Draft. Draft — assembled from the working notes by `tools/build_book.py`. The source document is the authority; this page is a copy for reading, and it is regenerated from that source.

Status: Solved. As a product analysis: the OCC1 is Australian Computer & Telecommunications’ 1982 Osborne 1 hard disk, and the mass-storage decision went the other way — emulate the Adaptec ACB-4000 inside the MD-10, not the ACT (docs/harddisk-decision.md). The ACT’s hardware interface below is read out of its own BIOS image. What remains is archival only: no ACT board photograph or schematic has surfaced (#8).

What this is. The ACT unit is a hard disk for the Osborne 1 that attaches through the Z80 socket, and it is one of two unrelated Osborne 1 hard-disk products in the archive — the other is the Media Distributing / Adaptec MD-10 (Two products, not one, below). The archive holds its driver disk and nothing else, so the ACT is documented here as prior art: it validates the Z80-interposer concept on this exact machine, and it is not the controller the project emulates.

Source: research/harddisk/OCC1_HARDDISK.IMD (Osborne 1 SSDD, FM 10x256). Extracted to research/harddisk/extracted/ via disk-analyse + cpmtools (osb1sssd).

What this is

The Osborne 1 hard disk product from Australian Computer & Telecommunications (ACT), (c) 1982 (string in BOOTHd.COM). This is the “winchester style” interface — and, significantly for us, it is one of the known Z80-interposer designs: the host-side hardware plugged into the Osborne’s Z80 socket.

Two products, not one

The archive drop (#71) added a second, unrelated Osborne 1 hard-disk product, so “the OCC1 hard disk” is a family, not a single thing:

  ACT (this doc) Media Distributing / Adaptec (the MD-10)
Source research/harddisk/OCC1_HARDDISK.IMD research/adaptec/*.img
Vendor string AUSTRALIAN COMPUTER & TELECOMMUNICATIONS COPYRIGHT (C) 1982 Copyright (c) September 1983, Media Distributing
Versions Version 31.05.82, Version 9-02-81 HARD version 1.93, PORTCHNG version 1.92, PREP version 1.93
Controller its own; prompts Enter disk controller address ( Adaptec ACB-4000; prompts Enter starting port address in hex / must be on a 4 port boundary
BIOS HARDBIOS.HEX, overlaid by LOADBIOS.COM HARDBIOS.SPR, HARDBDOS.SPR, HARDCCP.SPR, HD00BIOS.SPR, HD04BIOS.SPR
Other utilities MOVCPM10/5/F, DISKEDIT, DISKTEST, RESTORE, SAVEFILE MD10.COM, MD20.COM, PREP.COM, DRIVETBL.DAT, PORTCHNG.COM

The second product is identified: it is the Mega-byte MD-10. Mega-byte Inc.’s advertisement in The Portable Companion #5 (February/March 1983, PDF p. 25; research/technical-articles/TPC-Issue05-1983-pp25-mega-byte-md10-hard-disk-ad.pdf) sells the MD-10 — 10 MB formatted for $2295, “20 and 44 megabyte Winchesters also available” — and its footnote reads “MD-10 is a registered trademark of Media Distributing.” The disk in research/adaptec/ is copyright Media Distributing (September 1983) and ships MD10.COM/MD20.COM, matching the 10 and 20 MB models, so the “Adaptec” disk is MD-10 software. Two details from the ad bear on the interface question below: the MD-10 “does not tie up existing OSBORNE ports” and is “significantly faster than other winchester subsystems which use the IEEE port” — so it is not an IEEE-488 device, consistent with a controller that claims its own addresses (hence PORTCHNG.COM) rather than borrowing the serial/video ports. Its own parts list — “drive controller, case, power supply, cabling, Z-80 interface” — points at the same class of host attachment as the ACT unit: a Z80-socket or dedicated-port design, not the IEEE bus.

Both are port-relocatable controllers: the ACT disk asks for a “disk controller address”, the Adaptec disk for a “starting port address … on a 4 port boundary”, and it ships PORTCHNG.COM to change it. That is consistent with the rest of the O1, whose I/O decode only looks at the low address lines.

That is also what makes the port claim resolvable: 0x2A00 and 0x2C01 are the console ACIA and the video PIA on this machine (docs/o1-memory-io-map.md §3), so the ACT adapter is mapped elsewhere — the IEEE-488 PIA at 2900–2903, read out of the BIOS below.

So there are three routes to more storage than 2 × 182 KB — the two hard disks here, and Drive C (docs/drive-c-protocol.md) — and only Drive C is buildable without unobtainable hardware (#84).

