IBM Personal Computer 5150

This 5150 arrived in a well-worn cardboard box with most of its original packing foam. The previous owner had stored it in a garage for over two decades, and it showed: the case was heavily yellowed, and the keyboard had a sticky, unresponsive feel consistent with decades of ambient humidity exposure.

Initial power-on test produced a single long beep followed by two short beeps — the POST memory error code — and no video output. The CGA card was seated poorly in its ISA slot.

The power supply showed signs of swollen capacitors visible through the ventilation slots. A strong burning smell upon startup confirmed the PSU was in urgent need of attention before further testing could proceed safely.

1. Power Supply Restoration
The original IBM 63.5W linear power supply was fully recapped using Nichicon and Panasonic electrolytic capacitors matched to the original spec. The primary filter caps (2× 3300µF/16V) were replaced, as were all secondary side smoothing caps. Output voltages were trimmed and verified: +5V ±0.05V, +12V ±0.1V, -12V ±0.2V, -5V ±0.1V.

2. Motherboard Inspection
The original 16-64 KB System Board (revision E) was cleaned with IPA and inspected under magnification. All socketed chips were reseated. DRAM banks were tested individually; all 64K chips passed. DIP switch settings confirmed correct for the installed configuration.

3. Keyboard Restoration
The Model F keyboard (83-key) was fully disassembled. The PCB was cleaned with isopropyl alcohol. All keycap stems were lubricated with Krytox 205g0. The coiled cable was cleaned and tested; all 83 keys now actuate correctly with the characteristic Model F click.

4. Case Retrobright
The case was retrbrighted using a 12% hydrogen peroxide cream gel under UV light over a period of 6 hours. The result is a uniform cream-beige finish very close to the original IBM colour specification.

5. Disk Drives
Both 5.25" half-height drives were cleaned with IPA and the heads aligned using a diagnostic disk. Both drives now read and write reliably. Drive belts were intact and not replaced.

After restoration, the machine was run for a 24-hour burn-in test with POST looping and then booted into IBM PC DOS 2.1 for functional testing. All expansion slots were tested with a known-good CGA card and a multi-function I/O card.

The machine now passes all POST checks (single short beep), boots cleanly from both floppy drives, and runs BASIC, Lotus 1-2-3, and a selection of era-appropriate software without issue. VGA output via the CGA card is stable.