Dennis Nedry
Well-known member
I have a Macintosh IIci that came in very bad shape, which is being discussed in a different thread. Someone had taken a screwdriver to it. There was a deep scratch on the logic board, but the floppy drive received the most damage. The circuit board was cracked in several places and the metal was misshapen. It physically could not accept a disk.
So first I carefully unbent the metal so that a floppy could slide in. This was successful, but there were further problems because the carriage would not snap down all the way...
Carefully, I took the whole thing apart. I got the circuit board separated from everything else, and I also got down to the bare chassis. I removed the top half of the carriage as well. To the sink I went, with a toothbrush, dish soap, chassis, and carriage. All 3 were thoroughly washed and dried.
I then proceeded to take some automotive grease and apply it to all moving points of contact. (I didn't have any white lithium grease...) I reinserted the carriage into the chassis and hooked up the 2 springs. Physically works perfectly, accepting a disk and manually ejecting properly.
I then reinstalled the eject motor, heads, zero-point alignment sensor, and the piece that holds the heads apart when the drive is ejected. Next up was the circuit board.
I took a close-up picture of the board and mapped out where each crack was in Photoshop. Then I drew on top of each trace in a separate layer from one point to another that went through the crack so I could solder a wire to bridge each trace that went through a crack. Using chopped up solid copper Ultra-ATA ribbon (my favorite), I proceeded to solder wires connecting all these points together. Very tedious and full of errors. I don't know how many times I went back with a tester and discovered a problem, or could that I soldered to the wrong point...
Eventually I finished this process and plugged the drive into a IIci. I inserted a disk, and what do I hear but NOTHING, followed by a weird sounding eject after a while. The drive did not spin and the heads did not move. On top of that, the eject motor had stopped at a weird spot and I coudl not reinsert a disk.
I swapped in an eject motor from another drive temporarily. Still nothing. Visual inspections of the board revealed nothing. Everything looked fine. No spinning, no head movement. Dead for sure, right?? NO!
There is a small switch inside the eject motor mechanism that was dirty or loose or something. blew it out and tightened it. That fixed the problem.
Next I looked at the chips on the board to find the stepper driver that powers the motor that moves the heads. I plugged the drive into a running Mac and took the voltage of each pin with respect to ground for each pin of this chip, then compared by taking the same readings to a known-good drive. There were discrepancies. Glancing at the datasheet, I found a supply pin that should be 12V and it was something like 1 or 2V. So I traced around on the board and discovered that Q4 seemed to be suspect.
Shorting this transistor in such a way as to act like the known-good one powered the drive RIGHT UP. The disk spun and the heads moved. Then it ejected with the X floppy on the screen.
I desoldered this TINY SMD transistor and replaced it with a normal-sized one that I had that was similar in spec. The drive motors were now functional. But it still would not read a disk! I cleaned the heads thoroughly with alcohol. Still nothing. Surely the heads are out of alignment, or damaged somehow. So I thought I'd swap heads with the known-good drive. In the process of removing the connector on the good drive, I pulled too hard and RIPPED the ribbon.
At this point, I have TWO bad floppy drives. OH NO!!!!!
So I was a bit bummed. The ribbon is impossible to repair. I tried scraping to expose the copper, in hopes to solder across. Nothing. The printed lines on the ribbon are well under 1 mm.
Later, I put the heads from the bad drive into the drive that used to be known-good. It worked! This tells me that the heads were probably not the problem with the damaged drive. Interesting.
As I thought about this severed ribbon more and more, the thought occurred to me to slice between each printed line on the ribbon with a razor blade. Then weave the razor blade between these "fingers" to separate so that I would be able to solder one connection at a time to a small piece of Ultra-ATA ribbon. This prevents soldering the traces together. Then I could solder the other end of this ribbon to the connector that had sheared off. This WORKED. When I was done, I applied hot glue to prevent the soldered area from flexing. These heads work now and the repair looks really good. I tested with an ohm meter and by putting them in the known-good floppy drive.
So I was back to figuring out why this damaged drive won't read floppies. So I booted the Mac off of a hard drive and inserted a blank disk and formatted it. The format was successful and I was able to copy a Finder and System file to it and reboot off of that disk. This disk was not recognizable in the known-good floppy drive, and the damaged drive blatantly fails to read any normal disk. This is indicative of head alignment issues, but these heads are known good!! Hmmm...
Next I looked at the zero-point sensor. When you loosen the screw, there is a slight bit of play, forward to back, of this sensor. So I pushed it all the way back, then all the way forward, and accurately as possible, moved it to the point halfway in between, then tightened the screw.
