OSDev.org

I have always wondered how the different disk controllers are capable of actually reading the proper media locations at high speeds. I know that at electronic speed it is possible and that all of the pieces are properly placed with microscopic precision, but not much more as to satisfy the knowledge I want to get, about the requirements, the methods, the theory for actually calibrating the pieces, properly reading/writing and finding the data in the media, be it magnetic or optical which seem to be the only methods to store data in rotating media.

What books or documents could I read to get good and if possible very practical information about this, since the things I have found so far are too generic? I know that there are several applied fields involved here.

The problem here is that these things are jealously harbored Intelectual Property of the drive manufacturers. You can get high-level information plus a couple of tidbits when you Google for the stuff ("hard drive technology" came up with a couple of interesting links), but you won't find the kind of detailed data you're probably looking for.

Consider: Knowing things that the competition doesn't is what companies like Western Digital, Samsung, Seagate, and IBM are making their profits with. No-one not working for one of these companies will have to bother with the details anyway, since it's all handled by the drive firmware. They won't post such information on the web unless it's outdated and / or common knowledge anyway.

It's all based on high-precision engineering techniques, and electronic control feedback mechanisms. You learn both of these things in university engineering classes. Basically, to answer your questions in detail, you need to take a couple of years of engineering classes.

Step 1. Buy 5" 1.5tb hdd
Step 2. Install in computer
Step 3. Turn on PC
Step 4. ???
Step 5. Data out you requested.


??? will be open to you if you work for a HDD manufacturer,

The way the head mechanism is controlled, it is *almost* linear,

when the HDD is powered on,

The disc is spun up with a stepper motor which is controlled and the speed the disc is spinning is KNOWN
Next the head goes to its extremes and it *precisely* records the voltage to apply to the coil that moves the mech inside the magnets to reach these extremities.

The rest is mainly mathematics, i believe the new high speed data access drives even go so far as to apply inertia and momentum of the head to the algorithms, and i *think*? that the SATA drivers for some companies even handle multiple simultaneous requests and it sorts which one would be the next fastest sector to retrieve

If you grab a pen and paper you can easily imagine in your head how you would create some simple maths to track the head to a track/sector and if you know exactly the speed and orientation of the disc you could work out exactly how long the head will take to cover that sector, which would be the same regardless of the distance the head is from the center, and then you know at which time to start buffering the data the head is picking up/or putting down.

Higher accuracy manufacturing and higher accuracy DAC's are whats making things more reliable and larger these days.

Somebody has compared reading from a modern hard disk to flying a few feet above the ground in a 747, and counting every blade of grass as you pass over it - without making a mistake.