ST412 drive or controller fault?

Interesting The fact that u are getting solid AMNF errors in two 128 cylinder bands also suggests to me some sort of controller problem The drive doesn't know about AMs; that's handled by the controller. The large amount of AM errors suggests to me the tracks were not written (or overwritten with a different controller). If the tracks were formatted u should be seeing seek errors or data errors. The drive interface is step and direction; during a low level format it single steps so it's internal buffering of the step pulses shouldn't be invoked. But even if u wound up at a wrong track there should be count fields that tell the controller what track the drive is actually at and then the recovery becomes controller firmware dependent (I think some controllers just re-seek and continue) Have u done a low level (BIOS) format? Tom -----Original Message----- From: Jules Richardson [mailto:jules.richardson99 at gmail.com] Sent: Saturday, February 15, 2014 4:13 PM To: xx Classiccmp mailing list Subject: ST412 drive or controller fault? Just curious as to what the expert opinion is on this - I've got a ST412-type drive here (it's a Miniscribe with 615/4/17 ~20MB geometry) which refuses to read from cylinders 128-255 or cylinders 384-511 - giving address mark not found errors - although cylinders outside those two ranges are fine. The way the faults lie exactly within a couple of specific bit patterns (01xxxxxxx and 11xxxxxxx) for the cylinder number makes it seem more like a controller problem to me - but I don't know how much logic there is on a ST412-type drive, what with the buffered seek ability. Controller passes internal/RAM diagnostics OK (but how exhaustive those are, I don't know). Issuing a drive recalibration and/or controller reset prior to attempting access to the problem areas makes no difference. Maybe the drive was simply never formatted on those cylinder groups for some reason - but that seems rather unlikely (and the filesystem on the drive is definitely unwell, suggesting that it's trying to access data that's no longer there, although I've not gone to the extent of picking the FAT apart to verify this) I think all the errors I've seen before on these kinds of drives have been totally dead, seemingly random, or confined to all/parts of a surface - not in cylinders. cheers Jules