IBM PS/2 Model 70 Troubleshooting Guide

IBM PS/2 Model 70. Source: Wikimedia Commons.

This guide documents fault diagnosis for the IBM PS/2 Model 70 (machine type 8570, all submodels). The Model 70 shares much of its POST architecture with the IBM PS/2 Model 50 / 60 but has 80386DX-specific and Power Platform-specific failure modes, and is uniquely afflicted by the SMD electrolyte leakage that defines the PS/2 plague.

Reference Diskette and Diagnostics

The Model 70 requires the Model 70-specific Reference Diskette. The Power Platform-equipped 8570-B21 / -B61 uses a different Reference Diskette image with BIOS extensions for the 80486 daughtercard. Reference Diskettes are not cross-compatible between PS/2 submodels.

Boot Options:

• F1 — boot the Reference Diskette (Set Configuration, Set Features, Display Configuration, Copy Option Diskette, Run Auto Configuration).
• Ctrl-A from the Reference Diskette menu — Advanced Diagnostics. Advanced Diagnostics for the Model 70 includes 80386-specific tests (paging, protected mode entry/exit, 32-bit data path) not run on the Model 50 / 60.

POST Sequence

The Model 70 POST runs in this order:

1. Reset; CPU register check (80386 includes paging and protected mode tests).
2. ROM checksum.
3. CMOS / RTC battery check.
4. Planar RAM count (with one beep at 640 KB).
5. Onboard VGA initialisation.
6. Floppy controller and drive seek.
7. MCA adapter ID scan (each card returns a 16-bit ID).
8. ADF lookup against CMOS configuration record.
9. ESDI controller initialisation.
10. Boot device selection.

A four-digit numeric code displayed top-left on the screen is a fault code from the test block currently running.

Beep Codes

Model 70 beep codes

Beeps	Meaning
1 short	POST passed; normal boot.
2 short	Configuration error; numeric error on screen.
1 long, 1 short	Planar fault.
1 long, 2 short	Display adapter fault.
Continuous	Power supply or planar fault.
None, no display	Planar or PSU fault before video init.

Numeric POST Codes

POST codes follow the IBM convention where the test block is the leading digit(s). The Model 70 shares the 1xx–10xxx code families with the Model 50 / 60 but adds 80386-specific tests.

1xx — Planar / System Board

Planar POST codes

Code	Meaning	Notes
101	Interrupt failure	Planar IRQ controller fault.
102	Timer failure
103	Timer IRQ failure
104	Protected-mode failure	80386 protected mode test failed. Often planar SMD cap leakage near the CPU.
105	Keyboard controller failure
106	Logic gate / planar logic failure
107	NMI failure
108	Timer-bus timeout
109	DMA test failure
110	Planar parity check 0
111	I/O parity check
112	Watchdog timeout
113	DMA arbitration timeout
114	Paging test failure	80386-specific.
121	Unexpected hardware interrupt
151	Real-time clock failure	RTC chip itself (not just the cell).
161	CMOS configuration empty	Dead RTC battery; replace DS12887 module.
162	CMOS checksum bad
163	Time and date not set
164	Memory size mis-match with CMOS	Run SETUP from Reference Diskette.
165	MCA adapter ID mismatch	Run Auto Configuration.
166	MCA arbitration failure
167	Real-time clock failure

A 104 on a Model 70 frequently indicates SMD electrolyte leakage near the 80386 CPU socket has corrupted address or control lines. Recap before any other diagnosis.

2xx — RAM

Memory POST codes

Code	Meaning
201	Memory test failed (planar or memory adapter card)
202	Memory address line error
203	Memory address line error
215	Memory option-adapter failure (e.g., MCA Memory Adapter)
216	Memory option-adapter address conflict
225	Wrong memory speed for board (rare on Model 70)

The 201 number is followed by the failed memory block address (8 hex digits). Use the address to identify the failing SIMM bank or MCA memory adapter slot.

3xx — Keyboard

Keyboard POST codes

Code	Meaning
301	Keyboard not responding	Wrong port? Try a known-good keyboard.
302	User-indicated keyboard error
303	Keyboard / system unit interface fault
304	Keyboard or system unit interface failed
305	+5 V fuse on planar (keyboard fuse)
365	Keyboard mouse problem (PS/2 ports)

6xx — Floppy Drive

Floppy POST codes

Code	Meaning
601	Floppy / drive failure
602	Floppy drive boot record failure
604	Floppy diagnostic error
611	Floppy adapter / controller failure
613	Cable / drive A: failure
621	Drive seek failed
622	Drive CRC error
624	Floppy data bus addressing error
626	Floppy data compare error

8xx — Math Coprocessor

The Model 70 (E61 / 121 / A21 / A61) supports an optional 80387DX coprocessor. The Power Platform B21 / B61 has the math unit built into the 80486DX.

