent tracks to erase because of
the magnet fringes from the head. These magnetic fringes aren't
strong enough to effect 96TPI media because it has a higher
coercivity.
Finally, the head controls the size of the track. The width
of the magnetic poles sets the width of the track written,
plus a little fringing to provide track erasing to clean
the previously written data in case of track misalignment.
The bit density is controlled by how fast the head is moving
w/respect to the media, and how fast you can toggle the write
current. Think of it this way: if you have a stencil .1x.1 and
can fill it in in either red or blue. put the stencil down and
fill it in red. now move the stencil 1/3000 inch. put the
stencil down and fill it in blue. repeat a LOT. Now you have
10 TPI at 3000 bits/inch. Make sense?
I will look for a picture from a class I took many years ago
and put it somewhere. I think I still have a geocities webpage :)
clint
On Tue, 12 Oct 1999, Bill Yakowenko wrote:
Okay, one final question before I drop this topic
forever (or until tomorrow, whichever comes first).
How is it possible that DD media could be of such
poor quality that it can't (reliably?) do 96 TPI,
while still being just fine at 48 TPI?
I mean, it is recording something like 3000 bits per
inch along each track, right? How could that same
media not have enough resolution to keep the bits of
adjacent tracks separate at less than 100 per inch?
Can somebody who believes this can happen give me a
mental model of what is going on there? I mean, in
terms of physics or geometry or mechanics or anything
measurable and specific.
It sounds to me like it must be the drive, and not
the media, that limits the number of tracks to
anything below 1500 TPI.
Thanks,
Bill.
5:49 a.m.
Ok, so 48 TPI media has low coercivity, so a 96TPI
head with higher
write current will cause the adjacent tracks to erase because of
the magnet fringes from the head. These magnetic fringes aren't
strong enough to effect 96TPI media because it has a higher
coercivity.
???? reread that and see if it makes sense to you...
There are two media for 96tpi, one for DD and lower and is the SAME
magnetic material as 48tpi. THere is also the 1.2mb media and this is
VERY DIFFERENT and incompatable with any other drive or density.
So if we ignore 1.2mb media and the oddbal spindle speeds and data rates
that go with it we come down to one media (softs sector) and 6 different
drives over the years. They are:
48tpi single sided (sa400)
48tpi double sided (sa450)
96tpi single sided (teac Fd55E or DEC RX50)
96tpi double sided (teac FD55F)
100tpi single and double sided models (micropolus I think)
*there may have been others mostly forgetable. Apple and other GCR
unique drive not included as they are NOT standard floppies as often
they were standard drives with differences introduced.
They all record data at roughly the same linear density (FM at 125kbs or
MFM at 250kbs all rotate at 300rpm) and differ largely on the number of
sides, width of track and absolute position for tracks greater than 000.
Allison
12:24 p.m.
Allison wrote:
There are two media for 96tpi, one for DD and lower
and is the SAME
magnetic material as 48tpi. THere is also the 1.2mb media and this is
VERY DIFFERENT and incompatable with any other drive or density.
Not quite. The "1.2mb" media is compatible with Apple Twiggy diskette
drives, though the jacket cutouts are different.
3:55 p.m.
On Tue, 12 Oct 1999 allisonp(a)world.std.com wrote:
Ok, so not as much as at 7:00 this morning... HD media has a higher
coercivity, thus requires more write current to put a bit on the
disk. If you use that much write current with LD media (lower
coercivity) you will get a larger bit. The larger bit will partially
overwrite adjacent bits, leading to low amplitude, and read errors.
Ok, so 48 TPI
media has low coercivity, so a 96TPI head with higher
write current will cause the adjacent tracks to erase because of
the magnet fringes from the head. These magnetic fringes aren't
strong enough to effect 96TPI media because it has a higher
coercivity.
???? reread that and see if it makes sense to you...
There are two media for 96tpi, one for DD and lower
and is the SAME
magnetic material as 48tpi. THere is also the 1.2mb media and this is
VERY DIFFERENT and incompatable with any other drive or density.
So if we ignore 1.2mb media and the oddbal spindle speeds and data rates
that go with it we come down to one media (softs sector) and 6 different
drives over the years. They are:
48tpi single sided (sa400)
48tpi double sided (sa450)
96tpi single sided (teac Fd55E or DEC RX50)
96tpi double sided (teac FD55F)
100tpi single and double sided models (micropolis I think)
Ok, I wasn't aware of any 96tpi drives except the HD ones... Were
they ever used in the PC marketplace, or was it mostly a DEC thing?
After re-reading my uVAX manual about the RX50, I agree it is a
96tpi single sided drive, though the manual says: "Use only formatted
RX50 diskettes, available from DIGITAL or its licensed distributors"
So there are four drives using the same media:
SSDD 180K,48tpi,40tracks/side
DSDD 360K,48tpi,40tracks/side
RX50 360K,96tpi,80tracks/side
???? 720K,96tpi,80tracks/side - What was (is) this called? DSQD?
