Delicate 8” and 5¼” double-sided floppy disks – how to avoid a ‘bulldozer’ effect.

[SomeGuy]

(OP) Two further points for discussion.

2. Flux reading devices (GW, KF, CW, etc.) may default to reading each track multiple times, with the opportunity to specify the number of reads in the command line. Since loss of oxide may be proportional to the number of reads, reducing the number to 1 or 2 might be better than 3, 4 or 5.

Any comments/discussion on these points would be welcome.

When trying to reconstruct damaged data without the original media, having more revolutions usually increases the chance of being able to piece bits back together.

When a disk ISN'T shedding, a "bad" area may actually start reading OK after a number of revolutions as the heads scrape away dust and gunk.

In practice, I have found 5 revolutions, as the kryoflux defaults, to be a good midground between too many and too few.

 

[Al Kossow]

"the heads scrape away dust and gunk."

do you have evidence of this, as opposed to gunk being scraped off the HEADS?

I've argued back and fourth over the wisdom of moving heads a few tracks either side during the reading process

The one thing you don't want to do, and Chuck has said, is dwelling on one track too long

Ultimately, you want to be doing analog recovery from marginal disks. It's really obvious using that when there is a
dropout, as opposed to shed on the head.

 

[Al Kossow]

I use Siemens FDD-100 and FDD-200 drives myself. Build quality seems to be a bit better than the Shugarts.

I've never had an FDD-200 since they aren't common. Do they use a compliant head just on the top, like Shugart and IBM did?

 

[Al Kossow]

I've used a cleaning method that I talk about in these Winworld threads:

[help needed] bad floppy disk

I'm trying to dump an interesting software (Cullinet Goldengate) but I have problem with one the disks.

forum.winworldpc.com

Help needed - Fortran 3.13

I'm trying to dump floppies of Fortran 3.13 and I need advice about a disk with the surface that deteriorates.

forum.winworldpc.com

Cyclomethicone is my go to substance for any floppy that I think may shed. It's nice to see someone
getting the word out that you really do not want to use alcohol to clean floppy or tape media

 

[Chuck(G)]

I've never had an FDD-200 since they aren't common. Do they use a compliant head just on the top, like Shugart and IBM did?

I'd have to check to make sure, but I believe that they use a similar system to the Shugart bi-compliant setup.

Ever use a Persci 277 drive? Kind of curious about how they treated media.

 

[jonhales]

(OP) Further questions – How tough are Shugart read/write heads?

As mentioned earlier, the main challenge of reading Wabash disks is that the oxide they release is stuck to the heads. Using IPA on a cotton swab may be non-destructive, but it takes a lot of time and effort.

In an earlier post, Chuck(G) has listed chemical products that might work faster/better than IPA.

What else could work to deal with Wabash disk oxide residues?

Shugart SA850 heads
The SA850 manual states: “The proprietary heads are a single element ceramic read/write head. […]The head carriage assembly utilizes a combination flexured/rigid head mounting system. This allows the flexured head to load the media against its rigidly mounted counterpart”.

Shugart SA850 Side 0 head
The lower read/write head of the Shugart SA850 has two components – the means of reading/writing/erasing flux transitions and the white ceramic surrounding the read/write/erase element. The ‘head’ is embedded in an octagonal black plastic part, which might be damaged if certain chemicals were applied.

How resistant to cleaning is the ceramic part?

What would be an appropriate material for removal of the majority of the oxide prior to a final ‘rinse’ to deal with the last traces of contamination of the head with IPA?

Would, for example, a copper scraper be harder or softer than the ceramic? And (if copper is softer) how about a mild steel scraper … and so on? Or a ceramic tool?

I recognise the advice may be that there is no material that would clean off oxide without inflicting damage on the head. That’s OK, if that is the advice.

Shugart SA850 Side 1 head
This head is described as ‘flexured’. I assume this implies ‘easily damaged’. It’s also hard to access, so difficult to know when the oxide has been removed.

What experience do others have about the tendency of the side 1 head to collect oxide? Is the flexibility of the mounting of benefit in reducing the oxide build-up compared with the side 0 head?


‘Cleaning disks’
Long ago, I was advised that excessive use of ‘disk cleaning’ disks is likely to wear down the read-write heads. In the current situation, a small amount of abrasion of the edges of the head might be beneficial (both SA850 drives are known to have had little use since new).

What are your views about this?

Is the risk of damage to the side 1 (upper) head – from cleaning disks - greater than for the side 0 (lower) head?

 

[SomeGuy]

(OP) Further questions – How tough are Shugart read/write heads?

As mentioned earlier, the main challenge of reading Wabash disks is that the oxide they release is stuck to the heads. Using IPA on a cotton swab may be non-destructive, but it takes a lot of time and effort.

In an earlier post, Chuck(G) has listed chemical products that might work faster/better than IPA.

What else could work to deal with Wabash disk oxide residues?

The one time I let a wabash disk fall apart and foul my drive, alcohol didn't want to get everything off. I eventually got it off with Lysol, and some flat strips of paper towel rather than a q-tip.

But the idea is to avoid this in the first place. If you put a disk in the drive an hear a loud screaching, stop imediatly and do something to clean or treat the bad disk. If you do it this way, you should not need more than a cleaning disk.

