Mac 512Ke with a VERY strange floppy drive issue...

[DeChief]

I picked up a free Mac 512Ke with extended keyboard and mouse 6 months ago and have only just now gotten around to messing with it. When I first got it home, it powered on to a question mark floppy, so naturally I was ecstatic to have just been given a free working Mac and put it on my shelf to be cleaned later.

Fast forward to last week and it was throwing up a sad mac with error 028000, pointing to a faulty RAM chip at G12. So I de-soldered it and installed a socket in its place, and powering it up without the old chip gave me the same error. I ordered 3 new chips and installed one of them today, and it fired up to a question mark floppy symbol again! Before servicing the internal FDD I decided to give my external 800K drive a quick go and it booted first try with a copied System 6 disk I had laying around.

I'm not sure if that was a fluke or what, because it never worked with that floppy ever again, and I was forced to initialise it when I checked the disk in my SE/30. After that, every test I've conducted was with one of the two "512/800K System Tools" floppies the machine came with. 1 of them booted a few times but with a very slimmed down version of the menus (there wasn't even a Shutdown option and I couldn't eject the floppy by dragging it to the trash), and then it eventually stopped working. The other disk has continued to work on and off in the 512Ke despite reading and imaging perfectly in my SE/30 with DiskCopy 4.2. I was able to make a new disk using that image, and it works with the same inconsistency as the original.

In-between all of this, I cleaned and lubricated the internal drive and tested that it worked, and it displayed the exact same symptoms as the external drive... until I put it back into the 512Ke. For whatever reason, the drive only works when it's sitting at this specific angle outside of the machine, and even then it's just as inconsistent as the external:

I tried another yellow stripe floppy cable and it acted the same, so I re-flowed the solder joints of the FDD connector on both the FDD and 512Ke board itself and used some DeOxit D5 - nothing changed. I eventually managed to get it working while properly positioned inside the machine with the floppy cable twisted like this, but it was still inconsistent and was now booting at HALF the speed as when the drive was outside of the machine.

The last thing I tried was putting the internal FDD into the external enclosure, and the external FDD inside the 512Ke, and it gave me the same results as above. I've beep-tested both of my yellow stripe cables with a multimeter and both ends have perfect continuity.

TL;DR summary:

- known-good external and internal floppy drives

- known-good boot floppies

- working inconsistently, partially dependent on drive position

This is driving me absolutely up the wall, what the hell is going on here?!

I reflowed the entire analogue board but all of the capacitors are original, could this possibly be a voltage issue? Usually I replace the capacitors on Mac Plus analogue boards even if they're working fine, just for the sake of longevity, so I will eventually do that for the 512Ke as well - not before further investigating this floppy problem though.

 

[Dog Cow]

I read your story this morning, had an observation to make, but decided against it. Now that I see that still no one has replied, I will go ahead and remark that the CRT emits radiation which interferes with the floppy drive I/O. The metal mounting bracket shields the drive against this radiation, but depending on how you have disassembled your Mac and its drive, the radiation may still be affecting the drive.

1 of them booted a few times but with a very slimmed down version of the menus (there wasn't even a Shutdown option and I couldn't eject the floppy by dragging it to the trash),

The earliest Finder versions from 1984 had no Shut Down command, and you could not drag a disk icon to the Trash to eject it.

 

[DeChief]

I read your story this morning, had an observation to make, but decided against it. Now that I see that still no one has replied, I will go ahead and remark that the CRT emits radiation which interferes with the floppy drive I/O. The metal mounting bracket shields the drive against this radiation, but depending on how you have disassembled your Mac and its drive, the radiation may still be affecting the drive.

The earliest Finder versions from 1984 had no Shut Down command, and you could not drag a disk icon to the Trash to eject it.

I had considered that, but unless the drive is sensitive to a CRT with a millimetre or so of variation, it's unlikely to be the problem otherwise the Mac would have been designed very differently. It's in essentially the same position as it was from the factory.

That also still wouldn't explain why the external drive is flakey despite being so far away from the CRT.

Ah I see, I guess I've never used anything below System 6 before.

EDIT: could the tube be leaking more than it's supposed to? I could try swapping in the tube from my SE/30, but that's a bit of an effort for something that likely won't help. I'd have to swap the yoke and I just got it perfectly aligned! ;(

 

[Dog Cow]

Your response is why I initially didn't want to make that remark about interference with the floppy drive: I wasn't convinced beyond a doubt that it was applicable to your scenario, based on your written description and the photos.