And the HD drive:
DSHD 1.2M,96tpi,80tracks/side - can read older media, but writing
is unreliable (head isn't wide
enough to erase the whole old track)
If Bill indeed has DSDD media, bulk erasing the media before applying
it to the 96tpi (non-HD) drive should work fine. This erases the old
data from between the new (half as wide) tracks, so it doesn't
interfere with reading.
Allison
clint
4:44 p.m.
On Tue, 12 Oct 1999, Clint Wolff (VAX collector) wrote:
Ok, I wasn't aware of any 96tpi drives except the
HD ones... Were
they ever used in the PC marketplace, or was it mostly a DEC thing?
http://www.xenosoft.com/fmts.html#720K
will give you a list of some of them.
If Bill indeed has DSDD media, bulk erasing the media
before applying
it to the 96tpi (non-HD) drive should work fine. This erases the old
data from between the new (half as wide) tracks, so it doesn't
interfere with reading.
That seems to be working for almost everyone except Tony!
--
Fred Cisin cisin(a)xenosoft.com
XenoSoft http://www.xenosoft.com
2210 Sixth St. (510) 644-9366
Berkeley, CA 94710-2219
5:33 a.m.
How is it possible that DD media could be of such
poor quality that it can't (reliably?) do 96 TPI,
while still being just fine at 48 TPI?
???? Nonsense question, no context. I have no difficulty
nor is there any reason to expect difficulty with 96tpi DD
ops. I have expereinced using formerly formatted media of
96 or 48TPI that REQUIRED bulk erasure to be usable. I believe
that intertrack noise due to differnt track widths are why.
I regurally use PC360k, Vt180, RX50 and Visual1050 media and they
are 48 TS, 48 SS, 96ss, 96ss/TS all running at DD data rates.
media not have enough resolution to keep the bits of
adjacent tracks separate at less than 100 per inch?
The recorded track density is not he same as the TRACK WIDTH
or the TRACK position.
Can somebody who believes this can happen give me a
mental model of what is going on there? I mean, in
terms of physics or geometry or mechanics or anything
measurable and specific.
For all DD the number of bits per track will be the same (generally)
so whats different? The spacing of the tracks and their width.
Both 48 and 96 tpi drives write their trace in the same ~1" band on the
media.
The older 48tpi drives sliced this into 40 bands of .025"(approx) wide.
To allow for error the head writes and area less than that.
Not the newer 96tpi drives also use that same ~1" but put 80 bands
in that space, they are .0125(approx) wide. The head is narrower to
align and fit in that space.
The old 48tpi heads literally straddle two of the 96tpi tracks if you go
from 96 down to 48. Worse is that if you dilberatly
skip a track to
try and create a "48tpi compatable disk" you still write
a track that is
less than half the width. When you read it there is less than half the
magnetized area and half the signal. This promotes errors.
Going the other way, works better as the narrow 96tpi head gets to read a
track wider than itself, lots of signal. Can't write though as the head
is narrower than the old wide track, this means writing with a narrower
head can only erase a tunnel in the data previously written. read it
back and likely it's ok only less the drive is a different one and
alingment is off ever so slightly allowing both old data nd new data to be
visible, errors result. Try to read that disk on a 48tpi drive and the
wide head reads old data and new data, results will be poor at best.
It sounds to me like it must be the drive, and not
the media, that limits the number of tracks to
anything below 1500 TPI.
It's the .667 mil (0.0006666) wide head and the positioner.
Yes, specifically head width. The media also limits things some
as very narrow head will tend to dig into the media and the width of the
track also defines signal level. Zip disks are high track density
floppies using special cobalt media for high bit density and also high
track density. They aslo have servo info on the media to aid in finding
the rather narrow tracks.
This is why 135 tpi has been the limit for absolute mechanical positioners
like floppies. At some point the track is so narrow you have to be able
to hunt for it and stepper based positioners are inexpensive but lack
resolution (and speed) for that task. The next generation will be some
form of voicecoil and the media will have embedded servo (Zip, LS120,
Jaz and friends). At some point you bridge the differences between floppy
and hard disk.
There are really several questions in your posting.
Allison
8:36 a.m.
On Tue, 12 Oct 1999, Bill Yakowenko wrote:
Okay, one final question before I drop this topic
forever (or until tomorrow, whichever comes first).
How is it possible that DD media could be of such
poor quality that it can't (reliably?) do 96 TPI,
while still being just fine at 48 TPI?
Of my own experience, I do not believe that it is. I regularly use
generic unformatted 5.25" DD disks at 96tpi without problem.
- don
I mean, it is recording something like 3000 bits per
inch along each track, right? How could that same
media not have enough resolution to keep the bits of
adjacent tracks separate at less than 100 per inch?
Can somebody who believes this can happen give me a
mental model of what is going on there? I mean, in
terms of physics or geometry or mechanics or anything
measurable and specific.
It sounds to me like it must be the drive, and not
the media, that limits the number of tracks to
anything below 1500 TPI.
Thanks,
Bill.
9045
days inactive
9045
days old
7 comments
6 participants
participants (6)
-
allisonp＠world.std.com -
cisin＠xenosoft.com -
donm＠cts.com -
eric＠brouhaha.com -
vaxman＠uswest.net -
yakowenk＠cs.unc.edu