It works! Repaired against all odds. I talked to Mike Richardson and he says he has some auto-inject floppy drives that were out in the rain. With this experience, maybe I can get some of them to work! We'll see what kind of shape they're in and if I can fix them. I'll keep you all posted.
So first I carefully unbent the metal so that a floppy could slide in. This was successful, but there were further problems because the carriage would not snap down all the way...
Carefully, I took the whole thing apart. I got the circuit board separated from everything else, and I also got down to the bare chassis. I removed the top half of the carriage as well. To the sink I went, with a toothbrush, dish soap, chassis, and carriage. All 3 were thoroughly washed and dried.
I then proceeded to take some automotive grease and apply it to all moving points of contact. (I didn't have any white lithium grease...) I reinserted the carriage into the chassis and hooked up the 2 springs. Physically works perfectly, accepting a disk and manually ejecting properly.
I then reinstalled the eject motor, heads, zero-point alignment sensor, and the piece that holds the heads apart when the drive is ejected. Next up was the circuit board.
I took a close-up picture of the board and mapped out where each crack was in Photoshop. Then I drew on top of each trace in a separate layer from one point to another that went through the crack so I could solder a wire to bridge each trace that went through a crack. Using chopped up solid copper Ultra-ATA ribbon (my favorite), I proceeded to solder wires connecting all these points together. Very tedious and full of errors. I don't know how many times I went back with a tester and discovered a problem, or could that I soldered to the wrong point...
Eventually I finished this process and plugged the drive into a IIci. I inserted a disk, and what do I hear but NOTHING, followed by a weird sounding eject after a while. The drive did not spin and the heads did not move. On top of that, the eject motor had stopped at a weird spot and I coudl not reinsert a disk.
I swapped in an eject motor from another drive temporarily. Still nothing. Visual inspections of the board revealed nothing. Everything looked fine. No spinning, no head movement. Dead for sure, right?? NO!
There is a small switch inside the eject motor mechanism that was dirty or loose or something. blew it out and tightened it. That fixed the problem.
Next I looked at the chips on the board to find the stepper driver that powers the motor that moves the heads. I plugged the drive into a running Mac and took the voltage of each pin with respect to ground for each pin of this chip, then compared by taking the same readings to a known-good drive. There were discrepancies. Glancing at the datasheet, I found a supply pin that should be 12V and it was something like 1 or 2V. So I traced around on the board and discovered that Q4 seemed to be suspect.
Shorting this transistor in such a way as to act like the known-good one powered the drive RIGHT UP. The disk spun and the heads moved. Then it ejected with the X floppy on the screen.
I desoldered this TINY SMD transistor and replaced it with a normal-sized one that I had that was similar in spec. The drive motors were now functional. But it still would not read a disk! I cleaned the heads thoroughly with alcohol. Still nothing. Surely the heads are out of alignment, or damaged somehow. So I thought I'd swap heads with the known-good drive. In the process of removing the connector on the good drive, I pulled too hard and RIPPED the ribbon.
At this point, I have TWO bad floppy drives. OH NO!!!!!
So I was a bit bummed. The ribbon is impossible to repair. I tried scraping to expose the copper, in hopes to solder across. Nothing. The printed lines on the ribbon are well under 1 mm.
Later, I put the heads from the bad drive into the drive that used to be known-good. It worked! This tells me that the heads were probably not the problem with the damaged drive. Interesting.
As I thought about this severed ribbon more and more, the thought occurred to me to slice between each printed line on the ribbon with a razor blade. Then weave the razor blade between these "fingers" to separate so that I would be able to solder one connection at a time to a small piece of Ultra-ATA ribbon. This prevents soldering the traces together. Then I could solder the other end of this ribbon to the connector that had sheared off. This WORKED. When I was done, I applied hot glue to prevent the soldered area from flexing. These heads work now and the repair looks really good. I tested with an ohm meter and by putting them in the known-good floppy drive.
So I was back to figuring out why this damaged drive won't read floppies. So I booted the Mac off of a hard drive and inserted a blank disk and formatted it. The format was successful and I was able to copy a Finder and System file to it and reboot off of that disk. This disk was not recognizable in the known-good floppy drive, and the damaged drive blatantly fails to read any normal disk. This is indicative of head alignment issues, but these heads are known good!! Hmmm...
Next I looked at the zero-point sensor. When you loosen the screw, there is a slight bit of play, forward to back, of this sensor. So I pushed it all the way back, then all the way forward, and accurately as possible, moved it to the point halfway in between, then tightened the screw.
It works! Repaired against all odds. I talked to Mike Richardson and he says he has some auto-inject floppy drives that were out in the rain. With this experience, maybe I can get some of them to work! We'll see what kind of shape they're in and if I can fix them. I'll keep you all posted.