But that knowledge may be useful to you or others in future when working on a compact Mac and its internal floppy drive.

 

[DeChief]

Your response is why I initially didn't want to make that remark about interference with the floppy drive: I wasn't convinced beyond a doubt that it was applicable to your scenario, based on your written description and the photos.

But that knowledge may be useful to you or others in future when working on a compact Mac and its internal floppy drive.

Completely understandable, I'm glad you told me about that though. I put the 512Ke back together last night and now with the cover on and everything, the external drive is happily booting both 400K and 800K floppies while the internal tries for a while and fails with one of several errors (mainly ID = 15 or 26).

So if the CRT is causing these issues, what can even be done to fix it? Should I try my SE/30 tube in its place?

 

[Dog Cow]

while the internal tries for a while and fails with one of several errors (mainly ID = 15 or 26).

These are System Bomb errors? If so, both are Segment Loader errors.

I'm rereading your story again carefully, and I have some more questions and a hypothesis:

1.) The very first floppy you booted from you said  was a "copied System 6 disk I had laying around." Do you remember if this disk was locked or unlocked?

2.) Same question as before regarding the 512/800K System Tools disk that "eventually stopped working." Was this disk locked or unlocked?

My hypothesis is that your machine still has faulty RAM, but it's not a fault that the built-in RAM test can detect (Yes, this is possible because the startup RAM checks are not exhaustive). If your disks are unlocked, the Mac is writing some corrupted data back to the disk, which eventually causes the disk to stop working. Or, when reading from the disk into the file buffer, the contents are corrupted because of the faulty RAM.

So if the CRT is causing these issues, what can even be done to fix it? Should I try my SE/30 tube in its place?

If this Mac were in my hands, I would not monkey with the CRT until I had exhausted other avenues of investigation.

 

[LaPorta]

Another possibility: could the angle and cable-wrapping scenarios indicate a cable that has a broken wire that is only intermittently connecting? Meaning, when you test them they are ok, but if it moves just a tad, one wire breaks connection intermittently?

 

[DeChief]

These are System Bomb errors? If so, both are Segment Loader errors.

I'm rereading your story again carefully, and I have some more questions and a hypothesis:

1.) The very first floppy you booted from you said  was a "copied System 6 disk I had laying around." Do you remember if this disk was locked or unlocked?

2.) Same question as before regarding the 512/800K System Tools disk that "eventually stopped working." Was this disk locked or unlocked?

My hypothesis is that your machine still has faulty RAM, but it's not a fault that the built-in RAM test can detect (Yes, this is possible because the startup RAM checks are not exhaustive). If your disks are unlocked, the Mac is writing some corrupted data back to the disk, which eventually causes the disk to stop working. Or, when reading from the disk into the file buffer, the contents are corrupted because of the faulty RAM.

If this Mac were in my hands, I would not monkey with the CRT until I had exhausted other avenues of investigation.

I'm not sure what you mean by locked or unlocked, is that a software thing or are you talking about the write-protect notch?

Regardless, I managed to get the internal drive working perfectly! I tested with the external drive some more and realised that without the internal drive or cage, the further away the external drive was from the CRT, the more reliable it became (7cm+ was a good distance).

I talked to a few friends about it and brought up your CRT interference theory, and they suggested that the drive cage could have become magnetised somehow, possibly by spending a significant amount of time next to a large speaker. It turns out I own a spare drive cage that I must have salvaged out of a dead Mac years ago, so I swapped that one in and the drive has been working ever since.

I'm glad you told me about the RAM test not being exhaustive, is there a program you could recommend to do it properly?

Another possibility: could the angle and cable-wrapping scenarios indicate a cable that has a broken wire that is only intermittently connecting? Meaning, when you test them they are ok, but if it moves just a tad, one wire breaks connection intermittently?

I checked the continuity of all pins on both cables when twisted and when straight, and they checked out. It could still be a possibility but I can't eliminate that without hard-wiring a drive and I'm not going to bother until it stops working again.

 

[Dog Cow]

I'm not sure what you mean by locked or unlocked, is that a software thing or are you talking about the write-protect notch?

Yes.

I talked to a few friends about it and brought up your CRT interference theory, and they suggested that the drive cage could have become magnetised somehow, possibly by spending a significant amount of time next to a large speaker. It turns out I own a spare drive cage that I must have salvaged out of a dead Mac years ago, so I swapped that one in and the drive has been working ever since.

Good. I'm glad that you made some progress there.

the RAM test not being exhaustive, is there a program you could recommend to do it properly?

I use MacTest 7.0