Quantum GoDrive Failure

[techknight]

Well guys, All quantum drives are AFFECTED by this problem. Not 1 will live scott-free.

So, unless you have a special platter compression tool, you cannot fix this problem. The problem is the head crash post thats made of rubber that decides it wants to meltificate. The GoDrives, this is under the platter. Desktop drives its a mixed bag. sometimes its on the magnet and can be fixed. sometimes its under the platters and cannot be fixed.

The reason you cant take the platters out, is because if any of the platters rotate freely from each other, the cylinders do not match up any longer and the low-level formatting will be rubbish. You have to have a compression tool to remove the platters. A makeshift piston ring compressor can work, given some ingenuity. This way, the platters can be unscrewed and remove without slipping from one-another. This will allow for a successful repair. Also you must remove the head actuator first, which requires care of course. you have to fabricate an insert to go between the head arms as your removing the heads, because the heads cannot touch each other. if the heads touch, they are instantly destroyed and will not seek data.

Here i show pictures of the failed GoDrive: (the actuator picked up the melted rubber and plastered it all over the platter, instantly ruining the drive.)

https://www.dropbox.com/s/yilshgmpvnyxhd8/20130915_162032.jpg

https://www.dropbox.com/s/oegv7egsetmqj5g/20130915_162038.jpg

https://www.dropbox.com/s/8xkmp50mj9rxty0/20130915_162052.jpg

Unless you intend to do the repairs, Please retire all quantum drives immediately out of active use

If you can repair this, you can probably fabricate a new rubber peice to stick on there. But youll need a clean "box" which is obviously easier to fabricate than having a clean room. Then youll need the right platter tools. If its a single platter, you wont have any problems. If its multi-platter you need the tool.

I knew about this problem for awhile, but i decided to discuss it again with pictures this time.

 

[Paralel]

If the relative cylinder order between platters is disturbed, can't you just low-level format the drive?

(Excuse if this seems like a stupid question, but cylinder alignment, relative to low-level formatting, is something of which I know nothing)

 

[techknight]

Nope, because the platters contain servo tracks. Those servo tracks is what tells the controller where the head is at any point in time as it floats on the platter. I assume when they low level them, they use a special machine that plugs into the headstack/spindle and lays out all the tracks first using fixed positioning information and actuator drive signal, then they put the controller board on the drive. It is a servo, and all servos require a reference and/or feedback.

Once you disturb the servo tracks, the head will get confused where its at and it will slam and jump all over the place. worse than the click of death. Stepper motor drives dont have this issue because well, its angular position is constant. 1 step is going to move a fixed degree angle, never more or less.

Same thing with a CD player. There is a positioning track on the disc, so the laser knows how to focus and track itself along the disk, and knows exactly where it is at all times. Without this, the laser would move and bob everywhere eventually giving up. (happens when lasers start to get weak, they cant see this track as easily). Same with CD-Rs, there is still a spiral track recorded even though it isnt "burned" with information, and the laser records the information into the track as well as tracking "to" it. otherwise the laser would never know where it is, and it couldnt focus or track. a servo track has to be present regardless.

 

[James1095]

I saw a video a while back from some guys that were doing hard drive data recovery. I forget what they used for the clamp but it was something fairly simple they had made, I would never have thought it would be possible but it worked. Of course this was just to recover data from the drive, but if you can get it to work at all, I think there is a reasonable chance it will continue working.

A clean box is not strictly necessary. You want an area that is as clean and dust free as possible, and you want to have the drive open for as little time as possible, but careful use of a can of air duster should be sufficient to clean out any remaining dust just before installing the cover. Once closed up, there is an internal filter in the drive that will pick up just about anything remaining. If you wanted to make one, a cardboard box, plexiglass, packing tape, long rubber gloves, some computer fans and a HEPA filter for a vacuum cleaner could make for an ugly but functional one.

Where you get into trouble is running a drive with the cover off. The lack of a cover disturbs the airflow. Stray gusts interfere with the cushion of air the heads float on and cause them to crash. Particles of dust fly under them and do the same thing. I have opened a few drives before to fix stuck actuators and as long as I resisted the temptation to power it up open, some of them worked for years after that. The couple times I ran a working drive with the cover off, they failed within 20 minutes or so.