Coprocessor POST codes

Code	Meaning
801	Coprocessor test failed

104xx — ESDI Fixed Disk Adapter/A

The Model 70 uses the IBM ESDI Fixed Disk Adapter/A for its 60 / 120 MB ESDI drive. The 104xx error family is the same as on the Model 50 / 60 ESDI submodels.

See IBM PS/2 Model 60 Troubleshooting Guide for the complete 104xx code table — the same codes apply to the Model 70.

In summary:

• 10450 / 10451 / 10452 — drive-side errors.
• 10455 / 10499 — controller failure. Recap controller card before condemning.
• 10463 — write/read sector error. Same.
• 10464 — primary boot record read failure (drive or filesystem).
• 10480 / 10481 — drive 0 / 1 fatal error.
• 10491 / 10492 / 10493 — user-indicated read/write/controller error during Advanced Diagnostics.

Display POST Codes

The Model 70 has onboard VGA on the planar. The 2401 / 2402 / 24xx code family covers the planar VGA. A 24xx error is a planar / VGA fault — and on the Model 70 is one of the more common SMD electrolyte leak symptoms (electrolyte corrodes traces near the VGA controller chip first because they are the most densely-packed on the planar).

Display POST codes

Code	Meaning
2401	Onboard VGA POST failure
2402	VGA video memory failure
2403	VGA palette DAC failure
2408	User-indicated VGA failure (Advanced Diagnostics)
2409	VGA test failure
2410	Planar VGA card failure

SMD Electrolyte Leak Diagnostic Workflow

The Model 70 / 80 planar's surface-mount aluminium electrolytic capacitors are the leading cause of "the Model 70 won't POST" complaints. A leaky SMD cap can produce:

• A 104 (protected mode) error if the leak is near the CPU.
• A 201 (memory) error if the leak is near the SIMM controller.
• A 2401 (VGA) error if the leak is near the VGA chip.
• Random, intermittent reboots if the leak is on a bus signal.
• No POST at all if the leak has shorted a power rail.

If a Model 70 is exhibiting any persistent or intermittent POST fault, inspect the planar for SMD electrolyte leakage before any other diagnosis. Read IBM PS/2 Model 70 Capacitor Replacement Guide for the recap procedure.

MCA-Specific Faults

165 Card-ID Mismatch

The Model 70's three MCA slots make 165 less common than on a Model 80, but the fault mode is identical. Causes:

• Card was added or removed without running Auto Configuration.
• MCA edge contact oxidation.
• RTC battery died and CMOS lost the configuration record (cluster with 161/162/163).
• CMOS battery is fine but the saved configuration was corrupted.

Fix:

1. Boot the Reference Diskette.
2. Select Set Configuration → Run Auto Configuration.
3. Insert option diskettes as prompted.
4. Save and reboot.

If 165 persists after Auto Configuration:

• Reseat every MCA card. Clean edge fingers with eraser and contact cleaner.
• Try removing all cards, run Auto Configuration with bare planar, then reinstall cards one at a time.
• Suspect SMD electrolyte leakage on the planar near the MCA bus controller.

Power Platform-Specific Faults (8570-B21 / B61)

The Power Platform daughtercard adds its own failure modes:

• Power Platform regulator failure — the on-card +5 V regulator can fail and the daughtercard receives no power. Diagnosis: probe the +5 V test point on the Power Platform.
• Power Platform clock generator failure — the 80486DX receives no clock.
• Power Platform Reference Diskette mismatch — if a standard E61 / 121 / A21 Reference Diskette is booted on a Power Platform-equipped 8570, the 80486-specific BIOS extensions are not loaded and some diagnostics will fail.
• Cache test failure — the Power Platform exposes the 80486 on-die 8 KB cache for testing in Advanced Diagnostics. A cache test failure is a daughtercard fault, not a planar fault.

ECA Recalls and Service Bulletins

The Model 70 was affected by several Engineering Change Authorisation (ECA) bulletins:

• ECA 087 — Reference Diskette compatibility update for newer MCA cards.
• ECA 092 — Planar fix for one revision of the Model 70 with an erratic memory timing.
• ECA 117 — ESDI controller firmware update.

These ECAs apply only to specific FRU part numbers; check the planar / card FRU against the bulletin before applying.

