Contents copyright Herb Johnson 2014. Last update March 7 2014. Quoted material is copyright by the respective authors of that material and used with permission. For more info or for reuse or questions, email me via this Web link. Corrections are appreciated.
Old floppy diskettes often accumulate dust and mold (mould) over years and decades. The oxide coatings can become brittle. Crud on a floppy drive read head can scrap up the oxide and ruin not only the bad diskette, but subsequent disks. Here's some notes about diskette surface problems and how to clean them, and here's further notes about diskette quality by brand.
For other issues and technical information about floppy drives and diskettes, review my Web document at this link.
- Herb Johnson.
A common problem with floppy drives is a consequence of old diskette media becoming brittle and scraping away. The iron oxide coating actually strips away, as its binders become brittle. Other media problems include mold, growing on the plastic surface. A sharp ear can hear the difference in sound, when the drive head starts to drag or even scrape the diskette. It can cause damage not just to the "bad" diskette with fragile coatings, but to subsequent diskettes read on that drive.
Here's some photos of diskette media with mold damage. It's hard to see unless you use bright reflective light and possibly a magnifier. In these photos, I've removed the disk media from the envelope. In photo #1, the mold is widespread and looks like spilled and dried dirt. Mold can grow in that fashion. Or, in photo #3 it can look like filaments or an irregular web. When a moldy disk is read, the drive heads can smear the mold along tracks; in photo #2 you can see the circular traces where the diskette surface is marked, roughed up, or abraded.
Mold may also appear as small black dots. In general, any irregular pattern or off-color material on media may be mold. however, some brands of diskette coatings have regular pattern in the iron oxide coating. Look at several disks of the same brand and kind to make a determination.
When the head accumulates even a little iron oxide, it acts as a scraper, and picks up more material. A dirty or damaged read/write head, produces these scratches on media. Crud on the floppy disk drive head requires cleaning with Q-tip in alcohol. you can see in this photo, a Q-tip with oxide from a floppy head.
In response to the following posted question in comp.os.cpm, Lee Hart wrote this in Aug 22nd 2002 (quoted with permission):
[someone] wrote: > > I've had some 5.25" cp/m diskettes in storage for the past 17 years. > It appears that many of the disks are no longer readable or they are > readable with some file errors. > > Have any of you had problems like this? Is there a way to correct the > problem?
Lee Hart responded: I've been having the same problem. It appears that certain brands turned out to hold up much better in storage. There is also a significant correlation between amount of data on the disk and how easy it is to read. Single-density 40-track are the easiest; double-density 80-track the worst, etc.
If a disk has not been stored in a closed sealed package, then it is likely to have collected dust, or worse yet, mildew from storage in a damp environment. Attempting to read such a disk is likely to leave crud on the disk drive's read/write heads. Not only is this disk unreadable, but any subsequent disk you put in that disk drive may be contaminated, ore even destroyed!
Cheap disks can also have the oxide coating flake off. Some old discount brand disks I've found can only be read a few times before they become unreadable, and the heads get plugged with oxide dust. Head cleaning time, again!
Here's what I've started doing to test the disks and "refresh" the data on them:
1. Read the disk with a utility like FINDBAD that reads every single sector and marks bad ones.
2. Copy the intact files to a newly-formatted known good disk.
3. Erase, and reformat the old disk. 4. Run FINDBAD on it again, and insure that there are no bad sectors. If 100% good, then copy the original files back onto it.
I also try to use the lowest density format that is sufficient for the files that need to be stored."
---end quote-- Lee Hart
A subsequent poster (Aug 2002) suggested that a "moldy" diskette can be recovered as follows. (Note: a 5.25 inch or 8-inch diskette consists of a doughnut of plastic with the magnetic coating, in a sleeve or jacket.) Remove the diskette "doughnut" by cutting the edge of the jacket and removing the doughnut. Wash the doughnut in warm soapy water, CAREFULLY, and let it dry. Insert the cleaned doughnut in a SUBSTITUTE jacket that is known clean.
Another respondant suggested that the "mold" may in fact be glue from sealing the original jacket. I myself (Herb Johnson) have had diskettes with mold on them from dampness; I've also seen residues from glues on lables cause problems. Many diskettes were sealed using ultrasonic "welds" on the plastic, so that's not "glue".