 

[techknight]

I am thinking a clear tote or something of that, with holes cut in the side for the gloves taped/glued in place. Then of course the fan with HEPA. Thats the best and easiest bet.

The point is, we need to come up with a way to do this without destroying the drive in the process. That way, we can fix this issue before more drives succumb to it.

 

[Paralel]

What is the explanation for the rubber failing like this?

 

[techknight]

Crappy chemistry? Im not sure. Belts in some cassette decks did the same exact thing.

 

[Paralel]

I've had belts in old floppy drives do this as well.

I'm thinking you're right, it's a failure of the chemistry used at that time. The rubber wasn't stable, just long-term metastable. A common problem. No amount of testing would predict 30 year stability, so they had no way of knowing the composition would fail. Just like the bromine in the fire retardants that yellows cases. There is no way they could have predicted at that point it would happen given enough time. No one ever bothers to model the long term stability of something with a much shorter expected useful lifespan. It does create a really frustrating set of problems for preservationists and vintage collectors.

Unfortunately, I think it's a problem we're going to be contending with more and more as we move into yet more advanced materials. Apparently complex materials of the modern age don't hold up quite as well to the test of time as tried and true materials of previous centuries. Until we can figure how to deal with it there is going to be some pretty significant gaps in artifacts and collectables for this time period as a result in the future.

 

[James1095]

Environmental factors come into play too. Ozone in particular tends to break down rubber, and trace amounts of various solvents and other chemicals can attack it as well. Elevated temperatures increase the speed of any reaction, including the deterioration.

I suspect that the long term stability of the materials was frequently known, but if you are building something with an expected useful life of ~5 years, and a material spec'd to last longer than that costs more, you can guess what will happen. When you are designing something to be mass produced, you usually select the cheapest materials that are deemed good enough that a large majority of the product is projected to last well beyond the warranty period.

All of these drives are FAR beyond their intended lifespan. It isn't that they were designed to fail, but the designers assumed that most of them would be scrapped long before now due to obsolescence and I would wager that to be the case. If you buy a brand new hard drive today, you probably aren't too worried about whether it will still be working in 20 years.

 

[techknight]

And your right, I have an Akai GXC-39D that has its original belts, in perfect condition. But yet ive seen newer decks with belts so melted it gummed and jammed the entire mechanism.

 

[James1095]

I've seen some weird interactions between materials too. A friend of mine had an old radio that was stored with the power cord wrapped around it and the rubber insulation reacted with the plastic housing somehow and they melted into each other. I had a roll of that thin yellow tape that you find covering the windings in a lot of ferrite transformers and a roll of a different kind of tape that was left sitting on it turned the edges of the yellow tape into rubbery goo in the places they were touching.

Worse than belts, sometimes rubber pinch rollers break down and turn into tar-like muck. That was a fairly common problem with 8-track tapes too, the rubber roller being built into the tape cartridge. I'd be curious to know just what leads to this, whether it's inevitable with some materials, an interaction with vapors that outgas from nearby materials, or environmental factors.

 

[Gorgonops]

Another case of material breakdown is the "non-slip" coatings you'd find on quite a few portable dinguses from the mid 1990's.

.

 

[jwse30]

Another hobby of mine is collecting and playing with Lionel trains. In the mid 90's they made a few flatcars that had automobiles on top of them. The rubber (?) tires melted into the plastic car body. Shortly after that was discovered they started placing a thin piece of cardboard between the two.

I've got a Go drive that just recently gave up the ghost. If anyone wants to practice this repair (I don't understand enough of the words, let alone the repair itself :-/ ) I'll mail it to them after I get the new one in. If you're outside the continental US, I'll have to see how bad the shipping is before I make good on this...

J White

 

[insaneboy]

This definitely could explain the sudden failure of the go drive from the PowerBook 150.

 

[techknight]

it is the reason for the failure... But once its failed, its too late. Gotta fix it before it fails.