Difference between revisions of "RS03/04 disk drive"
(Mention swapping use) |
m (+cat) |
||
Line 41: | Line 41: | ||
{{semi-stub}} | {{semi-stub}} | ||
+ | [[Category: MASSBUS Devices]] | ||
[[Category: DEC Disk Drives]] | [[Category: DEC Disk Drives]] |
Revision as of 18:33, 7 December 2018
RS03/RS04 | |
Manufacturer: | Digital Equipment Corporation |
---|---|
Drive Controller(s): | RH11 (UNIBUS), RH70, RH10 |
Capacity: | 512KB (RS03), 1024KB (RS04) |
Transfer Rate: | 2 μsec/byte (RS03), μ1 sec/byte (RS04) |
Average Access Time: | 8.5 msec (60Hz), 10.2 msec (50Hz) |
Revolutions per Minute: | 3600 (60Hz) |
Media: | 16" diameter nickel-cobalt plated surface |
Total Surfaces: | 1 (RS03), 2 (RS04) |
Tracks per Surface: | 64 (plus 8 spares) |
Sectors per Track: | 64 |
Words per Sector: | 128 bytes (RS03), 256 bytes (RS04) |
Density: | xxx bpi (maximum) |
Recording Method: | MFM |
Physical Size: | 15-3/4" (high), 19" (width), 26-1/4" (depth) |
Weight: | 120 lbs |
Power Consumption: | 13A (starting), 6A (operating) |
The RS03 and RS04 disk drives were MASSBUS fixed-platter fixed-head disk drives. Their basic format was 18-bit words (for use in PDP-10 and PDP-15 machines), but they could be used in PDP-11s, where they simply ignored two bits.
They attach to any MASSBUS controller, e.g. RH11 (UNIBUS), RH70, or RH10. Up to 8 devices total can be attached to any MASSBUS controller.
The RS04 doubled the capacity (and bit rate) of the RS03 by recording on both sides of the platter; alternate bits in each word came from alternate sides.
They could store up to 512KB of data in 128 byte blocks (RS03), or 1024KB of data in 256 byte blocks (RS04). They used 64 fixed read/write heads per side for data (8 spares were provided), and 2 heads to read timing/address tracks (a main and a backup).
The second timing track did not provide reliability/redundancy; the two tracks were not recorded synchonously. (This seeming oversight may be explained from DEC's apparent intent that this device be used for swapping, where being unable to retrieve data would not be an issue. Indeed, the RS0x diagnostics scribbled over the disk - which caused an issue the first time they were run at Bell Labs, where UNIX used one as a root drive!) Switch-over from one to another was manual, by inverting a connector.
The drive had the ability to write-protect tracks 0-N ('N' set in switches).