Restoring the Plexus P20: Modern Revival of a Vintage Unix Workstation

Jul 3, 2024

Adrianne's Digital Basement: Plexus P20 Part 5

Introduction

  • Rare Plexus P20: dual processor multi-user Unix workstation (circa 1985).
  • Part five in the series documenting efforts to restore it to working condition.
  • Recap of parts 1-4:
    • Part 1: Disassembly, motherboard examination, hard drive imaged.
    • Part 2: Power supplies serviced, initial power on, serial port pinout figured out.
    • Part 3: Original hard drive reinstalled, system booted into Unix but then failed.
    • Part 4: Diagnostic journey, Discord community assistance, identified and replaced a faulty flip-flop chip.

Current Problems and Goals

  • Issues to resolve:
    • Diagnostic ROM error.
    • Clock circuit failure (likely clock chip).
    • Multibus error (related to expansion chassis).
  • Goal: resolve errors and achieve stable Unix boot.

Diagnostics and Investigations

Multibus & Power Supply Configuration

  • Multibus board: Not required for basic boot, supplies power to other cards if multiple power supplies are used.
  • Power Supply: Connected to backplane or directly to motherboard.

Initial Diagnostic Check

  • Powered up system: serial output shows system does not boot due to multibus error and clock calendar failure.
    • Ta (test all): Multibus passes but clock/calendar fails.
  • Clock chip address: D000000 (memory mapped IO in 68k system).
  • Clock chip pins: data not as expected, indicated possible chip or connecting logic issues.

Examining Clock Circuit

  • Probed and verified signals on clock chip using oscilloscope.
  • Identified power supply connections and chip select logic as possible failure points.
  • Simulation of circuit: Verified clock chip enable logic.
  • Contributors: Discord user SuperSVGA provided simulated logic for enabling clock chip.
  • Attempted direct grounding of chip enable line, observed expected results.

Replacing Components

  • Replaced clock chip: Confirmed faulty, socketed replacement worked with test runs.
    • Issues persisted with diagnostic calls.
    • Confirmed functionality by running timeframe tests on new chip.
  • Formatted new Dallas clock chip with external 2032 battery for persistence.
    • Added resistor to initiate Motorola mode for compatibility.

System Boot and Configuration

Boot Sequence and Error Resolution

  • Powered up system: verified retaining configuration.
    • Multiple iterations of setup confirming SC7 settings for tape drive resolved diagnostic errors.
  • Unix boot successful, confirming operational state.

Future Plans

  • Next Steps: Further Testing and Adjustments:
    • Continued validation: RJ45 check, system longevity stress testing.
    • Floppy drive repair.
    • BlueSCSI: potential for virtual storage, easier data transfer.
    • Hard drive image sanitization and sharing for broader access.
    • File transfer solutions: investigating viable methods suitable for Unix System V.

Reflections and Acknowledgments

  • Acknowledged Discord community for support and collaboration.
  • Thanks to contributors for data recovery and logic simulations.
  • Intention for further documentation to aid restoration and usage of vintage systems.

Conclusion: System operational and future-focused testing ongoing for stable and full utilization of the Plexus P20.