From Wikipedia - Floppy_disk:

to format or write to this high-coercivity media, the high-density drive switched its heads into a mode using a stronger magnetic field. When these stronger fields were written onto a double-density disk (having lower coercivity media), the strongly magnetized oxide particles would begin to affect the magnetic charge of adjacent particles. The net effect is that the disk would begin to erase itself.

This is a similar question to Was there a special drive, that saved much more than 1.44MB on an HD floppy disk?, but more dedicated:

How long can a MF-2HD disk hold 2.88MB of data, before it begins to erase itself?

And can any firmware hack allow enforcing 2.88MB onto 1.44MB?

The 32MB squishing technology from SuperDrives is different and requires hardware dedication.

  • 1
    (1) Are you somehow a floppy-disk-fetishist? (2) Can't find that quote under the link you provided
    – tofro
    Jan 17, 2018 at 12:51
  • 2
    @tofro (1) I'm pretty interested in floppy disks too. I'm not sure why though. (2) That quote was from retrocomputing.stackexchange.com/posts/5435/revisions, not from wikipedia Jan 17, 2018 at 14:07
  • @LateralTerminal Can't find that quote there as well. It is kind of unlucky and probably means to say that magnetizised areas are bleeding out - inastantly, not over time. Pure magnetic effects are one of the very few physical phenomenons that don't have something like "speed" attached to them.
    – tofro
    Jan 17, 2018 at 14:17
  • 1
    @tofro I see. He didn't mark where he got the quote from. Anyway I did manage to find the original original source. It's near the bottom of this section of History of the floppy disk wikipedia. (2) The_5¼-inch_minifloppy Jan 17, 2018 at 17:01
  • 2
    @TechLord In future, I suggest that you ask in chat before asking questions until you're more used to the site. Some of your questions are gems!... but others aren't.
    – wizzwizz4
    Jan 17, 2018 at 17:13

2 Answers 2


Experimentally verified data retention duration: 0 seconds.

Because you asked and I was curious too, I have just cut an ED notch in one of my old HD disks (to trick the drive into thinking it were an ED disk) and tried to format it in my ED disk drive set on my Sinclair QL - It wouldn't even format.

The drives have no problem whatsoever in handling both ED disks and HD disks (coded and treated as HD).

The disk could also be properly formatted as HD once i closed the ED notch.

Obviously, we shouldn't be generalizing from one single test, but for me it didn't work at all.

Possible reason could be that HD uses longitudinal recording (the "magnets" lie flat with N and S in the same layer) and ED perpendicular recording (the "magnets stand upright") - This doesn't seem to work at all with HD disk coating. But could be a gazillion of other reasons.

(You owe me a disk)

  • 1
    I'm not too surprised. The original quote is talking about formatting DD disks as HD so while its semi-logical that using HD as ED disks would work similarly, there's many possible reasons why that wouldn't necessarily be the case in practice.
    – mnem
    Jan 17, 2018 at 17:25
  • 1
    @mnem I'm not surprised as well, but had some fun trying.
    – tofro
    Jan 17, 2018 at 17:26
  • No doubt. I'd try and add another data point, but I don't have any ED disks on hand to confirm that my ED drive works properly normally, so it wouldn't really prove much by stuffing a modified HD disk in it.
    – mnem
    Jan 17, 2018 at 17:36
  • 1
    So... HD as ED is almost exactly as reliable as the machine's built-in Microdrives?
    – Tommy
    Jan 17, 2018 at 18:41
  • 3
    @Tommy The QL's microdrives are actually widely underrated - I just stuck a 33-year-old cartridge that was with the original computer (never written to since) in there and it worked much better than the HD/ED disk: Took 7 seconds to load a full Quill (The QL's Wordprocessor) without any error. The late Samsung-made devices work really well. Didn't try to format it ED, however.
    – tofro
    Jan 17, 2018 at 19:19

I'm not sure how ED drives work in particular, but a general issue with different densities of disk media is coercivity. The basic design goal of disk media is that the magnetic domains within it should flip their states in the presence of a magnetic field that exceeds a certain strength, and retain their states absent such a field. The design goal of the drive head is to produces a field which is strong enough to affect the parts of the disk that should be written, but not so strong as to disturb the parts that shouldn't be.

When using media with a low coercivity, a portion of the media which has just been polarized in one direction is moving away from the head and is escaping its influence, and the head starts trying to polarize the media in the other direction, the field produced by the portion just written will oppose the field from the head. This effect can limit usable data density. When using media with a high coercivity, and writing it with a stronger signal, this effect will be less pronounced, thus allowing a higher density. Consequently, media that need to support higher densities use higher corcivity materials.

If a drive is used with higher-coercivity media than it is expecting, it will drive the heads too weakly to reliably flip the states of the magnetic domains on that media. If it is used with lower-coercivity media, it will power the heads too strongly, thus affecting a wider area of the media than it should.

  • Very interesting.
    – neverMind9
    Jan 19, 2018 at 15:56
  • So ED has the highest coercity.
    – neverMind9
    Jan 19, 2018 at 15:57

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .