TR Forums

starts at 2:00

enjoy

https://youtu.be/J9P4UadRdNA

I wouldn't recommend attempting this in anything but a clean room and only with *REALLY* old drives. Modern drives aren't anywhere near as tolerant to contamination or as easy to disassemble.

That was really cool.

They are using a lammelor flow air unit, so it is basically a clean room. Not as good as a clean room, but more than enough to prevent contamination on the platters.

lammelor --> laminar

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

Epik fail lol, dont try to spell while late night posting...

I read that as laminar and didn't even notice until JBI pointed it out.

Redocbew wrote:

I read that as laminar and didn't even notice until JBI pointed it out.

So did I. Dare I look up whatever "lammelor" is?

Starfalcon wrote:

Epik fail lol, dont try to spell while late night posting...

They had a shot where they zoomed into the machine's label, clearly spelling "laminar".

So the scotch tape must have cost $10 out of the $1000s that are charged to a customer.

Psh, you and your spelling and rules.

Wow... do they really tend to use magnetized screwdrivers for that though?? If there was one part of a computer I would have thought would be most at risk or susceptible to a magnetic screwdriver it would be platters on a modern hard drive.

Kougar wrote:

Wow... do they really tend to use magnetized screwdrivers for that though?? If there was one part of a computer I would have thought would be most at risk or susceptible to a magnetic screwdriver it would be platters on a modern hard drive.

Modern HDD platters pack the bits so tightly together that the magnetic coating must have very high coercivity (resistance to changes in magnetization); otherwise adjacent bits of opposite polarity would erase each other! It takes a pretty intense field to change the magnetization on a modern HDD platter; the write heads are designed to create a field comparable to that of the strongest available rare-earth permanent magnets (but obviously concentrated on a microscopically tiny spot). The magnetic field at the tip of a magnetic screwdriver is going to be orders of magnitude less.

Unless you're touching the tip of the screwdriver directly to the platter (and probably not even then...), I'd say the risk of flipping bits is basically zilch. And if you're touching the platter with the screwdriver you've got bigger problems anyway.

just brew it! wrote:

Unless you're touching the tip of the screwdriver directly to the platter (and probably not even then...), I'd say the risk of flipping bits is basically zilch. And if you're touching the platter with the screwdriver you've got bigger problems anyway.

Aye, that makes sense. And I forgot about the super-powerful rare-earth magnets sitting an inch away anyway, whoops.

So since that rules out the hard drive, where did the whole fear of magnetic screwdrivers in a PC come from anyway? Seems like it's just another myth/urban legend now.

Kougar wrote:

just brew it! wrote:

Unless you're touching the tip of the screwdriver directly to the platter (and probably not even then...), I'd say the risk of flipping bits is basically zilch. And if you're touching the platter with the screwdriver you've got bigger problems anyway.

Aye, that makes sense. And I forgot about the super-powerful rare-earth magnets sitting an inch away anyway, whoops.

So since that rules out the hard drive, where did the whole fear of magnetic screwdrivers in a PC come from anyway? Seems like it's just another myth/urban legend now.

Common household magnets were capable of erasing floppy disks and magnetizing the shadow mask in color CRT monitors (causing shifts in image color). That's where the "keep all magnets away from the PC" rule came from. With floppies and CRTs both going the way of the dodo, the risk from magnets has basically gone away.

The everyday item most susceptible to magnetic damage these days are the credit/debit cards you carry in your wallet. Once all B&M merchants have switched over to chip readers that'll be history too.

Kougar wrote:

So since that rules out the hard drive, where did the whole fear of magnetic screwdrivers in a PC come from anyway? Seems like it's just another myth/urban legend now.

Probably comes from the very early days of PC 3.5" drives when bit density was so low that the IBM PC XT shipped with a 10 megabyte Seagate ST-412 HD. Bit density (and thus required coercivity) was so low that external magnets may very well have been able to affect the platters. As we have all seen here, once something becomes a "rule" in the computer hardware field, it tends to remain a rule long past its sell-by date.

Even with the early PC HDDs I doubt a magnetic tip screwdriver would've been able to do anything to a drive that wasn't opened up. You'd probably need to do something on the order of popping the lid of the drive and setting a fridge magnet on the platter. (And with the low bit densities and greater head flying heights back then, you might've even been able to get away with doing that and not cause enough platter damage to result in a head crash when the drive spun back up...)

I have to wonder how much random luck is involved in a platter swap like that. If you've got a platter (or worse a stack of platters) that was not formatted on the drive controller in use, the ability to find track 0 would seem to involve a great deal of chance. It might work, it might not. Perhaps the embedded servo code can redirect on the fly, but you still have to be able to line up on at least one track reliably.

My disk-fu is at least 15 years out of date at this point (stopped being concerned with the details sometime after GMR heads became common), so perhaps the mating process wherein the heads ask "where the f am I and who the f are you?" has changed significantly.

Given the extreme track densities these days (most of the capacity increase over the past couple of decades has come from packing tracks closer together), I had assumed that there was already something pretty hairy going on with regards to track positioning. IIRC about a decade ago HDD vendors moved to a 2-stage actuator (traditional rotary voice coil for coarse positioning, with a piezoelectric second stage actuators mounted on the head arms for fine control) because the tracks became too fine for a purely electro-mechanical solution.

FWIW a modern flagship capacity drive has a track pitch of just a few 10s of nm... positioning the heads with that kind of accuracy seems like black magic to me.