“Sigh-Quests”: More forgotten cartridges arrive, but fail to turn up data

A while ago, I was in contact with a reader about their long-forgotten set of Syquest cartridges and their Syquest drive. As it turns out, it was a set of 88Mb cartridges with matching drive, connected to a computer of an appropriate vintage, but no matter what was done, the cartridges stubbornly refused to mount.

After walking through the regular SCSI bus troubleshooting and basic Syquest LED diagnosis, it seemed the drive was fine and all was good but maybe something was lacking in the software on the Macintosh. I advised that they might have better luck trying with a PCI SCSI card and Linux, like my previous recoveries.

After a while, I got a response that the drive had failed entirely, and in the end nothing was achieved. As per my nature when it comes to these exotic media, I decided to extend an offer of free recovery provided the media are posted to me as I know my drive is still operational, albeit somewhat shaky.

Today, the stack of six disks arrived. I was not prepared for what was to happen next.

The Cartridge has Landed …

As I have been madly busy at this time, and somewhat absent minded, I immediately unpacked the box of cartridges, booted the recovery box, inserted a known good cartridge to verify the system was working (which it was) and then rolled up my sleeves to get my ddrescue running.

But this is where the first hint of problems arose. I picked out one cartridge to insert it and immediately it showed signs of distress. The cartridge didn’t quite fully get up to speed when the drive tried to load its heads onto the platters, failed and then retried several times. I immediately ejected it and reloaded it thinking it was just bad cartridge alignment.

The second time, the cartridge reached full speed and the heads loaded onto the media. This is when I heard the distinct light scraping noise that is the sound of a hard drive head meeting a contaminated surface. You would know this if you ever disassembled a hard drive, left a fingerprint on the platter and powered it up. The drive tried to read the format on the cartridge, took a little longer than usual and then went to idle, flashing a green LED which seems to indicate all is okay.

screenshot-def

Sadly, all was not okay. Any attempt to read resulted in an immediate response from the drive of Defect list error. This is an indication that the drive was not able to read the defect management area which describes the bad sectors and reallocations on the media. The same result was seen using sginfo. Nothing could be read. I ejected the cartridge.

I then tried to put in my known good cartridge just to check everything was good. It wasn’t. The known good cartridge too was reporting Defect list error.

A Cartridge Virus?

In a scene which is all too reminiscent of the furor that erupted around the ZIP click-of-death, I wondered if the cartridge had somehow damaged the drive. Judging from the sounds, it was indeed possible that the heads are now very much soiled or even damaged.

First aid was the order of the day – I disassembled the top lid by removing two screws and decided to clean the heads using a piece of exercise book paper dipped in high purity ethanol. The spacing between the heads was too small for a cotton bud, and its fibres could catch on the edges of the head and pull it off or misalign it from its fragile suspension. I very carefully rubbed the paper on the head surfaces, taking care not to poke or prod the head, and reassembled the drive hoping for the best.

Indeed, after the clean, the drive was able to once again read the known good cartridge. I then tried a second cartridge from the stack … no dice. Defect list error again! The other cartridges generally sounded better, but still had a nervous scratchy sound to them. The drive was not so dirtied as to stop reading the known good disk, so I alternated between the supplied cartridges and the known good disk until when the last cartridge was reached, the heads were once again dirtied to the point of no longer recognizing the known good disk and first aid was performed yet again.

The Coronial Inquest

Seeing as I was volunteering my time and equipment, and had made clear my terms included the rights to do whatever I felt was necessary and I was not to be held responsible for negligence, I decided the right course of action was to determine root cause of failure. To do this, a clean-looking cartridge was torn apart for examination.

20160201-1810-2319

The media surface shows white coloured contamination smeared across the media surface in the direction of rotation.

20160201-1809-2318

The soiling was rather extensive and affected the inner diameter severely, but was also spread elsewhere across the platter. Because of the design of the cartridge having an exterior plate that is riveted to the hub, inspection of the underside of the media was not possible.

Tearing down other cartridges in the set revealed very similar traces of contamination of the media surface. Such contamination will cause head crashes, fly height issues and damage to the head which will prevent proper readback of the data. The contamination alone rules out easy data recovery.

20160201-1821-2321

A close look at the cases seems to suggest it may have been stored in a damp environment, as some mould or mould residue can be seen. This suboptimal environment may have led to mould spores being deposited on the surface of the disk.

My hypothesis is: when the owner tried to read these in their Syquest drive, these spores were mushed across the surface of the disk as the drive tried to read the system formatting on the disks. The disks may have all been stored together in the suboptimal environment. Eventually, as the owner may have tried a few cartridges to rule out a single cartridge failure, the contamination was mushed on the surface of every disk, and potentially spread from one disk to another, or from the set of disks into the original drive causing the original drive to fail entirely and all disks to become contaminated.

