6 Apr
2016
6 Apr
'16
1:15 p.m.
I had one of those Japanese Koan moments recently when someone asked me
"Why do floppy disks stop working?" and I realised I... didn't actually
know. I thought I'd throw it to the group and get some theories/proofs.
Let's work on the assumption we're talking about 5.25" and 3.5" disks.
Several guesses:
- Repeated use slowly wears away the magnetic media layer on the mylar.
- When left in an unprotected state, or a poor environment, damp, mold and
dust can damage the surface, either degrading the magnetic layer or
causing the gap to shrink enough that the drive head physically damages
the disk?
- Quantum fluctuations in the state of the universe, caused by millions of
mostly non-interacting particles passing through a disk in any given
minute, alter the magnetic spin of the ferric atoms causing gradual data
loss over time (mostly tongue-in-cheek)
- Given the lack of use of most floppy drives they themselves pick up
'gunk' and on first reading a diskette after a long time of disuse damage
it.
It _seems_ like when you put a 3.5" disk down for ten years and pick it
back up, a disk that used to work fine no longer does. Of course, after
ten years, it could be your own memory that's failed.
Dare I ask, what's the consensus?
- JP
6 Apr
6 Apr
8:20 a.m.
On Wed, 6 Apr 2016, JP Hindin wrote:
- Repeated use slowly wears away the magnetic media
layer on the mylar.
There is no doubt that there is some friction there. However, I'd be
surprised if this was the chief cause.
When left in an unprotected state, or a poor
environment, damp, mold and
dust can damage the surface, either degrading the magnetic layer or
causing the gap to shrink enough that the drive head physically damages
the disk?
I have serious doubts here. While there is no question that contaminates
could ruin a disk, a drive, or both, I doubt it's the main cause. Many
floppy disks are kept in pretty pristine conditions. I know mine are, and I
still have bad sectors.
- Quantum fluctuations in the state of the universe,
caused by millions of
:-)
Of course, after ten years, it could be your own
memory that's failed.
That'd be my problem. That or being embarassed by what I find *uncorrupted*
on the floppies like college-age poetry and what-not.
Dare I ask, what's the consensus?
I'll go way off the rails and say that I beleive it's mostly due to
oxidization of the ferromagnetic particles on the floppy itself. I've also
heard smarter people than me claim that some floppies used "binders" for the
particles on the disk which are attractive to some forms of mold.
-Swift
6:07 p.m.
On 04/06/2016 10:20 AM, Swift Griggs wrote:
> When left in an unprotected state, or a poor
environment, damp, mold and
> dust can damage the surface, either degrading the magnetic layer or
> causing the gap to shrink enough that the drive head physically damages
> the disk?
Gap? There IS no gap on a standard floppy. The head
contacts the media surface.
(Bernoulli drives did have a gap, and spun at a much higher
RPM.)
But, on most single-sided floppies, there was a felt pad
that pressed the media against the head.
The only gap might be when the head load solenoid is
de-energized, the pad retracts and the media pulls away from
the surface. That was used on the old drives with AC motors
that spun the disk all the time.
Most double-sided drives pinched the media between the two
heads, and the DC motor shut off when there was no
reading/writing for a while.
I'll go way off the rails and say that I beleive it's mostly due to
oxidization of the ferromagnetic particles on the floppy itself. I've also
heard smarter people than me claim that some floppies used "binders" for the
particles on the disk which are attractive to some forms of mold.
All floppies and magnetic tapes have some kind of binder to
hold the oxide to the backing, and it does deteriorate over
time. Ozone and other air pollution probably makes it go
bad faster.
Jon
2:18 p.m.
Well, I don't know about the consensus, but in my experience, most
floppies go bad from wear and/or breakdown in the binder. It's
sometimes possible to "renew" a bad disk for use by either AC degaussing
or use of a strong DC field, but given the cost of the media, it's
scarcely worth the trouble.
After all, it's just rust spread on a cookie. ;)
--Chuck
6 p.m.
On Apr 6, 2016, at 5:18 PM, Chuck Guzis <cclist at
sydex.com> wrote:
Well, I don't know about the consensus, but in my experience, most
floppies go bad from wear and/or breakdown in the binder.
