dkedrowitsch
Experienced Member
In case anyone runs into a similar problem, below is a quick recap of a recent repair I just completed.
For reasons of pure nostalgia I've been looking for a clean Miniscribe 3212 5-1/4 HH 10-meg hard drive to use in my XT clone, and couple weeks ago a very nice looking one popped up on eBay for less than the usual fortune most sellers ask for hard to find drives, so I snagged it. Unfortunately after it arrived and I tested it, it would begin to spin up and eventually fail with a flashing 0100 fault code which meant the CPU didn't detect an index pulse.
Without schematics it was a bit of a challenge to troubleshoot, but I eventually traced where the index signal was expected to come from, a single pin on the spindle motor header. Turns out the hall-effect sensor in the motor was bad and was not pulling the index pin low to signal the CPU whenever the motor made a full revolution.
Fortunately this particular drive has an exposed flywheel on the spindle motor, so I was able to fasten a 3mm magnet to the flywheel and mounted a simple unipolar hall effect sensor to the underside of the HDA. I tied the new sensor into the index circuit on the PCB and BINGO, the drive spun right up and has quiet bearings to boot!
Unfortunately it was now flashing a different error code. This time is was 1000, unable to uncover track 0 sensor. This one was an easy fix, a single drop of light oil on the outer bearing of the head positioning stepper motor and the drive would spin up, perform self tests, and was ready to attempt low level formatting.
The LLF didn't go so well. It succeeded, but about 20% of the media was marked as defective. Plus I could hear what sounded like "dragging" as the heads were positioned around the outer 1/3 of the platter.
I decided to open the drive and take a look, and with a heat gun I was able to lift the warranty sticker without damage so the drive still looks untouched. The oxide on the platter appeared to be degraded due to I assume age or the environment, although externally the drive looks brand new. Under a magnifying glass the surface of the oxide film was scattered with tiny black deposits which the heads had been dragged across, hence the scraping sounds. Crap.
Since I have several common Miniscribe 3650s, I stole a "clean" platter out of one and transplanted it into the 3212. The hub, outer diameter, and thickness of the 3212 platter and 3650 platters appeared to be identical so I figured what the hell. Using nothing more than latex gloves and careful patience, I shaped folded pieces of paper into wedges that fit between the steel spring arms to lift the heads away from the disk, and unscrewed the head assembly from the positioners to allow the platters to be removed, and simply removed the 3212 platter and dropped one from the 3650 back in place. Reversing the procedure, I carefully reassembled the 3212 without incident and not a single mark on the surface of the disk.
I fired it back up to attempt another low level format--this time not only did it succeed...not a single bad sector was detected--just like the factory label indicates. Bonus!
I've been using the drive for a few days now and after a battery of tests and heavy use it seems to perform as good as it looks, brand new.
Anyway, hopefully this is interesting to some and maybe a little useful too.
Dieter
For reasons of pure nostalgia I've been looking for a clean Miniscribe 3212 5-1/4 HH 10-meg hard drive to use in my XT clone, and couple weeks ago a very nice looking one popped up on eBay for less than the usual fortune most sellers ask for hard to find drives, so I snagged it. Unfortunately after it arrived and I tested it, it would begin to spin up and eventually fail with a flashing 0100 fault code which meant the CPU didn't detect an index pulse.
Without schematics it was a bit of a challenge to troubleshoot, but I eventually traced where the index signal was expected to come from, a single pin on the spindle motor header. Turns out the hall-effect sensor in the motor was bad and was not pulling the index pin low to signal the CPU whenever the motor made a full revolution.
Fortunately this particular drive has an exposed flywheel on the spindle motor, so I was able to fasten a 3mm magnet to the flywheel and mounted a simple unipolar hall effect sensor to the underside of the HDA. I tied the new sensor into the index circuit on the PCB and BINGO, the drive spun right up and has quiet bearings to boot!
Unfortunately it was now flashing a different error code. This time is was 1000, unable to uncover track 0 sensor. This one was an easy fix, a single drop of light oil on the outer bearing of the head positioning stepper motor and the drive would spin up, perform self tests, and was ready to attempt low level formatting.
The LLF didn't go so well. It succeeded, but about 20% of the media was marked as defective. Plus I could hear what sounded like "dragging" as the heads were positioned around the outer 1/3 of the platter.
I decided to open the drive and take a look, and with a heat gun I was able to lift the warranty sticker without damage so the drive still looks untouched. The oxide on the platter appeared to be degraded due to I assume age or the environment, although externally the drive looks brand new. Under a magnifying glass the surface of the oxide film was scattered with tiny black deposits which the heads had been dragged across, hence the scraping sounds. Crap.
Since I have several common Miniscribe 3650s, I stole a "clean" platter out of one and transplanted it into the 3212. The hub, outer diameter, and thickness of the 3212 platter and 3650 platters appeared to be identical so I figured what the hell. Using nothing more than latex gloves and careful patience, I shaped folded pieces of paper into wedges that fit between the steel spring arms to lift the heads away from the disk, and unscrewed the head assembly from the positioners to allow the platters to be removed, and simply removed the 3212 platter and dropped one from the 3650 back in place. Reversing the procedure, I carefully reassembled the 3212 without incident and not a single mark on the surface of the disk.
I fired it back up to attempt another low level format--this time not only did it succeed...not a single bad sector was detected--just like the factory label indicates. Bonus!
I've been using the drive for a few days now and after a battery of tests and heavy use it seems to perform as good as it looks, brand new.
Anyway, hopefully this is interesting to some and maybe a little useful too.
Dieter