• Please review our updated Terms and Rules here

Resuscitating Old Storage/Hard Drives

L

lyonadmiral

Veteran Member
Joined
Jun 3, 2009
Messages
2,430
Location
Peru, New York
I think this topic has been beat to death; not just a dead horse but a greasy stain in the pavement, but I'd like to start fresh if I can.

There are a lot of modern day solutions to permanent storage for vintage systems; XT-IDE/CF is the big one, but in the spirit of keeping systems wholistic, what do you do for old MFM/DLL hard disk drives? I have an IBM 5161 with two full height 20 meg hard disk drives in them, they're loud but not screeching and both successfully low (unconditional) and high level format. On the other hand, I have two PS/2 8525's each with hard disks that while will successfully low level (unconditional) format, don't want to low level (conditional) or high level format at all. They're not loud, they don't screech, but just don't want to fully function. Any hope at all for those or chalk it up to just bad luck?
 
lyonadmiral;n1206763 not just a dead horse but a greasy stain in the pavement said:
Can't help with your problem, but you could quit your job and go to NYC and be a comic. Haha Hoho Hehe.
Click to expand...
 
Spinning media forever!

The problem with most low-level programs, especially those in ROM BIOSes, is that they usually do no verification whatsoever. Heads can be crashed, or a data cable can be disconnected, and they will still report success. It is up to the high-level format programs (such as DOS Format) to detect errors, but those will unhelpfully just bomb out at the first serious error.

So before giving up on the drives, I would at least try to get a better idea of what is really going on. SpeedStor: http://minuszerodegrees.net/software/speedstor.htm can do a read verification scan. If there are NO readable tracks, then I might suspect a cabling or electronics failure. If tracks on one head are bad, then it might be a crashed head. If track zero is bad, but others are readable then it might be a damaged track zero.
 
Another thing to keep in mind to is that the tolerances on electronic devices (chips, capacitors, resistors, etc.) change as they age. On older devices it is more likely that somebody can debug a problem, or even an intermittent problem, but it's going to require skills and equipment - software alone isn't going to do it.

Here is a simple example - one of my Xircom Ethernet adapters went bad. It turned out to be the on-board FLASH that holds the MAC address and device specific data. It can be recovered, but I don't have the programmer or soldering equipment I need to it myself. And the Xircom is nowhere near as complex as a hard drive.

Spin your drives up and exercise them once in a while. That's the best solution for preserving them. Or at least detecting when things start to go bad.
 
Old MFM/RLL ST506 interface hard drives go bad for several reasons in my experience:
  1. The medium (coating) itself degrades. Gradually, bad spots grow and erupt. Nothing to be done here.
  2. Head crashes--if the failure damages the head, the drive is beyond repair. If it damages the medium, you may get away with just a bad spot.
  3. Mechanical failure--bearings go dry; plastic parts degrade go gummy or crumble. Motor winding short or go open.
  4. Drive electronics fail--capacitors die, ICs fail, etc.
Of these, perhaps 3 has some measure of recovery and 4 is possible only if the problem can be diagnosed and repaired by a skilled person. The others likely relegate the device to paperweight status.

I have no doubt that in 30 years, people will be wondering how to recover/repair flash-based devices.
 
I finally cracked one open and you wouldn't (or maybe you would) believe the amount of specks inside from where either the head scraped surface material off the platters or it began flaking off. If you can check out this video https://youtu.be/aFcGn8IDV5s (once it finally uploads).
 
You must log in or register to reply here.
Back
Top