Perhaps this might be useful to others: the drive I did this on is a Micropolis 1325.
From a discussion on the cc-tech mailing list, I got a description on how to fix the "sticky bumper" problem on a Micropolis ST-506 drive. These drives have rubber "bumpers" as end-stops for the drive head. And after many years this rubber becomes sticky, resulting in the head not having enough power to "unstick" itself from the bumper when powered up.
The solution is (for me at least) quite scary: open up the drive, move the head and put a piece of paper between the rubber "bumper" and the head part. Then all will be well (if the drive doesn't have other problems).
So I did. First I had set up everything I needed: a psu to test the drive with, some post-it paper, screwdriver, scissors, a small flashlight.
The top cover has ten screws, and six of them is partly or totally obscured by the label. I simply peeled away the label above the screws.
After unscrewing, I used a small flat-blade screwdriver to loosen the cover; the gasket made it stick to the housing. Then I lifted the top cover carefully off (on the inside, the airflow duct is connected to it), and started to look for the bumpers.
Since I haven't seen the inside of one of these drives before, it took a while before I realized that the "bumpers" where inside the head assembly; there is only two small slits where you can see them when you move the head. There is also a locking mechanism that prevents the head from moving freely if the drive isn't powered on.
When I moved the head assembly with my fingers, there was very little resistance; I started thinking that maybe there was something else wrong with this drive.
But I cut a small strip of post-it paper (maybe 2 or 3 mm wide), moved the head a bit and put it into the hole between the head and the bumper that the head rested on when turned off. Simply releasing the head made the paper stick to the bumper.
Testing; I powered up the drive (yes, with the cover off) and now the head moved out when the drive had finished spinning up. Before the paper, the head hadn't moved at all. Aha!
It was the "sticky bumper" problem after all. I re-assembled the drive, powered it on, and let it be on for about four minutes. Success!