A Last Ditch Attempt

Contaminated media is something most people, including myself, are not well equipped to handle. Worse than this, the design of the Syquest cartridge uses brass rivets to secure the outer load plate to the platter itself (picture below is with the protective aluminium tape removed). This prevents the ability to survey the underside of the media or to treat the media with any ease.

20160201-1837-2322

My first attempt, knowing that the media likely has a diamond-like overcoat, was to clean the top platter only using high purity ethanol and buff it clean with a new oil-free microfibre cloth (while wearing gloves) and then use air from a photographic blower bulb to remove the excess dust.

Sadly, while this made the cartridge quieter, the drive was still not able to establish its defect lists, and hence, could not read any data from the disk (even using SCSI read-long commands).

20160201-1841-2323

In my desperation, I tried to clean the rear surface of the media as well, by sliding low-lint tissue paper into the rear, soaking that in high purity ethanol and rotating the platter. Unfortunately, due to the low clearances, buffing the rear to a clean spot-free finish was very difficult, and it seems that the spots may have interacted with the head resulting in a drive sounding a bit rattley and no improvement on read-out.

Trying to gain better access to the rear, I tried to break apart the rear plastic itself, leaving the platter assembly intact. I reasoned that if the rear was cracked into half to release the hub assembly, due to the copious amount of screws and top-cover support, it should be possible to load the cartridge back into the drive.

Sadly, this attempt went wrong as the plastic was much more pliable than brittle, and the hub itself ended up bending rather than the plastic snapping as I had intended. I had irreparably damaged this cartridge, and I felt really bad about it. Had it been built with a torx screw hub like other drives, it may have been possible to remove the platter, clean it, wash it with distilled water, and have it potentially read back barring no other faults but the design really worked against me.

Conclusion

Very regrettably, I had hoped to be of assistance, but instead, I did not achieve anything of value to the owner, and I even managed to destroy one of their cartridges pretty much for good. No wonder I call this the sigh-quest. At this stage, I’m unsure if my drive has survived unscathed after seeing all this contamination – my bet is that my drive has probably accumulated enough damage from this ordeal that it be permanently retired as well. Very unfortunate outcomes all round. I just hope I will be forgiven.

I suppose the take-away message here is not to be too hasty and skip physical media inspections on the assumption that previous media of the same sort were “just fine”. Where possible, spotting things like fungus, mould and other contamination before putting it in your drive could save its life. In the case of Syquests, this means removing screws and opening up the cartridge itself to get a good look, as the case design prevents you from getting a clear image of the platter.

The other message might be to be careful when trying out old hardware. Don’t assume that one piece (e.g. the drive) is functioning fine, and start trying each and every disk you have as that could damage all the disks. Likewise, proper storage of equipment is necessary to ensure they don’t succumb to humidity and temperature impacts – maybe in future, store your media in sealed bags (or doubly sealed) with desiccant packets.

About lui_gough

I’m a bit of a nut for electronics, computing, photography, radio, satellite and other technical hobbies. Click for more about me!

This entry was posted in Computing and tagged , , , , . Bookmark the permalink.

4 Responses to “Sigh-Quests”: More forgotten cartridges arrive, but fail to turn up data

  1. Mark says:

    I recovered data from several rather heavily contaminated Syquest cartridges by cleaning them in an ultrasonic cleaner (using the “If it jams force it, if it breaks it needed fixing anyway” philosophy.

    Scrupulously cleaned the cleaner. Ran three cycles with DD (distilled / deionized) water. Change water and re-clean the cleaner each cycle. Cartridges suspended vertically to reduce chance of schmutz re-settling on the plater. Added a couple of drops of Kodak PhotoFlo wetting agent on the last cycle. All but one cartridge was recovered… that one had visible physical platter damage. You probably don’t want to use the highest setting on industrial grade cleaners…

    • lui_gough says:

      Interesting tip. I don’t have access to any ultrasonic cleaners, and maybe not one large enough for a whole cartridge, but I might investigate.

      Thanks,

      Gough

      • Mark says:

        Somebody at your university has got to have one…

        Beware of the really high power ones. Imploding micro-bubbles can have a horrendous energy density that can actually “texture” the surface of materials.

        • lui_gough says:

          Indeed they may have, as we do have a lot of equipment, but finding it and getting access/work-safety approval/etc is the issue. You would think that they would operate on a “quest for knowledge” style basis with friendly people willing to help, but sadly, this is rarely the case as departments are getting cash strapped etc. Even getting access to equipment for legitimate PhD research was such a pain in some cases, we had to rely on friends in other universities to post us gear which we would post back because it was easier to do it that way!

          – Gough

Error: Comment is Missing!