Acorn Risc PC 600 Troubleshooting Guide

This guide gives diagnosis for the Acorn Risc PC 600 (ARM610, board 0197,000). It uses the Power On Self Test (POST), the on-board diagnostics and the parts/electrical data in the Acorn Risc PC Technical Reference Manual, which also contains the full circuit diagrams (Appendix C).

Acorn Risc PC 600. Source: Wikimedia Commons.

Mains warning

The Risc PC has a fan-cooled mains PSU in the base slice. Disconnect the mains lead before removing the cover, and never work on the supply while it is plugged in. The PSU primary side and its reservoir and mains-suppression capacitors can hold a lethal charge after the machine is unplugged — a shock can be fatal. The module can be opened and recapped for a restoration, but only if you are competent with mains switch-mode supplies: discharge the reservoir capacitor and confirm 0 V before touching the primary side.

Power On Self Test (POST)

The Risc PC always runs a POST immediately after power-on. The progress is shown by the screen colour, and a failure is signalled both on screen (red) and by the floppy-disc LED flashing an 8-digit hexadecimal fault code.

Normal sequence (screen colour):

1. Purple — testing has started; brief ROM and RAM tests run, then IOMD and VIDC20 are initialised and tested. This phase is quick. A crash or halt here, with no further activity, indicates a major fault, most likely in the I/O system.
2. The screen mode is set from the video ID[0] pin: VGA if a VGA/SVGA/Multiscan monitor pulls ID low, otherwise a 15 kHz TV-style mode. If no stable display appears, suspect the monitor connection, a video fault, or a fundamental fault preventing the test software from running.
3. Blue — a more extensive memory test, then brief video and sound tests.
4. Purple again — the ARM processor ID is read (needs good RAM).
5. Green (pass) or Red (fail). On a pass the green screen is usually not seen, because RISC OS starts immediately and shows a black screen with the memory-size message. On a fail, the screen stays red and the floppy LED flashes the fault code.

Reading the LED code: the 8-digit hex code is flashed as 8 groups of 4 flashes; a long flash = 1, a short flash = 0. For example a ROM failure on an ARM610 machine is code 00000219.

POST fault bits (only bits 9–31 indicate faults)

Code	Meaning	First-look
00000200	ROM failed checksum	Reseat/replace the RISC OS ROMs; check ROM data/address lines
00001000	IOC register test failed	IOMD or its bus; check seating and tracks
00004000	VIDC (Virq interrupt) timing failed	VIDC20 or its clock/interrupt path
00008000	Sound (Sirq interrupt) timing failed	VIDC20 sound / sound DMA path
00010000	CMOS unreadable	I²C path to the CMOS/RTC; battery/clock circuitry
00020000	RAM control line failure	Classic leaked-electrolyte symptom — inspect/recap and repair corroded RAM-control tracks; reseat SIMMs
00040000	Long RAM test failure	Faulty SIMM; substitute

POST status bits (informational, bits 0–7)

Code	Meaning
00000001	Self-test due to power-on
00000040	Integrated I/O controller fitted
00000080	VRAM present
00000100	CMOS RAM checksum error

A POST interface box (plugged into the parallel/printer port) shows each test and a PASS/FAIL bitmap on an LCD, running the tests slowly so the screen-colour phases are clearly visible.

First: power

If the machine is completely dead:

1. Disconnect the mains lead and check the mains-plug fuse.
2. With the lid open (mains reconnected, take care), confirm the PSU rails and that the PSU fan runs. A dead rail points to the PSU module.
3. Check the on-board fuses FS1 (2 A) and FS2 (800 mA) for continuity — a subsystem dead with the main rails present is often a blown on-board fuse.

Always suspect leaked electrolyte

On a Risc PC, intermittent and progressive faults are frequently leaked surface-mount electrolytic capacitors rather than a failed IC. The POST RAM control line failure (00020000) code in particular is the classic result of leaked electrolyte corroding the RAM-control tracks. Inspect the board under magnification for crusty residue and discoloured pads around the surface-mount electrolytics before deeper diagnosis; recapping and cleaning often clears otherwise-baffling faults — see Acorn Risc PC 600 Capacitor Guide.

Symptom table

Symptom	Probable cause	Action
Completely dead	Mains fuse, PSU, or on-board fuse	Check plug fuse, PSU rails/fan, FS1/FS2
Stays on a purple screen / crashes early in POST	Major I/O-system fault (IOMD) or RAM	Inspect for leaked electrolyte; reseat SIMMs; check IOMD
Red screen, floppy LED flashing	POST fault — read the code	Decode the 8-hex LED code (table above)
No stable display, machine otherwise alive	Monitor/cable, video ID pin, or VIDC20	Try a known monitor; check the VGA connector; verify VIDC20
Random crashes, RAM errors (code 00020000/00040000)	Leaked electrolyte on RAM control lines, or a bad SIMM	Recap and clean; reseat/substitute SIMMs
Lost configuration / wrong clock	Flat or leaked NiMH battery (BT1)	Clean corrosion, repair tracks, replace cell
No / distorted sound (code 00008000)	VIDC20 sound path or leaked audio-stage caps	Recap; check the sound DMA path
Hard disc not found	IDE cable/power or dead drive	Reseat the cable; replace the drive (mind the RISC OS 3.5 ~512 MB limit)
One subsystem dead, rails present	Blown on-board fuse	Check FS1/FS2
Won't start with a second CPU/PC card fitted	CPU-card seating or a faulty card	Reseat; test with the primary CPU card only

CPU card, SIMMs and VRAM

The processor is on a plug-in card in one of two CPU sockets (the second socket takes a second processor or an x86 PC card). A machine that will not start should be retested with only the primary CPU card fitted, the card reseated, and the DRAM SIMMs and VRAM reseated. The POST reads the ARM ID late in the sequence, so a CPU-card or RAM fault often shows as a hang on the second purple phase.

Storage and ports

• Hard disc not detected: check the 40-way IDE cable and drive power; reseat. RISC OS 3.50 limits ADFS discs to about 512 MB.
• Floppy: clean the head and try a known-good disc before suspecting the FDC37C665 'Combo' controller.
• Serial/parallel: handled by the FDC37C665; check the connector and the controller.
• Expansion: clean the podule backplane connector and the network-card slot; a machine that misbehaves only with a card fitted usually has a dirty or shorted contact.

Common fault catalogue

• Completely dead — mains fuse, PSU module, or on-board FS1/FS2.
• Crashes early (purple), or RAM-control fault (00020000) — leaked surface-mount electrolytics on the RAM control lines; recap and clean (the most common Risc PC cause).
• ROM checksum fault (00000200) — reseat/replace the RISC OS ROMs.
• VIDC/sound timing faults (00004000/00008000) — VIDC20 path; check after a recap.
• Lost time/config — flat or leaked NiMH battery (BT1).
• No display — monitor/cable, video ID pin, or VIDC20.
• No hard disc — IDE cable/power or a dead drive.
• Subsystem dead, rails good — blown on-board fuse FS1/FS2.

Related pages

• Acorn Risc PC 600
• Acorn Risc PC 600 Maintenance Guide
• Acorn Risc PC 600 Capacitor Guide
• Acorn Risc PC 700 Troubleshooting Guide
• Acorn Risc PC Technical Reference Manual · Capacitor Failure Symptoms