PSU Faults

Symptoms and diagnosis:

• Dead — no fans, no power: Bulk capacitor or mains rectifier; verify mains; PSU recap required. See IBM PS/2 Model 70 Capacitor Replacement Guide.
• Fans spin briefly, then click-retry: Power Good not asserted in 150 ms. Could be PSU fold-back or shorted planar tantalum / leaked SMD electrolyte.
• Boots cold, fails when warm: Secondary electrolytics aged. Recap.
• Audible whine, smell of fish: RIFA X2 cap is venting.
• Rails low/high: PSU feedback path issue. Recap and retest.

Drive Stiction

ESDI drives in the 8570 are prone to spindle stiction after long storage. Same procedure as the Model 60:

• Power off.
• Open the chassis. Locate the drive.
• Gently rotate the drive case 45–90° around its spindle axis in both directions to free the heads.
• Re-install. Power on.

After the drive boots, immediately image its contents to a modern disk image file. The drive cannot be trusted to spin up reliably again.

Floppy Faults

Same as Model 60 (see IBM PS/2 Model 60 Troubleshooting Guide).

Keyboard / Mouse Faults

• 301: Keyboard or mouse in the wrong port.
• 305: +5 V fuse on planar blown.
• 365: PS/2 mouse fault.

Memory Faults

• 201 with address in the on-planar range: SIMM (planar memory) failure. Note: the Model 70 uses IBM proprietary 72-pin SIMMs.
• 201 with address in an MCA memory adapter range: Memory adapter card failure.
• 215 / 216: MCA Memory Adapter configuration error.

When to Suspect the Planar

• 1xx errors that persist after CMOS battery replacement and Auto Configuration.
• 10x errors with no other peripheral fault visible.
• No video and no beeps after PSU verified known-good.
• Repeated 165 errors after every card removed and reseated and a full Auto Config done.
• Any visible SMD electrolyte leak.

The single most common Model 70 planar failure is SMD aluminium electrolyte leakage. See IBM PS/2 Model 70 Capacitor Replacement Guide.

Diagnostic Workflow

1. Visually inspect planar for SMD electrolyte leak first. If found, do not power on.
2. Power on. Listen for beep.
3. Note POST screen — any leading numeric code.
4. If 161/162/163 cluster — replace DS12887 module.
5. If 165 — run Auto Configuration.
6. If 17xx (n/a — no ST-506 on Model 70) / 104xx — verify drive type in CMOS; reseat drive cable; consider stiction; consider controller cap failure.
7. If 24xx — planar VGA fault, often SMD electrolyte leak.
8. If 104 — 80386 protected mode fault, often SMD electrolyte leak near CPU.
9. If no beep and no video — PSU first (rail check), then planar SMD inspection.
10. Run Advanced Diagnostics (Ctrl-A) once basic POST passes.

⚠️ Power-supply RIFA capacitor and tantalum shorts

Two age-related failures are near-universal on this era of IBM hardware:

• RIFA mains-filter capacitors in the power supply are metallised-paper parts that crack and fail short with age, producing acrid smoke shortly after power-on. Replace them pre-emptively with modern X2-class parts.
• Tantalum capacitors on the planar (system board) and on ISA cards fail short with age. A shorted tantalum will prevent the power supply from starting (dead machine, PSU protection latched) — look for a cracked or discoloured tantalum and lift suspect ones to find the short.

IBM PC/XT switching supplies also need a minimum load to start, so a bare supply on the bench may not run without a dummy load.

⚠️ CMOS / RTC battery

This machine keeps its configuration in battery-backed CMOS, and the battery is a common failure. On AT-class boards the clock/CMOS is often a Dallas DS1287/DS12887 module with the cell sealed inside; it lasts about ten years and then dies, giving 161 / 163 CMOS and clock errors at POST (and sometimes spurious floppy-drive errors). PS/2 planars use a rechargeable barrel or pack battery that leaks and corrodes the board. Replace a dead Dallas module (or rework it with an external coin cell), and on a leaking planar battery remove it and clean the corrosion before it eats the traces.

Related Pages

• IBM PS/2 Model 70
• IBM PS/2 Model 70 Maintenance Guide
• IBM PS/2 Model 70 Capacitor Replacement Guide
• IBM PS/2 Model 80 Troubleshooting Guide — tower sibling, same planar family
• IBM PS/2 Model 60 Troubleshooting Guide — predecessor

References

• PS/2 Error Codes — Ardent Tool of Capitalism. Source for the 1xx–104xx error code tables.
• IBM PS/2 Model 70 — Ardent Tool Quick Reference. FRU data and ECA history.
• IBM, IBM Personal System/2 Hardware Maintenance Manual (S52G-9971-02, October 1994). POST error code reference.