I have no experience with this issue in floppies. But I have a distressingly large
quantity of audio cassettes that have gone bad over 10 or 20 years. It wasn't wear;
they weren't played regularly. Instead, something bad happened with the structure of
the coating so that they would squeak loudly when played, both over the playback and
physically (noisy passage over the head). The problem is clearly incompetent chemical
engineering, because it showed up only in one brand, which as a result is now on my
"never again" list for any of its products.
paul
7:22 p.m.
Wear happens. Particularly on directory tracks, or where you encounter
the snake in Adventure.
But, that doesn't account for the loss of data over time.
Entropy: Could the rust on the cookie be de-oxidizing, and turning back
into non-oxidized ferrous compounds? :-)
How long were they supposed to last?
Will they honor those "lifetime guarantee"s?
10:56 p.m.
On 04/06/2016 07:22 PM, Fred Cisin wrote:
How long were they supposed to last? Will they honor
those "lifetime
guarantee"s?
Kodak, when it was entering the market with floppy media, even offered
to recover data from bad (Kodak) floppies. That offer, obviously didn't
last very long.
--Chuck
7 Apr
7 Apr
2:23 a.m.
"Try getting a replacement for one of those Lifetime vacuum tubes at Radio
Shack."
Beat me to it. I tried that twenty years ago (with a bad quartet of 7591s)
and needless to say, got nowhere. "We haven't sold those in years!" - as if
that makes any difference, in terms of the warranty?
Sure, I can appreciate their position, but still.. they _did_ garner a
premium price for those 'Lifetime' tubes with the fancy-dancy gold pins.
And IIRC, the warranty is "For the lifetime of the equipment in which the
tubes are originally installed". So it's not as if it's limited to either
the original owner or purchaser.
Sigh, the brilliance of "Huh?!??!" ;-)
On Thu, Apr 7, 2016 at 1:25 AM, Brent Hilpert <hilpert at cs.ubc.ca> wrote:
On 2016-Apr-06, at 7:22 PM, Fred Cisin wrote:
How long were they supposed to last?
Will they honor those "lifetime guarantee"s?
Try getting a replacement for one of those Lifetime vacuum tubes at Radio
Shack.
12:39 a.m.
On 4/6/2016 8:22 PM, Fred Cisin wrote:
Wear happens. Particularly on directory tracks, or
where you encounter
the snake in Adventure.
But, that doesn't account for the loss of data over time.
Entropy: Could the rust on the cookie be de-oxidizing, and turning back
into non-oxidized ferrous compounds? :-)
How long were they supposed to last?
Will they honor those "lifetime guarantee"s?
You mean like the ones on Radio Shack Vacuum Tubes,a year or so before
they all vanished from the face of the Earth.
Ben.
6:50 a.m.
On Apr 6, 2016, at 10:22 PM, Fred Cisin <cisin at
xenosoft.com> wrote:
...
Entropy: Could the rust on the cookie be de-oxidizing, and turning back into non-oxidized
ferrous compounds?
No. If that were true, iron ore would spontaneously turn into iron. Unfortunately, that
does not happen.
paul
8 Apr
8 Apr
12:13 p.m.
Entropy: Could
the rust on the cookie be de-oxidizing, and turning back int$
No. If that were
true, iron ore would spontaneously turn into iron. Unfort$
6 Apr
6 Apr
7:36 p.m.
On 04/06/2016 06:00 PM, Paul Koning wrote:
I have no experience with this issue in floppies. But
I have a
distressingly large quantity of audio cassettes that have gone bad
over 10 or 20 years. It wasn't wear; they weren't played regularly.
Instead, something bad happened with the structure of the coating so
that they would squeak loudly when played, both over the playback and
physically (noisy passage over the head). The problem is clearly
incompetent chemical engineering, because it showed up only in one
brand, which as a result is now on my "never again" list for any of
its products.
Floppies, tapes, etc. can suffer breakdown in binder, leading to the
dreaded "sticky shed" issue, for which baking is often prescribed.
