ST412 drive or controller fault?

On 02/16/2014 07:27 PM, Jules Richardson wrote: Well, I got some time just now to hack some code for testing this. The controller doesn't seem to support seek operations, but doing a read (which takes c/h/s parameters) and ignoring the returned data seemed like a fair substitute. And... wow, is this thing ever confused... So, at power-on (and after the drive's up to speed), the stepper goes through more than one revolution before settling at what I'm assuming is cylinder 0 (the interrupter on the end of the shaft is in line with the detector) - in other words, available head travel doesn't limit it to < 1 revolution. If I attempt successive reads from cylinders 0-614, I can see the head positioner advance as expected from cylinders 0-63. At cylinder 64, the stepper jumps to a completely new part of the arc, but then advances "linearly" (if you see what I mean) from there in the range 64-127 - I don't know why this range wouldn't simply follow on from where cylinder 63 was! At the sector level there's obviously valid data there, although whether it's the data that DOS is expecting to be in that physical spot on the disk is another matter. Now, in the cylinder range 128-255, the stepper doesn't move at all, it just stays stuck where it left off at cylinder 127. At cylinder 256, it appears to reset to the cylinder 0 position (but it's hard to tell for sure where the heads physically are on the disk; remember that that stepper is capable of > 1 revolution, so I'm just going by the interrupter position). The stepper then advances "normally" for reads in the range 256-319, but at 320 it mimics the behaviour of cylinder 64, suddenly jumping to a different part of the arc, before once again advancing "normally" for the 320-383 range. Between 384-511, we see a repeat of 128-255, with the stepper not advancing (staying where the read from cylinder 383 left it). At cylinder 512, it suddenly jumps to - apparently - cylinder 0 position, and then the stepper advances "normally" up until the last cylinder on the drive (611) I'm not sure what my next move is - possibly hooking some magic up to the step/direction lines so I can monitor exactly what they're doing, which would at least enable me to determine if it's the drive or the controller at fault. I'm particularly curious whether there really is a sudden burst of step pulses between read attempts from cylinder 63 and 64, for instance. cheers Jules