IBM PC XT/370 Troubleshooting Guide

This guide documents fault diagnosis for the IBM PC XT/370. POST is the standard IBM PC XT (5160) POST in the 1xx–19xx range; XT/370-specific failures present as VM/PC boot errors or as cards-not-detected by the XT host BIOS.

Reference Documents

• IBM SA38-0037-00 — Personal Computer Family Service Information Manual (July 1989), Chapter 6 (XT/370). Authoritative service manual.
• IBM 6137739 — Virtual Machine/Personal Computer User's Guide (December 1984).
• IBM PC XT documents — POST codes, BIOS, hardware troubleshooting carry over directly.

Initial Diagnosis

Power-on sequence on a working XT/370:

1. IBM PC XT POST runs to completion (no error message; single short beep).
2. PC DOS 2.10 boots from the C: drive.
3. AUTOEXEC.BAT calls the VM/PC loader.
4. VM/PC initialises the 370PC-P card via the ISA bus.
5. VM/PC contacts the IBM 3174 or 3274 mainframe controller via the PC3277-EM card.
6. VM/PC downloads CMS system files from the host mainframe to local minidisks (first run only).
7. CMS prompt appears.

If any of these does not happen, stop and diagnose at that stage.

Stage 1 — XT Host POST Fails

Standard IBM PC XT POST codes apply. The 1xx–19xx code series identifies the failed FRU at the chassis level. Cross-reference with IBM PC XT (5160) Troubleshooting Guide.

Common POST codes:

Selected XT POST codes that apply to the XT/370

Code	Meaning	First action
1xx	Planar / system board	Reseat / replace planar
2xx	Memory error	Identify failing 64 KB block from on-screen address
301 / 302	Keyboard fault	Verify keyboard connected; check for stuck keys
6xx	Floppy drive fault	Cable, drive, controller
17xx	Hard disk fault	Cable, drive, fixed-disk adapter

Stage 2 — XT POSTs OK but VM/PC Won't Load

• DOS boots normally — XT chassis is healthy, problem is in the 370PC subsystem.
• VM/PC initialisation message appears but hangs — 370PC-P card not responding. Reseat the 370PC-P card. Reseat the 370PC-P to 370PC-M backplane connector.
• VM/PC reports "memory error" or "card not present" — 370PC-M card not detected. Reseat M card; clean backplane connector with deoxidising contact cleaner.
• VM/PC reports "host connection failed" — PC3277-EM card or the cable to the 3174 / 3274 controller. Check the BNC coax cable, terminator, and controller-side configuration.

Stage 3 — VM/PC Loads but CMS Crashes

• S/370 Program Check — usually means a guest CMS program has executed an opcode the modified 68000 does not implement. The instruction trap then fails because the unmodified 68000's interpreter does not recognise the opcode either. Resolution: update CMS / VM/PC to a known-good level; or restrict the guest program to instructions documented as supported by the XT/370 emulator.
• Spurious memory errors — 370PC-M DRAM failure. Run the VM/PC built-in memory test to identify the failing chip.
• Disk I/O errors — XT 10 MB ST-412 drive failure (more likely than a card fault).

370PC Card Diagnostic Workflow

Because the XT host BIOS does not know about the 370PC cards, the cards do not produce POST codes — they are detected only when VM/PC loads. Diagnostic workflow:

1. Verify the XT chassis POSTs cleanly with the cards installed.
2. Verify DOS boots from C:.
3. Run a small DOS program from C: to confirm RAM is intact (e.g. CHKDSK).
4. Load VM/PC and observe the initialisation messages.
5. If VM/PC reports a 370PC-P or 370PC-M error, power off and reseat the cards and the backplane connector.
6. If the error persists, swap with a known-good card set.

There is no in-system diagnostic for the modified 68000s or the modified 8087. If a card fails, the verdict is replacement — and replacement cards are unobtainium.

3270 Host Connection

The PC3277-EM card (or the standard 3278/79 Emulation Adapter on late-production units) connects to the IBM 3174 or 3274 cluster controller via BNC twinax. Common faults:

• No carrier on the host link — BNC cable, terminator, or 3174/3274 port.
• Host connection times out — controller-side configuration. The XT/370's host LU (Logical Unit) must be defined on the controller as a 3278 or 3279 emulator.
• VM/PC sees the host but cannot download CMS — VM CP (Control Program) configuration on the host mainframe; the user ID must have appropriate permissions and minidisk allocations.

VM/PC Operating System Faults

• VM/PC boots but CMS commands not recognised — minidisks not formatted. Format using the VM/PC FORMAT command (formats a DOS file as a CMS minidisk).
• CMS files do not export properly to DOS — character set conversion issue. Use the EXPORT command with explicit EBCDIC-to-ASCII translation.
• CMS application crashes with "no such opcode" — guest software executed an instruction outside the XT/370 emulator's supported subset. Replace with a different version of the application or a CMS-included tool.

Common Field Symptoms

• XT POSTs, DOS boots, VM/PC won't load — 370PC card subsystem. Reseat in order of likelihood: backplane connector → 370PC-P → 370PC-M → PC3277-EM.
• VM/PC loads, CMS prompt appears, but every command hangs — 370PC-P custom 68000 microcode issue, or 370PC-M memory failure.
• VM/PC loads, host connection fails — coax cable, terminator, controller-side configuration.
• Random reboots during VM/PC operation — PSU at capacity. Recap PSU; see IBM PC XT/370 Capacitor Replacement Guide.
• Smell of fish from PSU — RIFA X2 mains-suppression cap is venting. Power off immediately and replace.

Diagnostic Workflow Summary

1. Power on; observe POST. If POST fails, treat as standard XT chassis fault.
2. If POST passes, observe DOS boot. If DOS fails, suspect XT hard drive or floppy.
3. If DOS boots, load VM/PC. If VM/PC fails, reseat the 370PC cards.
4. If VM/PC loads, attempt host connection. If host connect fails, check the BNC cable and the 3174/3274 controller.
5. If host connect succeeds, attempt a CMS command. If CMS commands hang, suspect 370PC-P custom 68000 failure.

⚠️ 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.

POST beep and error codes

The IBM Power-On Self Test signals faults by beeps and, where a display works, by a numeric code:

IBM POST beep codes

Beeps	Meaning
1 short	Normal POST — system OK
2 short	POST error (numeric code shown on screen)
No beep	Power supply or system-board fault
Continuous / repeating short	Power supply or system board
1 long, 1 short	System-board fault
1 long, 2 short	Display-adapter fault (MDA/CGA)
1 long, 3 short	Display-adapter fault (EGA/later)

Common numeric codes include 161/163 (dead CMOS battery/clock), 201 (memory), 301 (keyboard) and 1701 (hard disc). A code ending in the family prefix identifies the failing subsystem.

Related Pages

• IBM PC XT/370
• IBM PC XT/370 Maintenance Guide
• IBM PC XT/370 Capacitor Replacement Guide
• IBM PC XT (5160) Troubleshooting Guide — chassis-level POST codes

References

• IBM SA38-0037-00 (July 1989), Chapter 6.
• Kozuh, Livingston, Spillman (1984), IBM Systems Journal 23(3):245.
• HelpPC — IBM PC diagnostic codes.