Interesting reddit blog

[xsoliman]

IamA Ex-employee of a Hard Drive Manufacturing Factory in South East Asia. AMA!

http://www.reddit.com/r/IAmA/comments/1 ... facturing/

[labtech]

Interesting read

[LarrySabo]

Fascinating read. I was surprised to read the importance of regularly defragging to prolong drive life, because mechanical/head/platter wear and tear contributed most to mortality. With defrag scheduled in VISTA+ OSes (1:00 am Wednesdays) by default, I usually tell customers not to bother with it. Also, too frequent defrags just add to wear and tear, I understood. Not sure it's worth changing the default schedule, or not.

[fzabkar]

"althing" states that "I was an Equipment Engineer building and maintaining machines that make Hard Disk Drives". Elsewhere s/he states that s/he has a Mechanical Engineering degree, so s/he is not an electrical engineer and not a programmer. Therefore I'm wondering just how much s/he really knows about the product.

For example, someone asks whether "the temperature thresholds that SMART uses really matter". althing replies "I have no idea. This stuff was handled by the RnD teams. I made the machines that made the drives, remember?"

Elsewhere althing states ...

"Caviar Black is either constant 5400/7200RPM whereas Caviar Blue is variable RPM. This means the disc will spin at variable speeds to suit the read/write speed."

... and ...

"Blue and Green utilize tech to spin the disks at variable speeds."

ISTM that althing has been conned by WD's IntelliPower marketing BS. AIUI, whenever the heads are flying over the platters, the RPM is fixed. If it weren't, then the flying height and read amplitude would vary. Some models do spin their platters at a reduced speed during a particular power management mode, but the heads are parked on the ramp during this time.

The following [expired] URL used to state that "for each drive model, WD may use a different, invariable RPM":
http://www.wdc.com/en/products/greenpow ... nology.asp

This Red spec sheet still makes the same claim:
http://www.wdc.com/wdproducts/library/S ... 771442.pdf

Regardless, the blog was very useful and enlightening.

[Russwinters]

agreed fzabkar; I don't think this person really knows HDD; they are just a productions manager and all about the %'s (failure rates,etc) I don't think their knowledge goes very deep as far as the inner workings of the HDD.

[HaQue]

Well he did say he made the machines that made the hard disks.
Kind of like expecting someone to know how to tune an engine when he really made the casts for the engine blocks.

It was a very interesting thread nonethelkess, and we seldom get insight like this.

I would love to see a similar thread from someone that worked at Spectek, Micron, Sandisk etc

[Russwinters]

Well he did say he made the machines that made the hard disks.
Kind of like expecting someone to know how to tune an engine when he really made the casts for the engine blocks.

It was a very interesting thread nonethelkess, and we seldom get insight like this.

I would love to see a similar thread from someone that worked at Spectek, Micron, Sandisk etc

No doubt; it is still interesting and there is still a lot of useful information we can learn from people who work in that area.

[xsoliman]

I would like to ask such a person how the drives are low level formatted on the production line.

Do they relay on some secret internal firmware command and wait the hours needed
and work out the best adaptives etc needed for each individual drive

Or is there some faster method they use (maybe driving the heads from a test rig or something)

[Keatah]

Disks are low-level formatted with a servo-writer.

[xsoliman]

Ah yes, I remember now

Some good info and pictures here
viewtopic.php?f=1&t=9747&hilit=servo+writer&start=20

[fzabkar]

I'd like to see how the spindle motor and platters are balanced.