hutch@net1.ucsd.edu (Jim Hutchison) (04/06/89)
I have been reading about read/write optical disks after finding I knew
nothing about them, now I know a little. A few things came to mind,
and this seems to be a good place to check out the "facts".
1) In the magazine I was reading (its around here somewhere), the type
discussed was one where the media was heated with a laser in a magnetic
field and then held the field. Writing 0's and 1's was a process of
setting the field for 0's, hitting the track (zeroing it), inverting
the field and writing the 1's. With this in mind, does anybody switch
which step they do first to cut some of the delays in settling/stablizing
of the polarity? Is the rotational delay long enough that this is not
even an issue (delay >> settling time)?
2) If I moved back to balmy El Cajon (which sports 100 degree weather
right now), would the constant heat be enough to allow entropy to have
an effect on the legibility of the datas state? Similiar to the way
wood will "spontaneously" if kept warm enough long enough?
3) Is there usually some kind of interlock to make sure that the field
actually changes polarity everytime you write a sector? Could be real
nasty if it didn't.
4) With this technology, is it possible to do an inexpensive read-after-
write?
/* Jim Hutchison {dcdwest,ucbvax}!ucsd!net1!hutch */shs@uts.amdahl.com (Steve Schoettler) (04/08/89)
In article <1589@ucsd.EDU> hutch@net1.UUCP (Jim Hutchison) writes: >I have been reading about read/write optical disks after finding I knew >nothing about them, now I know a little. A few things came to mind, >and this seems to be a good place to check out the "facts". > >1) In the magazine I was reading (its around here somewhere), the type >discussed was one where the media was heated with a laser in a magnetic >field and then held the field. Writing 0's and 1's was a process of >setting the field for 0's, hitting the track (zeroing it), inverting >the field and writing the 1's. And there's a third pass of verification. > With this in mind, does anybody switch >which step they do first to cut some of the delays in settling/stablizing >of the polarity? The magnetic materials involved have anisotropic curves, i.e., it is easier to magnetize them along some crystal direction than others. In iron, the easy directions are cube edges, in nickel the easy directions are body diagonals. In both cases, rotating the magnetism from one easy direction to another involves roating through a hard direction, but the energy required (and therefore time) is the same, whether going from 0 to 1 or 1 to 0. (The plane of rotation of the plane is actually quite small, ~0.30) > Is the rotational delay long enough that this is not >even an issue (delay >> settling time)? As of about now, I think the rotational delay is about the same as the settling time (on the order of 35ms). >2) If I moved back to balmy El Cajon (which sports 100 degree weather >right now), would the constant heat be enough to allow entropy to have >an effect on the legibility of the datas state? Similiar to the way >wood will "spontaneously" if kept warm enough long enough? Of course, there are always particles of any given energy at any positive temperature, but in this case, the probability is such that data will be there for at least 1000 years. Actually, conventional magnetic disks have a much lower coercivity at room temperature, and are even more volatile. But the coercivity decreases sharply with increasing temperature, and it turns out you can change the data at around 75-100 degrees Centigrade. I think the real problem involves the stability of the transition layer. Terbium is almost always used, and it unfortunately has a high affinity for oxygen. Even though there is a protective coating on magneto-optic disks, the media slowly degrades. One NeXT salesman told me that they thought data on such a disk would only last 5 years or so. Geoffrey Bate, a scientist at a Santa Clara semiconductor consortium (sorry, I forget its name) estimates data would be good for probably at least ten years, but is not likely for >30 years with current magneto- optic technology. -- Steve Schoettler shs@uts.amdahl.com {sun,decwrl,pyramid,ames,uunet}!amdahl!shs Amdahl Corp., M/S 213, 1250 E. Arques Ave, Sunnyvale, CA 94088