• Please review our updated Terms and Rules here

HDD Track 0 repair tool

J

jammaster

Member
Joined
Oct 15, 2016
Messages
24
Location
France
Hello!
I'm not sure if the topic was already posted?
My ST-4096 MFM seem to be dead.
I have opened it and the park lock is ... locked (!)
The head still blocked on the inner of the drive.
In this drive, I believe the lock system release the heads only when the track 0/system area was found (?). St-4096 have autopark function with voice coil heads... (great drive)
Is there a tool to perform a factory calibration. (seek commands resulting mounting a new base track)
The Led drive light at startup and still blinking very fast (searching track 0 i presume) and i got a 1780 error.

Spinrite II ???
 
Wouldn't a magic bullet be nice? If anybody has one I'd like to try it on several of my drives with faulty track 0s.

FWIW, while SpinRite II is quite useful it can *only* work on a valid, readable partition. So if the drive does not have the ability to be accessed by DOS SpinRite is useless.
 
Maybe someone clever could hack the BIOS to add a few tracks on every operation, then we'd just need to partition the same number of tracks short.
 
It's not that simple, James. Usually, part of the command set at initialization is a "recalibrate" command, where the heads retract to track 0, the IDs are read, and things proceed from there. In fact, when a soft error is encountered, part of the recovery sequence is usually a "recalibrate" operation.

It's been suggested that simply moving the track zero sensor inwards a bit might produce enough new surface under the heads to make things work, but I've never heard of that working successfully. Of course, it's completely out of the question where a servo track is involved.
 
Yes sometimes if it has an external actuator it can be adjusted by adding a shim to the stop
 
Chuck(G) said:
It's not that simple, James. Usually, part of the command set at initialization is a "recalibrate" command, where the heads retract to track 0, the IDs are read, and things proceed from there. In fact, when a soft error is encountered, part of the recovery sequence is usually a "recalibrate" operation.

It's been suggested that simply moving the track zero sensor inwards a bit might produce enough new surface under the heads to make things work, but I've never heard of that working successfully. Of course, it's completely out of the question where a servo track is involved.
Click to expand...


I think, "modern" voice coil actuator seeking is always dynamic/relative
The ref track is magnetised in factory i believe (?), and without this segment, the calibration procedure is almost a fail.
Am i wrong?
 
Last edited:
jammaster said:
I think, "modern" voice coil actuator seeking is always dynamic/relative
The ref track is magnetised in factory i believe (?), and without this segment, the calibration procedure is almost a fail.
Am i wrong?
Click to expand...

The servo appears across the entire platter
Some use servo signals between sectors, some use a dedicated surface
If you have a drive with an odd number of heads, it probably has a dedicated servo surface

Dave Gesswein was able to recover data from a Miniscribe by injecting an offset into the servo loop
http://www.pdp8online.com/3b1/drive.shtml
 
Last edited:
jammaster said:
My ST-4096 MFM seem to be dead.
I have opened it and the park lock is ... locked (!)
The head still blocked on the inner of the drive.
In this drive, I believe the lock system release the heads only when the track 0/system area was found (?). St-4096 have autopark function with voice coil heads... (great drive)
Click to expand...
Of course, it is up to the drive's engineers to decide on what criteria they use for releasing a park lock.
Adequate spindle speed could be one criterion (i.e. heads known to be 'flying').
 
Chuck(G) said:
It's been suggested that simply moving the track zero sensor inwards a bit might produce enough new surface under the heads to make things work, but I've never heard of that working successfully. Of course, it's completely out of the question where a servo track is involved.
Click to expand...
Stone said:
Same here. Talk is cheap.
Click to expand...
mikey99 recorded success for a stepper motor based drive at [here].
 
One of my ST 4096s has a park lock that sticks. When the drive comes up to speed, you can hear it try to release three times, then gives up. If you cycle it enough times, it will finally release. Then it works fine.

It doesn't seem like it would be a track zero issue since I don't think the heads could read anything until the park lock is released.
 
I don't recall--does the park lock take the form of a solenoid with a pad that rubs against the spindle drive? I've seen those fail--generally, the cheap way out is to simply discard the thing. But perhaps a little light oil on the armature may free things up.
 
Chuck(G) said:
I don't recall--does the park lock take the form of a solenoid with a pad that rubs against the spindle drive? I've seen those fail--generally, the cheap way out is to simply discard the thing. But perhaps a little light oil on the armature may free things up.
Click to expand...

When the HDD shut down the lock mechanism is open a couple of seconds before entering in is lock position.
Si can seek the heads in the center of the platters.
When restarting the HDD, nothing. The heads sleep on the center of this dawn granddad!

http://theangel540.free.fr/WP_20161118_23_58_31_Pro.jpg
 
On the Seagate 4096 the park lock is a solenoid, but I think it is separate from the spindle brake. The head carriage is a voice coil actuated affair mounted on rails with roller bearings, so if there wasn't something to hold the heads in place when the drive is off, they could just slide out away from the parking zone.
 
I'm wondering what are the chances of bringing such a drive back to life, after it's been open, and not in a dust-free chamber.
 
You must log in or register to reply here.
Back
Top