Floppy Drive Tuning

[jmacz]

Normally I haven't bothered too much with the floppy drives (aside from getting them kinda working and ejecting), but now that I have two disk centric models (512K and Plus), I have been getting more and more annoyed with the lack of consistency across the floppy drives in my 11 vintage Macintoshes. A few of them seem to have zero problems sharing disks with each other, but the rest are sketchy.. partially work, totally don't work, etc. I believe it's tied to the zero track alignment where depending on which direction it's off, you'll get some of the machines unable to share disks with others.

While I wait for a replacement screw for my external 400K drive, I decided to take a look at some of my other floppy drives.

What technique/strategy have you used to ensure your floppy drives are working to near factory spec?


For now, I've been doing the following using Disk Copy 4.2:

• Read a known working commercial floppy disk and note the tag checksum and data checksum.
• Make a duplicate, lock it, and then read it to get its tag checksum and data checksum.
• Compare the result and ensure it's spot on.
• Use both the original commercial floppy and the duplicate on the next floppy drive, check both sets of checksums on both floppies. Then start with the duplication step again.
• Do the above for 400K, 800K, and 1.4MB floppies (depending on what the floppy drive supports).

I've only got 3 drives done so far and had to tweak the zero track alignment on one of them (read errors). All three can now share disks (haven't had an error yet) and can also read known working commercial disks.


Is there a better technique? Is it better to use the Floppy Drive test on MacTest Pro? (although not an option my older Macs that don't have a 68030 and thus not supported by MTP). What have you guys done? Or do I have massive OCD and none of the rest of you care?

 

[jmacz]

Note, with the above, I think the drives still could be off by a little bit and result in A working with B, and B working with C, but A doesn't work with C. The use of multiple disks and across sizes helps provide more coverage but still there's room for a difference.

Hoping someone has a better technique or tool.

 

[croissantking]

I’ve been able to tweak my drives by unscrewing and rotating the stepper motor.

I don’t understand how adjusting the zero track sensor could fine-tune things. It’ll either be off and the drive won’t work at all, or it’ll be aligned right and the drive will work. It just sets where the drive heads think track 0 is, but the motor moves in complete steps so if it’s slightly off then the motor will either be 1 complete step too far forward or 1 complete step too far backward.

 

[jmacz]

I don’t understand how adjusting the zero track sensor could fine-tune things. It’ll either be off and the drive won’t work at all, or it’ll be aligned right and the drive will work. It just sets where the drive heads think track 0 is, but the motor moves in complete steps so if it’s slightly off then the motor will either be 1 complete step too far forward or 1 complete step too far backward.

In terms of 'tuning', I was using that term more in aggregate, ie to get all my drives to be consistent with where track 0 is.

But in terms of the adjustment, perhaps I have the wrong mental model of how it works? I did not think there were known physical tracks on the disk itself (like a vinyl record) but that the tracks are purely determined by where the head writes it, which is governed by the stepper motor and tweaked by rotating it on each floppy drive. So across many drives, you could have a different start position for track 0 and thus the exercise is to get them all to start at the same spot. I guess part of the reason I thought this is because the rotation of the stepper motor isn't something that has clicks to allow you to select the start track, but it's a free rotation meaning near infinite starting locations depending on how much you turn it. And then with that in mind, I had thought there was some tolerance and thus if it was slightly off between drives A and B, it might still work, and if it was again slightly off between B and C, it might still work, but between A and C, it might not. But again, maybe I have the wrong mental model?

 

[croissantking]

In terms of 'tuning', I was using that term more in aggregate, ie to get all my drives to be consistent with where track 0 is.

But in terms of the adjustment, perhaps I have the wrong mental model of how it works? I did not think there were known physical tracks on the disk itself (like a vinyl record) but that the tracks are purely determined by where the head writes it, which is governed by the stepper motor and tweaked by rotating it on each floppy drive. So across many drives, you could have a different start position for track 0 and thus the exercise is to get them all to start at the same spot. I guess part of the reason I thought this is because the rotation of the stepper motor isn't something that has clicks to allow you to select the start track, but it's a free rotation meaning near infinite starting locations depending on how much you turn it. And then with that in mind, I had thought there was some tolerance and thus if it was slightly off between drives A and B, it might still work, and if it was again slightly off between B and C, it might still work, but between A and C, it might not. But again, maybe I have the wrong mental model?

Actually I think maybe I misunderstood what you were saying. I agree with all of the above - I thought you were talking about moving the zero track sensor (the one that moves back and forth) but it now sounds like you’re rotating the motor, which makes sense to me.

 

[jmacz]

Ah ok.

So far played with six floppy drives. 3 of them seem spot on and fully interchangeable in terms of reading/writing disks. The other 3 have some issues. One can read but errors writing (something else is wrong with this one). Second one seems to work just fine for 800K/1.4MB disks but cannot write 400K disks. The third one seems to have random errors from time to time.

Six more to go.

 

[jmacz]

Haha, well I guess this exercise was fruitful. Just figured out that the one drive that can't write, well, that drive works fine in another Mac so something's wrong with the SE/30 it came out of. I must have another busted trace somewhere, already fixed a few on this one.