I've had half-inch tapes suffer from binder "bleed", where the sticky
stuff ends up on the oxide surface and sticks to just about *anything*,
including heads, guides, etc. Running said tape through a cleaning
machine does next to nothing, other than to cause the tape to stick to
the cleaning blade.
Short term, your friend is D5/cyclomethicone applied as a thin layer and
the tape read immedately (D5 is somewhat volatile and evaporates).
I'll emphasize that my interest is strictly in *reading* and not writing
or re-use of this stuff.
Again, my money's on binder breakdown. I have lots of samples if
someone wants to investigate further, but the audio tape guys have done
a lot of work on this.
--Chuck
7:56 p.m.
On 04/06/2016 08:00 PM, Paul Koning wrote:
That is usually the pressure pad has gotten some wipe-off
On Apr 6,
2016, at 5:18 PM, Chuck Guzis <cclist at sydex.com> wrote:
Well, I don't know about the consensus, but in my experience, most
floppies go bad from wear and/or breakdown in the binder.
I have no experience
with this issue in floppies. But I have a distressingly large quantity of audio cassettes
that have gone bad over 10 or 20 years. It wasn't wear; they weren't played
regularly. Instead, something bad happened with the structure of the coating so that they
would squeak loudly when played, both over the playback and physically (noisy passage over
the head). The problem is clearly incompetent chemical engineering, because it showed up
only in one brand, which as a result is now on my "never again" list for any of
its products.
from the tape, and becomes a sticky, instead of
slippery,
material. If you can get to it, you can sometimes scrape
the sticky material off with a knife to produce a better
surface. On a (one sided) floppy drive, the pressure pad is
part of the drive, not the media.
Jon
Jon
7 Apr
7 Apr
10:06 p.m.
As a number of folks have pointed out it's not the magnetic particles - rust
is pretty stable; they read a 60 year old RAMAC at the CHM every week or so.
And it's not likely the binder, fully cured epoxies are also very stable.
Note disks do not have the tension problem that tape has.
And it's not likely the substrate, again Mylar is very stable
My guess it is both contamination and wear of the head. After all, this is
contact recording.
Tom
10:46 p.m.
On 04/07/2016 10:06 PM, Tom Gardner wrote:
And it's not likely the binder, fully cured
epoxies are also very
stable. Note disks do not have the tension problem that tape has.
There, I have to respectfully disagree. Floppy disks do not use an
epoxy binder, but usually a polyurethane-based one. (Most are
proprietary formulations). Hard disk coatings usually use an
epoxy-phenolic binder, but they're not on a flexible substrate.
Polyurethane isn't forever and tends to be somewhat hygroscopic.
--Chuck
8 Apr
8 Apr
12:11 p.m.
Sorry, to a EE they are all epoxies, but:
"BINDER: A polymer such as polyester-polyurethane used to bind magnetic
particles together and adhere them to the base film in the manufacturing of
magnetic media."
http://www.fujifilmusa.com/shared/bin/Magnetic_Media_Terminology.pdf
"Polyester polyurethane is the workhorse of polyurethane films. It offers
excellent strength and long-term stability"
http://www.apiusa.com/polyesterpolyure.html
I think properly cured all binders used are pretty stable.
Tom
-----Original Message-----
From: Chuck Guzis [mailto:cclist at sydex.com]
Sent: Thursday, April 07, 2016 10:47 PM
To: General Discussion: On-Topic and Off-Topic Posts
Subject: Re: Why do good floppy disks go bad?
On 04/07/2016 10:06 PM, Tom Gardner wrote:
And it's not likely the binder, fully cured
epoxies are also very
stable. Note disks do not have the tension problem that tape has.
There, I have to respectfully disagree. Floppy disks do not use an epoxy
binder, but usually a polyurethane-based one. (Most are proprietary
formulations). Hard disk coatings usually use an epoxy-phenolic binder, but
they're not on a flexible substrate.
Polyurethane isn't forever and tends to be somewhat hygroscopic.
--Chuck
12:12 p.m.