Files

File Role
HARDBIOS.HEX Intel HEX of a full replacement BIOS, loads at 0xDF00-0xEB00 (17-jump BIOS entry table at 0xDF00)
LOADBIOS.COM Overlays HARDBIOS.HEX onto the running BIOS (“Your BIOS has been overlayed”; validates 17 JMPs, checksums, warm-boot vector)
BOOTHd.COM Hard-disk boot (“ERROR DURING BOOTING TO HARD DISK”)
MOVCPM10/5/F.COM Stock Digital Research MOVCPM.COM builds — three relocations of CP/M under the new BIOS, not 10 MB / 5 MB volume sizes
DISKEDIT.COM, DISKTEST.COM Winchester utilities
RESTORE.COM, SAVEFILE.COM Backup/restore to floppy
AUTOST.COM, SYSGEN.COM, PIP.COM, XDIR.COM Standard

Hardware interface (from HARDBIOS.HEX disassembly)

Read by disassembling HARDBIOS.HEX (z80dasm, org 0xDF00, 0xDF00–0xEB00). The interface is not the 0x2A00/0x2C01 pair those addresses would suggest — they are the console ACIA and the video PIA (docs/o1-memory-io-map.md §3) — and the image drives exactly one device. docs/harddisk-decision.md §3 has the same reading in the decision’s own words.

  • Same Osborne bank discipline as Drive C: OUT (0x00) = bank 2 (IO), OUT (0x01) = bank 1 (RAM), DI/EI bracketed (subroutines at 0xEA6A / 0xEA75 wrap all IO accesses).
  • The disk transport is the IEEE-488 PIA at 2900–2903, bit-banged: set 3,(02902h) / res 3,(02902h) is a PB3 strobe around ld a,(02900h) (read) and ld (02900h),a (write), with cpl for active-low. It is the only device routine in the image, reached from the disk path READ (E32A) → sub_e448h → E475 → sub_e520h → sub_e58fh (E58F).
  • 0x2A00/0x2C01 are not the disk. 0x2A00/0x2A01 are the console ACIA (0x2A00 status read / control write, 0x2A01 RX/TX — the console helpers sub_e992h/sub_e99bh test bit 1 = TDRE, bit 0 = RDRF) and 0x2C01 is the video PIA (docs/o1-memory-io-map.md §3). The image contains no floppy-controller references at all: it is a hard-disk-only BIOS.
  • The “indirect OUT (C),A, target TBD” is resolved: C comes from [0xEF08], which the bank helpers set to 0 or 1 — it is the bank restore on exit, not a disk port.
  • The BIOS preserves a full register environment (pushes IX/IY/HL/DE/ BC/AF, switches to a private stack at 0xEFC1) around hard-disk calls — it hooks deep into the 17-entry BIOS jump table as a full BIOS replacement, not just a disk-driver wedge.

Comparison with Drive C

  Drive C (RAM disk) ACT hard disk
Physical interface IEEE-488 port (6821 PIA bit-bang @ 0x2900-0x2903) IEEE-488 PIA @ 0x2900-0x2903, bit-banged (PB3 strobe)
Host code Driver wedge (DCL/DFD.SPR) Full BIOS replacement @ 0xDF00
Boot QUICKPAC autoload from floppy BOOTHd.COM + SYSGEN to winchester

Implications for the interposer project

  1. The ACT product validates the whole interposer concept on this exact machine — a Z80-socket interposer was a shipping commercial Osborne 1 peripheral in 1982.
  2. Its transport is the same PIA bit-bang as Drive C (2900–2903), so an ACT personality would reuse that machinery — but its command protocol is undocumented and would have to be reverse-engineered from HARDBIOS.HEX (docs/harddisk-decision.md §3).
  3. The mass-storage decision is made: emulate the Adaptec ACB-4000 (the controller inside the Media Distributing MD-10), not the ACT. See docs/harddisk-decision.md for the reasoning and the plan.

Open questions

The ACT-specific questions are closed or superseded — the ACT is not being emulated (docs/harddisk-decision.md §3):

  1. Which ACT product exactly? Superseded. MOVCPM10/5/F.COM are stock Digital Research MOVCPM.COM builds, not volume sizes.
  2. Command set at 0x2A00 — resolved as not the disk. The transport is the 2900–2903 PIA and its protocol is undocumented.
  3. The OUT (C) target port at 0xEF08 — resolved: the bank restore.
  4. Whether any ACT hardware photos/schematics survive online — still unattested, and now only of archival interest.
  5. Re-derive the port claim — done (above; harddisk-decision.md §3).
  6. The MD-10 / ACB-4000’s remaining questions — the DPB, the host-adapter bit map, the drive letters, “36 user areas” — are in docs/harddisk-decision.md §10.