ZIP SCSI General Issues - COD

[RSX11M+]

I know this thread is one of several on ZIP disks, but I have a specific concern that bears talking about.


Since I'm a new member of the SCSI / PDP-11 club myself, I've been lurking on those threads waiting for it be covered - but nobody has yet brought up the "COD" (Click Of Death) problem once well known to ZIP drive owners.

There is an excellent technical overview of this on grc.com and even some free software to help detect susceptible drives, write-damaged disks, along with specific recommendations for cartridge care and inspection. Prepare yourself for a bit of reading, as he delves into the subject in several pages. Please read them all if you're not already acquainted with the topic.

Steve Gibson is a highly reputable technical guy and I usually find his thinking and software well founded. His writings on COD are among these. Remember though, he's a geek like the rest of us, so his style reflects this heritage.
​

I am assembling resources to be able to share ZIP disks between PC based PDP-11 emulators, and my real systems. It turns out my MAC II collection has a number of drives that will serve nicely in this capacity, and should be of old enough vintage to be reliable (prior to the mfg runs known for COD) Nonetheless, I plan to qualify them by testing using Steve's utilities and to inspect all my disks for damage and readability before using or relying on them.

Having recently researched these potential problems and what I might do to preclude them, my purpose here is to both share this information and ask for other's comments and experiences.


To summarize:


The ZIP drive Click Of Death (COD) syndrome resembles characteristics of a Hardware Virus. It is possible to damage a ZIP cartridge in such a way that it is highly likely to damage the drives in which it is inserted. The resulting drive damage is then capable of inflicting the same harm upon other disks, and thus spreading to more units as if a "contagion" had taken root.

Additionally, some disks can have their format overwritten by an errant drive or one with intermittent power during a write operation. This can not be easily detected until data loss occurs, unless specifically tested for.

Both types of error, can cause a ZIP drive to exhibit COD behavior.


Perhaps I'm just late to the party, and this information is already universally known to everyone else. But with so many new members "discovering" the usefulness of this hardware for the first time, I thought it prudent to start this thread.

 

[Chuck(G)]

This is a rather old problem. I can remember chuckling along with my co-workers over a note posted by an overnight operator who did not have a lot of experience with CDC 844 disk drives. He saw a drive go "not ready" with a strange sound, so he moved the pack to a different drive, then tried a new pack on the old drive, etc. You could see the whole tragedy horribly unfolding blow-by-blow. By shift's end, he'd clobbered about 19 packs and 11 drives.

 

[SomeGuy]

I used to use Zip drives all the time back when they were popular and still have a SCSI Zip in my computer. I never encountered a problem where a faulty disk would damage another drive, but I always took care of my disks, kept them clean, and tested them often. If I thought there was a significant problem with a disk it would go in the trash before it caused problems.

Occasionally a sector would go "bad" or become weak for any number of reasons (writing error, power surge, magnetic interference, dish washer spots, who knows). Of course any drive would click like crazy trying to read damaged existing data, but using the SCSI card's low-level format, 98% of the time would restore it to perfect operating condition. No errors or-retries. The other ~2% of the time it was a physical disk error, and low leveling turned it in to a "hard" error that formatting would just skip over.

There was one drive that I accidentally dropped on a concrete floor (oops!), and after that it would just sit and click trying to read a disk. There was another drive that, I think, was jammed up with dust and was also just clicking. Oddly soaking that drive in alcohol fixed it - but don't try that except for a last resort!

 

[RSX11M+]

Yes - IOMEGA's problem is old, for sure.

To reiterate: I've personally seen the COD in "modern times" (less than 10 years ago) and can assure you the scenarios described above are real.

Since there is an inspection / prevention method available for both drives and media, I thought it would be a good idea to re-surface the information. You never know what eBay purchases have been subjected to. ZIP drives are a hot topic on the forums here lately - so be advised.


I think anyone from the PDP-11 era, must be aware of at least one story such as your CDC 844 wreck-a-thon, if not more. The most spectacular of these come from DEC service personnel, who were called in to deal with the carnage. The price tag of such a day, could be totally awesome - and that's in equipment damage alone, not counting the cost of data lost.

