12

In the IBM PC 5150 documentation, it is stated that the floppy drive has 40 tracks, 8 sectors/track and 512 bytes/sectors, for a total of 163 840 bytes of data. It makes sense, 40×8×512 = 163 840 bytes. Ok. It is also said that a floppy drive has 45 cylinders...

The IBM 5-1/4″ Diskette Drive is a single sided, double density, 40 track unit. The Diskette Drive has a formatted capacity of 163,840

(page 2-110 from the IBM 5150 Technical Reference, 1st edition)

1 Head, 45 cylinders, 8 sectors/TRK, 512 bytes/sector,MFM.

(page 2-104 from the IBM 5150 Technical Reference, 1st edition)

Both manuals are available from https://www.pcjs.org/documents/manuals/ibm/.

How is it possible? How can the number of cylinders be different from the number of tracks? What happens if I want to access cylinder 42 when there are only 40 tracks?

I thought cylinders were tracks on top of the others, there should be the same number of tracks as cylinders, shouldn't there? Or are there 45×8×512 = 184 320 bytes, but 163 840 usable and some tracks are disabled? I suppose this is not the case because the address mode would have been THS for Track/Head/Sector instead of CHS.

2
  • most of the FDDs I came to contact with had 42 tracks possible some 43 but no more ... so my bet is 45 is the virtual "mechanical limit" of the media ...
    – Spektre
    Commented Jul 29, 2020 at 21:26
  • Pu_1700.exe can do (on ibm dd floppy) up to 41 track/10 sector; on typical late 80s dd floppy drive up to 81/10, some cases 83/11. Commented Feb 16 at 5:32

3 Answers 3

16

As far as I can tell, the first edition of the IBM 5150 Technical Reference is the only IBM manual which mentions 45 cylinders. Few Technical References document the FDC, which is where the “45 cylinder” mention appears, but those that do, other than the 5150 first edition, don’t reproduce this — see for example the first edition of the IBM 5160 Technical Reference, page 1-167.

This suggests that it could have been a mistake (and indeed, it’s impossible to use 45 tracks on a 5.25” drive).

I initially thought that another explanation could be that the section mentioning 45 cylinders is documenting the FDC’s limits, but the FDC isn’t limited to 45 cylinders, and the text seems to me to be specifying a disk format.

6
  • 1
    Original engineering goal = 45, actual achievable result = 40?
    – dave
    Commented Jul 29, 2020 at 12:07
  • 2
    Maybe not 45 tracks, but it was possible to use at least 42 tracks on a IBM PC floppy disk.
    – Grabul
    Commented Jul 29, 2020 at 12:51
  • 1
    @TEMLIB yes, depending on the drive, 42 or 43 tracks; I’m not sure what the original IBM drive was capable of (I no longer have a working 5150 drive). Commented Jul 29, 2020 at 12:54
  • 1
    @another-dave: Another factor is that the if there's any uncertainty about the location of the head within the carriage assembly, the range of motion for the carriage assembly within the drive would need to be larger than the range of motion that would be used for any particular carriage assembly. If e.g. the head was unusually far toward the end near the center of the disk, reading the outermost track would require moving the carriage further out than would be needed if the drive head were centered.
    – supercat
    Commented Jul 29, 2020 at 16:52
  • What about later 360KB drives (which is double sided 180KB)?
    – RonJohn
    Commented Feb 16 at 6:58
9

Floppy disk drives could usually mechanically access more than 40 tracks, but only 40 were actually used, "formatted".

First track is on the outer edge of the disk, there is a detector than signals when the disk head has reached the edge. Other tracks are reached by moving by a known number of phases a stepping motor.

Simple drives use "CAV" = Constant Angular Velocity : The disk turns always at the same speed. Combined with the fact that all tracks have the same number of sectors, data on outer tracks is occupying more area, which makes these tracks more reliable. Even if a drive could phycically access up to 45 tracks, inner tracks can be too short to be reliably used.

A bit more recent drives such as Macintosh floppies or CDs use variable motor speed ("Constant Linear Velocity") and different number of sectors per track to compensate from differing track lenghts (and in CDs data starts from the innermost part of the disk to support different diameters).

4
  • 1
    Bear in mind that Apple switched to CAV for 1.44MB floppies. Given that their drives still needed to be custom for backwards compatibility with 400K and 800K floppies and to support software eject, I'm assuming that was just for easier inter-compatibility with PCs.
    – ssokolow
    Commented Jul 29, 2020 at 23:07
  • 1
    @ssokolow: Another factor may be that the time needed for the drive speed to stabilize when moving between inner and outer tracks was greater than the time needed to move the head.
    – supercat
    Commented Jan 6, 2022 at 15:41
  • All 5 1/4" (and 8") floppy drives use CAV -- the variable motor speed was not introduced until 3.5" floppy drives.
    – Chris Dodd
    Commented Feb 14 at 21:42
  • @ChrisDodd Early Apple Lisa and some Macintosh prototypes used special 5 1/4' "Twiggy" floppies with variable RPM drives (I've just searched). Anyway it's just an obscure format quickly abandoned.
    – Grabul
    Commented Feb 14 at 22:20
2

Cylinders are a hardware property, while tracks are logical. These are not necessarily the same.

A floppy disk (drive) may feature a given number of cylinders, while the disk operating system may only use (and assign) a subset of these as tracks. The disk operating system may even assign tracks in some sort of abstraction of the underlying hardware. E.g., some CP/M systems appended the tracks of the second side of double-sided floppies on top of those of the first one. Meaning, while the floppy physically features 45 sectors on each side, it may appear like a single-sided disk with 80 tracks.

You must log in to answer this question.

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