20

The User's Guide for the Commodore 1541 disk drive includes a vague warning, on page 7, to "remember to always remove the diskette before the drive is turned off or on". Some third-party books and magazines issue the same warning even more forcefully—for example, Jim Strasma, writing in the June 1984 issue of RUN, goes so far as to claim that "if you turn the drive on or off with a disk in place and the door latched, you will probably damage the data on the disk".

Now, I had a Commodore 1541 and I must have turned the drive on or off with a disk inside thousands of times. This never resulted in any obvious damage to the physical disk or to its digital contents. Was I just lucky, or is there really some mechanism by which powering up or down the drive might damage a disk left inside? If so, what exactly is this mechanism, and under what sort of conditions might it be triggered?

Note that I'm not talking about turning off the drive while it has a file open for writing; obviously that can result in things like splat files and corrupted block allocation maps. I'm asking whether it can be dangerous to shut off the drive when no disk activity is ongoing.

2
  • 2
    Just anecdotal evidence, I powered my drives on/off with disks in them more times than I can possibly account for. I never had any obvious side effects. I guess it's possible I blew up an unused sector here or there, but I don't think I ever had any data loss. - Now turning off or ejecting the disk in the middle of a Write operation is a whole other ballgame... :-P
    – Geo...
    Dec 16 '21 at 15:59
  • 3
    @Geo... - I thought ejected a disk during a write was the best way to preserve your Castle Wolfenstein and Ultima saved games. Dec 16 '21 at 17:44
16

According to the 1541 service manual, the device contains power-up and power-down write protection. On power-up, the CPU controls the power to read/write electronics so they are not turned on before CPU is properly running. On power-down, the power to read/write electronics turns off before CPU loses power.

That said, of course it is possible that there are still mechanisms that may damage data, in which case it is best to not to leave disk inside if you want to be sure.

3
  • 1
    Thanks for the information about power-up and power-down write protection. If these aren't sufficient to prevent damage to a disk, then I'm very interested to learn the details of what other "mechanisms that may damage data" could exist.
    – Psychonaut
    Dec 16 '21 at 6:47
  • 1
    @Psychonaut: I would guess that the protection mechanisms are designed to work correctly in cases where the supply voltage ramps straight from 0 to 5 volts, or from 5 volts to zero, but might not always work if the supply voltage wobbles up and down. If one had a 1541 plugged into a power strip with an induction-motor-based device and a fluorescent light, it's possible that turning off the power strip could cause the voltage to fluctuate based if the motor initially acts as a generator but the fluorescent light draws down the voltage, but then the voltage rebounds when the...
    – supercat
    Dec 16 '21 at 22:22
  • 1
    ...current through the light becomes insufficient to maintain ionization and thus drops to almost nothing. Advising against having a disk in the drive when cutting power is easier than trying to anticipate all the situations where supply voltage might fluctuate up and down.
    – supercat
    Dec 16 '21 at 22:23
12

I don't know about the 1541 in particular, but I have personally lost data by powering off a computer with a diskette in the drive. One problem that could occur, independently of whether you remembered to save your work or, whether the OS finished writing out buffered data, was that somewhere in between the power supply and the read/write head, the power-off event could create a "glitch" that would ruin the formatting of the track on which the head was resting.

3
  • 3
    Given @Justme's point that the 1541 has power-up and power-down write protection, I'm not sure that what you describe is a concern for this drive. But even if it is, I'd like further details on how such a "glitch" might arise.
    – Psychonaut
    Dec 16 '21 at 6:45
  • @Psychonaut I think it's more that it's hard to make guarantees about voltage levels and things like that when the rug's been pulled. This is kind of like the Undefined Behaviour in C.
    – OmarL
    Dec 16 '21 at 8:54
  • 3
    FYI, some modern rotating hard drives use the momentum of the rotating platters to generate enough electric power to retract the heads to a safe "parking" location when they detect a "brownout" at the power inlet. So, by the time a glitch might happen, the heads are in a location where it won't affect any data. Dec 16 '21 at 15:49
10

In the thousands of times I turned my 1541 off and on I always ejected (at least partially) the disk in the drive, because I also heard (probably read somewhere) that leaving it in place could damage the disk. The reasoning I (rarely) heard for this precautionary measure was that when you turn the drive on, nobody can guarantee that no accidental electric pulse (from turning on the whole drive/electronics) could reach the r/w head and affect the magnetic surface on the disk.

Pulling out the disk just halfway always seemed enough for me, as opening the latch also moves the r/w head vertically away from the disk, and - as we know - it has to be very close to read/write data.

Good question and very interesting that you didn't experience any data loss. I couldn't tell because I got into the habit to pull the disk out always. I even had the one or other corrupted file/disk nevertheless, but that's a different story.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.