 

[Lou - N2MIY]

The zip disk is basically a miniature RL02. It has embedded servo information, one platter, and two heads. (it's even more similar than that with a reserved bad block replacement space too, but I digress...) The way that a head crash trashes an RL pack, and the trashed pack then trashes other drives' heads, the zip click of death appears to look similar to me. For this problem, on both types of drives, if recognized right away, the damage can be limited.

Your point about the unknown condition of ebay zip drives is a very good one. I suppose just as how I never put a pack with irreplaceable data in an unknown RL drive, the same caution should be taken with an ebay zip drive. Running Steve's diagnostics on unknown drives seems like a very good idea!

Lou

 

[krebizfan]

The Zip drives most susceptible to triggering contagious Click of Death came from production in late 1995 or early 1996 so it was resolved before the lawsuits started. A cost cutting method was not fully tested before implementation. I have an earlier Zip drive, nary a problem and later Zip drives work fine as well. I remember Iomega presented an article indicating the number of drives returned for repair was about 10% of production at about the time of the lawsuit which dropped back the more normal 1% after modified drives were introduced. I wish Iomega had followed up on the issue with a line of clear plastic cartridge shells. Opening the disk cover and rotating the disk to find a rip seems like a great way to kill the disk from dust exposure.

I don't purchase used Zip disks. The cost of not yet opened ones is still quite cheap; sometimes even cheaper than the used disks are offered.

 

[RSX11M+]

The Zip drives most susceptible to triggering contagious Click of Death came from production in late 1995 or early 1996 so it was resolved before the lawsuits started. A cost cutting method was not fully tested before implementation...

I knew of the issue back in the day, but wasn't personally effected. Do you recall what specific drive types it involved? I.E. - External, Internal, SCSI, Printer Port - it might be helpful.

Of course, a damaged disk might be able to hurt any drive it's inserted into, so all used drives should be treated as suspect until tested. I now keep a test disk specifically to use as a "sacrificial lamb" so no data is risked. All old media should be inspected prior to use.


There's another wrinkle I've been wondering about... (besides COD)

• Can anyone give us a dissertation on MAC vs PC formatted ZIP cartridges?
• Are they interchangeable in PDP-11 applications?

I recall a little driver-mania back in the day involving ZIP disks. (cobwebs clearing)

There were two flavors of ZIP disks: PC and MAC, plus a "universal" disk called a ZIP 100 Tools disk. The tools disk had the unique capability of configuring itself to a system type - once - upon insertion. Thereafter, it was dedicated to that system type. I never discovered a way to recreate a universal tools disk, as all became dedicated upon first insertion.

Users could encounter an interesting problem in their MAC systems. As the story goes, MAC ZIP disks contained a copy of the driver to access the disk. When inserted into a MAC, the driver in the OS was replaced by the driver on the disk under certain conditions. This could lead to problems, I don't recall the specifics, where the MAC OS had an inappropriate ZIP driver revision level.
​

All this begs the question of how a PDP-11 SCSI controller will recognize or deal with - cross-platform disks. Will a PDP-11 SCSI controller recognize an empty ZIP drive, or must a disk be present? You can kind of see where this is leading.

With regard to SCSI ZIP drives: I was never aware of a platform dedicated ZIP drive, so all drives should be interoperable on PDP-11 / VAX SCSI controllers. ... Right?

 

[RSX11M+]

Has anyone used a ZIP 250 ATA drive to make ZIP 100 PDP-11 destined disks? Iomega designed the ZIP 250 drives to be able to read/write ZIP 100 disks.

BTW: Iomega never made a SCSI ZIP 250 drive. (A competitor did)


I purchased two of these ZIP 250 ATAPI drives with the intention of using them in a PC to create ZIP 100 disks for use in ZIP 100 SCSI drives on a PDP-11 QBUS controller.

Delivered today, well shipped - appears to be NOS.


Will report when I've tested this chain. (may take a while)