Stiction marks appear as tiny circles under microscope?

[CleanestRoom]

Hi there

Was doing some dead drive autopsying on a:

Seagate Barricuda 2000GB
ST2000DM001
9YN164-302

Had a massive amount of trails, and stiction marks on every platter, for the first time i got to try out my new microscope on some stiction marks, then i noticed something freaky.

The stiction mark appeared to be a series or tiny circles at max (90X) zoom.

Circles? Why circles? Can anyone explain this strange phenomena to me?

Attachments

Here's the best looking image i found.

[Dmitri]

The stiction mark appeared to be a series or tiny circles at max (90X) zoom.

Circles? Why circles? Can anyone explain this strange phenomena to me?

This drive uses patterned media (aka BPM).

[CleanestRoom]

Fascinating! So each circle is a bit.

https://en.wikipedia.org/wiki/Patterned_media

http://www.toshiba.co.jp/rdc/rd/fields/11_e09_e.htm

Thanks for the input. If anyone else is wondering what microscope i used, it was an Amscope:

http://www.amscope.com/microscopes/3-5x ... amera.html

[Leo]

And this pattern on the posted picture is not a stiction mark.

[CleanestRoom]

Enlighten me. What do we call it?

[HaQue]

I would have thought an individual bit would be a LOT smaller than this.. but I have no experience, research or knowledge to actually back this up. I notice on the link you posted they used a SEM to see the bits.

[CleanestRoom]

I would have thought an individual bit would be a LOT smaller than this.. but I have no experience, research or knowledge to actually back this up. I notice on the link you posted they used a SEM to see the bits.

I also imagined they would be a lot smaller, strange they showed up on only 90x zoom.

After fiddling around with the scopes software i was able to get pixel measurements, at max (90x) zoom.

1mm at MAX zoom = ~750px

1 circle at MAX zoom = ~23px

23 / 750 * 100 = ~0.0306mm
or ~30.66 μm (micrometers)
or ~30666.6 nm (nanometres)

So yes these circles are about 3,000 times larger than the '10nm' quoted from that toshiba link. (If my math isn't off.)

Attachments

1mm compared to circle. (Circle is zoomed.)

[Dmitri]

If anyone else is wondering what microscope i used, it was an Amscope:

Basically duplicating my PM to CleanestRoom here, but anyway:

Yes, these dots aren't magnetic grain for sure.
I probably should sleep more, but I asked CleanestRoom about the microscope as I thought he got his hands on some sexy equipment and decided to (ab)use it properly :)

This is an amateur judgement, but I believe optical microscopes can't produce such a detailed picture of the grain on a modern platter due to their diffractional limit.

[jermy]

but I believe optical microscopes can't produce such a detailed picture...

even with digital camera connected to it ?

Thanks for the input. If anyone else is wondering what microscope i used, it was an Amscope:

http://www.amscope.com/microscopes/3-5x ... amera.html

details from the link above:

3MP digital camera captures still images, streams live videos and is compatible with Windows XP/Vista/7/8, Mac OS X, and Linux, as well as user-friendly advanced software for Windows offers Stitching, EDF, video recording "and measurement functions"

[fzabkar]

Using just the measurable parameters (max transfer rate, outside diameter, RPM), we can arrive at an upper limit for the size of a "bit".

(210 MB/s) / (7200 per minute) / (3.5 inch x pi) = 50 bits per um = 20 nm per bit

http://www.google.com/search?q=210+MB%2 ... its+per+um

Obviously the size would be lower if we were to account for the "overhead" bits.

According to the product manual (http://www.seagate.com/staticfiles/supp ... 686584.pdf), the maximum recording density is 1807kFCI (kilo flux changes per inch).

That works out to 13.6 nm per flux change.

[CleanestRoom]

details from the link above:

3MP digital camera captures still images, streams live videos and is compatible with Windows XP/Vista/7/8, Mac OS X, and Linux, as well as user-friendly advanced software for Windows offers Stitching, EDF, video recording "and measurement functions"

Software only measures pixels, there's no electronic reading of the scopes zoom so calculations must be done manually as I've done.

Hypothetical. I'm a billionaire, and i want a SEM that can read to less than 10nm, how much might that cost?

[HaQue]

http://www.technicalsalessolutions.com/items.php?CID=3

There is one here for 8 grand...
http://www.labx.com/item/hitachi-s-2400-scanning-electron-microscope/624867

of course I haven't looked at specs, but it would do me

not insanely expensive, some bigger DR labs could probably afford one