On Apr 8, 2016, at 3:11 PM, Tom Gardner <t.gardner
at computer.org> wrote:
Sorry, to a EE they are all epoxies, but:
"BINDER: A polymer such as polyester-polyurethane used to bind magnetic
particles together and adhere them to the base film in the manufacturing of
magnetic media."
http://www.fujifilmusa.com/shared/bin/Magnetic_Media_Terminology.pdf
"Polyester polyurethane is the workhorse of polyurethane films. It offers
excellent strength and long-term stability"
http://www.apiusa.com/polyesterpolyure.html
I think properly cured all binders used are pretty stable.
That may be, but if so, there are some big name vendors who do not know
"properly".
paul
12:43 p.m.
On 04/08/2016 12:11 PM, Tom Gardner wrote:
Sorry, to a EE they are all epoxies, but:
Gorilla glue is an epoxy? News to me.
Polyurethane is far from stable over the long term and exhibits
hygroscopic properties:
For example, vide:
http://www.ptonline.com/knowledgecenter/Plastics-Drying/Resin-Types/Hygrosc…
...and many others.
Floppy and tape binder also is somewhat of a witches' brew as it usually
includes a fatty acid or some other substance for lubrication.
As far as "stability", that's subjective. Most plastics will eventually
degrade; some are worse than others. There's a big problem in the
museum archiving business that not much, other than freezing seems to
retard eventual decomposition of plastics.
--Chuck
7 Apr
7 Apr
10:49 p.m.
This from Fuji may also help a bit when examining the makeup of a floppy
disk:
http://www.fujifilmusa.com/shared/bin/Magnetic_Media_Terminology.pdf
--Chuck
6 Apr
6 Apr
5:55 p.m.
On Apr 6, 2016, at 3:15 PM, JP Hindin <jplist2008
at kiwigeek.com> wrote:
I had one of those Japanese Koan moments recently when someone asked me "Why do
floppy disks stop working?" and I realised I... didn't actually know. I thought
I'd throw it to the group and get some theories/proofs.
Let's work on the assumption we're talking about 5.25" and 3.5" disks.
Several guesses:
- Repeated use slowly wears away the magnetic media layer on the mylar.
- When left in an unprotected state, or a poor environment, damp, mold and dust can
damage the surface, either degrading the magnetic layer or causing the gap to shrink
enough that the drive head physically damages the disk?
- Quantum fluctuations in the state of the universe, caused by millions of mostly
non-interacting particles passing through a disk in any given minute, alter the magnetic
spin of the ferric atoms causing gradual data loss over time (mostly tongue-in-cheek)
- Given the lack of use of most floppy drives they themselves pick up 'gunk' and
on first reading a diskette after a long time of disuse damage it.
It _seems_ like when you put a 3.5" disk down for ten years and pick it back up, a
disk that used to work fine no longer does. Of course, after ten years, it could be your
own memory that's failed.
Dare I ask, what's the consensus?
- JP
On a related note... What causes an old mfm / roll drive to die if the heads aren't
touching the surface except in the landing zone?
I have read that it's not impossible for the "pigment" to lose its
coercivity over time.
6:07 p.m.
On Apr 6, 2016, at 8:55 PM, Geoff Oltmans <oltmansg
at gmail.com> wrote:
On a related note... What causes an old mfm / roll drive to die if the heads aren't
touching the surface except in the landing zone?
I have read that it's not impossible for the "pigment" to lose its
coercivity over time.
I would like to hear an explanation of the physics behind that claim. Off hand, I
don't believe it.
A likely cause of drive failure is contamination. That could come from inadequate seals
(or seals that degrade over time). It could also be from contaminants released by
components of the drive itself, if their chemistry is not quite right. It doesn't
take much to cause a head crash.
paul
3180
days inactive
3182
days old
21 comments
12 participants
participants (12)
-
bfranchuk＠jetnet.ab.ca -
cclist＠sydex.com -
cisin＠xenosoft.com -
drlegendre＠gmail.com -
elson＠pico-systems.com -
hilpert＠cs.ubc.ca -
jplist2008＠kiwigeek.com -
mouse＠Rodents-Montreal.ORG -
oltmansg＠gmail.com -
paulkoning＠comcast.net -
swiftgriggs＠gmail.com -
t.gardner＠computer.org