I have nothing to say; I just love this picture. Someday I will have one of these old-school hard drives on my wall and I will be a happy man. (Thanks, O2!)
Look – New Hard Drive [Gizmodo]
Update: Daniel V. Klein offers up some insight on this drive (and a hilarious anecdote about moving men and inertial precession) after the jump.
The picture is of a fixed-head disk, very similar to a Borroughs unit I had the pleasure of disassembling (in 1975) after a catastrophic head crash (I got authorization from Gordon Bell himself to do it). It took me 3 days to whittle it down to nuts and bolts, and the platter weighed 18 pounds. The hub upon which the platter was mounted was phosphor bronze, and weighed an additional 17 pounds. So imagine the inertia of 35 pounds spinning at 3600 RPM. It had electric brakes, because if you just switched off the power, it would spin for a loooong time. There is an (apocryphal) story of movers just hitting the circuit breaker (not the off switch that engaged the brakes), and after waiting the requisite 5 minutes for spindown, loaded the drive into a truck. All the moves and hallways were right angles, of course. Since brakes had not been engaged, it was still spinning at 2000 RPM or so by the time it was loaded. When the truck turned a corner, the drive precessed right out through the side of the truck.
It held a few megabytes at most, if I recall correctly (a similar unit was used as a swap disk on the PDP-10, so it would have held 256K or so). I can get exact numbers from co-workers, if you care.
“Fixed head” is an interesting technology – you can see the technician lowering one bank of heads, and this appears to be a dual-platter unit, so it has four banks of heads. Each head array (there are 13 array’s per bank in this picture, hand wound magnetics, of course) accessed between 8-12 tracks. Each track therefore had its own head, so the seek time was zero – you just electronically switched heads.
What you can’t see here is the plumbing. When the drive was spun up, the heads were designed to be floating high above the disk – too far to work. An air pump activated a piston which drove each of the gimballed head arrays towards the disk.
I still have the platter, both head banks (one in an “exploded” state so folks can see how it worked), the makers logo, and some drive electronics, along with the drive motor and air pump, and a box of nuts and bolts that *still* comes in handy.
