Matth79 wrote on 2022-04-13, 23:13:

A permanent magnet strong enough to wipe the disk, would leave it harder for the drive to write as the disk would be fully magnetised

A working floppy drive should generally create a field strong enough to correctly write to the medium, as long as it is a supported medium. DD drives may fail to handle HD media, and even more so when the disk is fully magnetized. A floppy drive write head is built in a way that it writes data to the center of the track and erases data (using AC magnetization) at both edges of the track, to prevent old data disturbing the reading of the new data, in case the new track doesn't perfectly align with the old track.

This means that low-level formatting a floppy disk just once in a properly working drive is enough to remove any disturbing old magnetization that could cause problems when you read the floppy afterwards in the same drive (or same kind of drive, as long as both drives are aligned in the same way). The only moment the method of writing the center and erasing the edges fails is when the writing drive and the reading drive do not agree on track width, such that the reading drive reads tracks wide enough that it reads outside of the area erased by the writing drive. This can happen when you write on an 80-track drive and plan to read with a 40-track drive. In case of PC-compatible drives, this only applies to writing using a 1.2MB drive and reading using a 360KB drive. Bulk erasing (preferrably using proper AC degaussing) might actually help in this isolated case. In all other cases, degaussing a floppy won't improve results over just trying to reformat it without degaussing it in-between.

The pure act of reformatting a floppy can in fact "cure" bad sectors, if the problem is caused by the surface being dirty. The dirt might be wiped by the tissue inside the medium, maybe only after being gently "scratched" by the head. Furthermore, as the heads push from both sides onto the medium, just repeatedly formatting a disk might straighten out dents in the medium. In both cases, the surface improves due to the mechanical treatment during formatting, not due to the magentization applied during formatting, so degaussing can't assist that process.

mkarcher wrote on 2022-04-18, 19:48:

degaussing a floppy won't improve results over just trying to reformat it without degaussing it in-between

I have a bunch of second hand floppies, which I had to clean with water and IPA first (I mean the magnetic surface, though the envelopes were dirty too) to remove weird stains, dirt, and possibly mold before I even put them in the drive. A lot would not format without bad sectors even using /U and /C, and I did try at least 3 times each. All the "bad" floppies then got the magnet treatment and most formatted fine on the very next try after. And that's with me being lazy and not even properly pulling the magnet away to reduce the residual field - so I'm going to stick to my methods.

And in case anyone wonders why I even resucue such floppies, I've fixed a bunch of floppy drives and I need test disks to make sure they work before I risk using my good floopies. And sometimes I find such nice gems on the old, used floppies like DOS 1.10 bootable one with Bushido game on it. It's a single-side 160k format, and it shows, the floppy in question has the upper surface in pretty rough shape due to these drives not having the upper head but a pressure arm instead - which often got dirty and scratched the media.

Deunan wrote on 2022-04-20, 10:29:

mkarcher wrote on 2022-04-18, 19:48:

degaussing a floppy won't improve results over just trying to reformat it without degaussing it in-between

I have a bunch of second hand floppies, which I had to clean with water and IPA first (I mean the magnetic surface, though the envelopes were dirty too) to remove weird stains, dirt, and possibly mold before I even put them in the drive. A lot would not format without bad sectors even using /U and /C, and I did try at least 3 times each. All the "bad" floppies then got the magnet treatment and most formatted fine on the very next try after. And that's with me being lazy and not even properly pulling the magnet away to reduce the residual field - so I'm going to stick to my methods.

And in case anyone wonders why I even resucue such floppies, I've fixed a bunch of floppy drives and I need test disks to make sure they work before I risk using my good floopies. And sometimes I find such nice gems on the old, used floppies like DOS 1.10 bootable one with Bushido game on it. It's a single-side 160k format, and it shows, the floppy in question has the upper surface in pretty rough shape due to these drives not having the upper head but a pressure arm instead - which often got dirty and scratched the media.

And how long did these "repaired" disks last? Have you tried doing a full surface scan on them a week or a month later to see if they were still free of bad sectors?

I have not, but the magnetic degradation of a floppy stored in a box is very slow. I don't expect any of them to fail after just a month or two. I suppose one can get a "false positive" when a damaged floppy seems to format fine after some treatment but then fails read/write soon after. Haven't happened to me yet but if/when it does the floppy will be downgraded to 1-sided or a head separator for transporting drives. I'd need to see way higher failure rates to consider my magnet to be ineffective.

I have had some floppies that fail only hours after the magnet trick, which is still enough for the quick "I must install this software" kind of things I tend to use such floppies for. Some have remained working for longterm (months after).