Sources: as above, quoted with permission
In Aug 2009 I (Herb Johnson) responded to comments about the idea of "soap and water" and problems with "noisy drives" as follows:
"Soap and water" is not crazy, water with a very small bit of soap is a very benign cleaner. A few minutes or a slight NOT HOT breeze of air will dry the surface. DO NOT USE A HAIR DRYER - this is thin plastic, it will MELT. With care you can clean a diskette's media without removing it from the envelope, just work on a section at a time and rotate the "doughnut" in the envelope by hand.
Diskette media surfaces are essentially a coating of iron oxide in a plastic binder with embedded lubrication, on mylar or similar material. Unfortunately, this is also a great surface for mold. Also, the binder material can degrade over decades and become brittle, the lubricants can dry out. Read my Web page at the link above for more information.
The description of NOISE while reading diskettes means either 0) the disk media is rubbing in the envelope - a distinctive noise; 1) the rotating hub which grabs the center of the diskette is dry and noisy; 2) the motor which drives the hub is dry and noisy or - worse - 3) the read/write head is digging up iron oxide and scraping it off the disk and chattering in the process.
LIFT THE DRIVE DOOR IMMEDIATELY if these noises occur, inspect the diskette, and see if there is physical damage to the diskette's media. Inspect the drive - use a 'scrap" diskette to determine a source for the noise. Clean the heads with a long QTip and alcohol.
Inspection of the diskette may show some circular scuffing, which suggests the heads are dirty and/or the media is fragile. Good media have lubricants to reduce scuffing. Scuffs won't lose data (much). Close inspection will show if the "tracks" are just scuffs, or gouges of removed material (fatal damage to the data and diskette). CLEAN THE HEADS, clean the heads, clean the heads - crap on the heads will damage your diskettes, you will kill five disks to "test" the drive after one of these "dirty head" events.
I can't stress enough, the value of knowing the sounds and sights of your system when working "normally". Odd noises, odd smells, odd sights - these are all diagnostic of problems which, if resolved promptly, will extend the life of your ancient systems and data.
From an independent discussion in Usenet newsgroup comp.sys.tandy in Sept 2009, Mike Yetsko made the following comments. They are quoted here with permission. - Herb
There was ..a problem in that some of the early [8-inch floppy] drives, like the Shugart , had a 'DC balanced' amplifier on the front end for the read head, and as such, had to be adjusted so that pulses were symetrical for flux changes. Later drives had an AC coupled amplifier and no adjustment was necessary.
One thing to warn you about though is the 'head load pad' [on single-sided drives]. This is the little fiber pad that pressed on the diskette and caused the media to deform properly across the head. If they picked up particulate matter (and they did a LOT!) they could score a diskette. What makes them really bad is that if a diskette is scored and there is a 'hard spot' in the pad the diskette will deform slightly the WRONG way over the head and not be read. Or worse, it can physically damage a diskette. A hard spot on the pad can cause material to build up on the head in a coresponding spot. While that wasn't as bad in the 5.25" drives as it was in the 8" Model II's (I've seen Model II diskettes that you could SEE through in a ring from the drive spinning all day long on one track!) it could still cause problems.
I always cleaned my heads with a Q-Tip and denatured alcohol. Don't use rubbing alcohol as it has too much water in it. And make sure the head is clean and dry before you let the arm back. Do NOT let the load arm 'snap down' against the head after you've cleaned it. Finally, I always used an X-Acto knife on the pad. No, not to 'cut' anything, but to 'pet' it like you would do to a cat with your hand, keeping the blade at a small angle to the surface and taking it BACKWARDs across the fiber, not so that it could cut anything, but to dislodge any particulate matter and to 'fluff up' the pad a bit. Then I'd put in a diskette rotated so that the window was NOT over the head, lower the arm, and PRESS it gently to 'pack' the fiber slightly.
Replacement pads used to be available (I think about 15 cents) and were easy to snap out and replace, but I think anything like that are long gone as replacement parts. Oh, I should say, make sure you even HAVE a pad! I've seen drives where the pad has fallen out! If that happens, go to a hobby shop and get some of those felt pad to glue onto the bottom of metal box. Then go to a stationary store and get a 'punch' (looks like pliers) a bit bigger then 1/8". Punch out a pad, and stick it in there. If you can't get a punch, go to WalMart and buy a 'leather punch'. Looks like a big pair of pliers with a 'star' wheel on the end with all different sizes of holes.