 

[jonhales]

(OP) What about a dessicant – with or without heat?

Photographers use silica gel to protect film negatives and transparencies from deterioration. It drinks any ambient moisture, which can be driven off by heating the silica gel in an oven. It's re-usable.

Would silica gel have any benefit as a means of drying/hardening the oxide surfaces of Wabash disks (as a preliminary to reading in a drive)? As a single method, or together with heat?

I’m thinking of experiments to investigate this.

Are there any experiences or suggestions?

 

[m_thompson]

I've had good results dealing with fungi and loose residue by removing the cookie and washing it carefully, fully immersed in lukewarm water with a touch of dishwashing liquid. While the cookie is immersed, I rest it on the palm of my hand and gently run a soft cloth around both sides about 4 times. I then removed the mylar from the soapy water, dip it in a bowl of clean water to rise, let the excess water run off, then let it dry while resting on a paper towel. After drying, I insert the cookie into a clean jacket and read.

Tez

I have a PDP-8 bootable 8" diskette with Sykes floppy diskette drivers on it that I would like to image. Unfortunately a mouse family built a giant nest on top of the drive and diskette. The diskette is covered with mouse pee, so I hope your procedure will also work for cleaning up that mess.

 

[Chuck(G)]

(OP) What about a dessicant – with or without heat?

Baking seems to be the best way to accomplish this. If you're wondering about treatments, there is a ton of information at tapeheads.net, There is, I believe, even a patent on baking. Really, take your hints from the audio folks--we're babes in the woods when it comes this. I can think of no worse recovery job than trying to recover audio from an old smelly, wrinkled, acetate-based audio tape. We digital folks have it easy--pretty much all of the stuff you'll run into uses Mylar (note that I didn't say "everything").

 

[Al Kossow]

There is, I believe, even a patent on baking.

ampex 5236790

Silica gel is used to remove low levels of humidity in sealed packages over long periods of time,
which is how it is used for photo preservation.
I've seen it used in commercial low-humidity cabinets where the gel is heated and refreshed
It will have little effect in the short term.
My 1/2 tape baking oven is one of those modified cabinets. I thought it might help but the build-in
hygrometer tells me the chamber dries out to a few percent RH pretty quickly once the heater kicks in.


I bought the brand of hexane chuck recomended in another thread, haven't tried it yet, since I have several bottles of TFE reserved just for head cleaning

 

[Al Kossow]

How resistant to cleaning is the ceramic part?

the ceramic isn't the problem, any organics used anywhere around the ceramic is, glues, cable insulation, etc.

you absolutely, positively don't want to use anything hard around the ceramic head because of the dangers of scratching the surface

 

[Al Kossow]

Ever use a Persci 277 drive?

I've never had one. Drives of that era all seem to use the same design as the one IBM came up with in the 3740
I suppose I could go down the rathole of taking pictures of all of the head assemblies.

The heads I really dislike are the rectangular ones, especially if they are on the bottom.

 

[shirsch]

I've been very happy with my TM848s

I know tandon doesnt exactly scream quality but that model may he the exception...

I have just finished archiving over 200 8" diskettes using a Panasonic JA-751 1/2-height drive and Mitsubishi M2896. Absolutely flawless reliability and not a trace of diskette damage. Not all 8" 1/2 ht. units are crap, but the ones that aren't are getting very hard to find in working condition.

 

[tezza]

I also have a M2896 that I use quite often. Totally reliable!

Tez

 

[Al Kossow]

"Not all 8" 1/2 ht. units are crap"

Qume/YE Data is. I bought a box of NOS ones years ago and gave them all away
to someone that has a Fairlight.
If you dig back in the archives, I'm not the only person to dislike them.
Unfortunately, the Panasonic and Mitsubishi HH 8" drives are not common. I don't
think I've ever found more than a couple Mitsubishi and I've never seen a JA-751

 

[shirsch]

I had never seen nor heard of the JA-751 prior to finding a bunch of NOS units on eBay a few years back. They were not cheap (~150 USD each IIRC), but I picked up three of them. Just gorgeous pieces. The M2896s (I have 2) were $10 bargain box items at a 1998 hamfest. I long for the good old days...

 

[shirsch]

I also have a M2896 that I use quite often. Totally reliable!

Tez

Just keep an eye on the rubber "slap absorber" bits glued to the nylon head-load cyclinders. They degenerated into goo on both my units. Easy enough to fix, if a bit tedious.

 

[maxtherabbit]

I had never seen nor heard of the JA-751 prior to finding a bunch of NOS units on eBay a few years back. They were not cheap (~150 USD each IIRC), but I picked up three of them. Just gorgeous pieces. The M2896s (I have 2) were $10 bargain box items at a 1998 hamfest. I long for the good old days...

M2896s are selling on ebay now (maybe twice a year tops will one even come up) for over $300

 

[Dwight Elvey]

I had a set of floppies that had what looked like some type of fungus. It was kind of like blotches on the disk. I later found it was the adhesive used to hold the liners in the envelope. This had bleed through the liners to the disk. My guess is they'd been stored in an attic.
I was able to read the disk after carefully cleaning the surfaces with rubbing alcohol and drying ( of course a new envelope as well ).
I did follow chucks instructions on cooking.