- Mike Yetsko
IN 2013 I had to recover data from 5.25" diskettes with severe mold damage. Some photos of that media are above. To read the diskettes, here's what I did.
I cut the entire edge of the envelope, removed the entire "cookie" or circular disk media from the envelope. Then I washed the surfaces with 90% isopropyl alcohol on a clean cotton ball (cotton with no oils or other chemicals). The disk was allowed to dry completely and I inserted it into a different clean envelope. The cleaned disk was read - but if any read errors occurred or the drive sounded like it was scraping, I immediately removed the diskette and cleaned the drive heads.
This very aggressive approach allowed me to recover most of the files and minimized further damage to the diskettes and the drive.
Over the years, many people have reported that some diskette manufacturers - "brands" - produced better diskettes than others. Better means that over time, the oxide surface of the media has not become brittle, it does not dry out. For 8-inch diskettes this is an issue because some of them are 20, 25 or more years old as of the 21st century. Mold is always an issue for any plastic surface, including the "cookie", the oxide-coated disk inside the floppy envelope. And there can be other issues of manufacturing which affect long life.
In general, old "name brand" diskette brands have faired better than secondary names; and disks made in the 21st century are often not as good as those made decades earlier.
My colleague "Greegor" posted in comp.os.cpm in April 2009 the following; "How about a mention of how cookie substrate actually has a ""grain"" to it and humidity and temperature can deform the substrate "cookie" making it slightly oval shaped? I remember that while there were (at one point) dozens of diskette brand names there were only like 3 makers of the oxide coated cookies. Dysan's standout quality reputation was based on their reprocessing (burnishing/polishing) and QC [quality control] before they put the cookie into the diskette carrier."
Mike Yetsko apparently read the above and told me: "Oh, I was going to tell you to warn people about 'flippy' media. [This was the practice of adding a second index hole to single-sided media and "flipping" it over to use the other side on drives which had only one read/write head. - Herb] Yeah, I had a punch and did it back then, but quickly discovered some brands of media would wear out quickly if you flipped them. It SEEMED that some brands of media would get a 'grain', almost like fur, and when you flipped it you would be reversing the 'grain', which seemed to cause accelerated wear. Media would start to flake fairly quickly. Some media. Other media didn't seem to care. If you were flipping for archive storage, that was fine, but not for everyday use."
Mike added: "CDC [brand 8-inch] media was s**t no matter what you did. Tandy bought a ton of the stuff for the Model II, since they put CDC drives in the bay. (I didn't like CDC drives any better than I liked their media.) But the stuff came apart if it was not stored in cool temps. Tandy had it in uncooled warehouses in Texas and it was a mess. The heat destroyed the binder and the media just flaked almost from first insert."
A 2011 conversation with Stewart Kay (microbee emulator and diskette utility author, see my S-100 Web pointers of 2011 for details) included the following: "I've also had problems with the disk heads removing the disk [coatings], and as we have both found this emits a sound from the drive. It's mostly been poor quality media where I have found this to be a problem. One hint I give in my [disk utility's] README file is that flaky disks can be tested for by scraping the disk surface from inside of the [center] mounting hole, which never comes in contact with the heads. All old disks will flake somewhat but the bad ones are quite obvious."
Steward continues: "I'm in the process of recovering data from some 8" disk. I've found that most read fine (Verbatim brand) but I have several disks where the head is removing the surface of the disk (Computer Resources Company and Wabash brands) and some of these disks I'm eager to recover. Has anyone played around with counter-weights on the drive heads?"
I advised not to do adjust the pressure on the heads, as the heads are designed to read with a certain amount of pressure on them, to make good contact with the coating which originally included lubricants. Less pressure will reduce signal and introduce errors. But the problem of scraping off coatings does depend on the head design: heads that are small and fragile and square will twist and SCRAPE the media. Heads that are large, rounded and massive can still scrape off media, by accumulating debris on the head which, again, acts as a scraper. I have no solutions at this point for "flaky media" beyond these considerations of lubrication and head design.
More comments about diskette media quality is in this document of a discussion with Chuck Guzis, formerly of Sydex.
Speaking for myself, 3M, Verbatum, Dysan brand 8-inch and 5.25-inch diskettes have in general been very good over the years.
Copyright © 2014 Herb Johnson