Every so often a machine comes through the workshop that you've wanted to work on for years. For us, this was one of them: a Sharp X68000 Compact, the striking red “Red Zone” finish, that had stopped booting and clearly hadn't been powered on in a long time. The brief was simple to state and a lot less simple to execute: redo every capacitor, undo whatever damage the old ones had done, and get it running with a modern SCSI storage adaptor in place of its long-dead hard drive.

1993

Compact released

Sharp X68000 Compact XVI

68000

Motorola CPU

16 MHz in the Compact XVI

100%

Caps replaced

Main board, PSU and drive

Why the X68000 Is Worth the Effort

If you didn't grow up in Japan, the X68000 is easy to miss. Launched by Sharp in 1987, it was a home computer with near arcade-grade graphics and sound at a time when most Western machines couldn't come close. Software houses used it to develop and port coin-op games, and its arcade-perfect conversions became legendary. The original machines are instantly recognisable by their twin-tower “Manhattan shape” case.

The Compact, released in 1993, condensed that hardware into a much smaller cube-style case. Ours was the Compact XVI, a 16 MHz Motorola 68000, and like every X68000, it relies on SCSI for its storage. That single design decision is what makes these machines so satisfying to revive in 2025: a dead 30-year-old hard drive is no longer a dead end.

The X68000 was so capable that Capcom used it as the basis for the CPS (Capcom Play System) arcade development environment. When people say a home computer could run “arcade-perfect” ports, the X68000 is the machine that actually earned the phrase.

First Assessment: The Usual Suspects

Before anything gets powered on, an X68000 gets a full visual inspection. These machines are notorious for surface-mount electrolytic capacitors that leak their electrolyte onto the board, where it quietly corrodes traces, vias and component legs underneath. By the time a machine “won't boot,” the caps are rarely the only problem, they've usually already started eating the board.

Sharp X68000 Compact (XVI)

1993 · Motorola 68000 @ 16 MHz

High Risk
Surface-mount electrolytic caps leak and corrode the board beneath them
Power supply caps degrade, a genuine reliability and safety concern
Floppy drive eject mechanism and belts perish with age
Original SCSI hard drives are usually long dead
RTC / battery leakage on some units adds another corrosion source

Never power on an untested X68000 'to see if it works', a failing PSU or leaking caps can turn a repairable machine into a much bigger job.

⚠️ Don't power it on first

The single most damaging thing you can do to a stored X68000 is plug it in to “check if it still works.” Aged power-supply capacitors can fail under load and take other components with them. Every machine we restore is assessed, recapped and bench-tested on a current-limited supply before it ever sees mains power normally.

The Recap: Every Last One

The brief was a full recap, and that's exactly what these machines need, not just the ones that look bad. A capacitor that measures fine today on an X68000 is still 30 years old and sitting next to ones that have already leaked. Doing half a job means a return visit. So the work covered three areas:

01

Main board

Every surface-mount electrolytic removed, the pads and surrounding area cleaned of old electrolyte, and fresh capacitors fitted. Where electrolyte had crept under the caps, the board was neutralised and cleaned before anything new went down.

02

Power supply

The PSU is the most important, and most overlooked, recap in the whole machine. Aged primary and secondary capacitors were replaced so the supply delivers clean, stable rails the rest of the system can trust.

03

Floppy drive

The internal drive was serviced, caps replaced, mechanism cleaned and the perished components in the eject system addressed so it actually loads and ejects disks reliably.

With the old caps out, the real work begins: undoing the damage they left behind. Leaked electrolyte is mildly corrosive and conductive, so every affected area is cleaned and neutralised, traces are checked for continuity, and any that the electrolyte had eaten through are repaired. This is the part that separates a machine that “boots on the bench” from one that's genuinely reliable.

💡 Match the caps to the job

A good recap isn't just “new caps”, it's the right type, value and voltage rating in each position, with the correct footprint and polarity. On a board this age, lead-free reliability and low-ESR parts matter as much as getting the numbers right.

Storage: A Modern SCSI Adaptor

With the board healthy, the last hurdle was storage. The X68000 boots from SCSI, and its original mechanical hard drive, like almost all of them, was no longer viable. The fix we used here is one of the best things to happen to retro computing in years: a PiSCSI.

PiSCSI (formerly RaSCSI) is a small board that sits on top of a Raspberry Pi and emulates SCSI devices for vintage computers. To the X68000 it looks like one or more period-correct SCSI hard drives; in reality those “drives” are just disk-image files living on the Pi, which you manage from a web interface on any modern computer. It's well supported on the X68000, there's even a dedicated setup guide for the machine, and it handles the X68000's SCSI quirks cleanly.

01

Wire it onto the SCSI bus

The PiSCSI board is connected to the X68000’s SCSI bus, assigned a SCSI ID the machine expects, and set up with correct termination so the bus reads cleanly.

02

Attach a bootable Human68k image

A partitioned, formatted hard-disk image with Human68k, the X68000’s operating system, is placed on the Pi and mapped to the emulated SCSI ID through PiSCSI’s web interface.

03

Format, partition and verify

On the machine itself, the volume is checked, partitioned to the X68000’s expectations, and confirmed bootable and writable before the job is signed off.

Because every “drive” is just an image file on the Pi, the owner gets something the machine never originally offered: trivial, complete backups. Copy the image and a catastrophic storage failure becomes a quick restore, and you can keep several system configurations on hand at once.

Bringing It Back to Life

First power-up after a recap is always done carefully and on a controlled supply, watching current draw before committing to a full boot. With the rails clean, the corrosion repaired and the SCSI adaptor configured, the Compact did what it hadn't done in years: posted, found its SCSI volume, and dropped into Human68k. From there it was testing, floppy access, video output, sound, and a proper soak under power to make sure nothing marginal was waiting to misbehave.

The “Red Zone” Compact went back to its owner as a genuinely usable machine, every capacitor fresh, the leakage damage undone, and storage that will outlive any of us. That's the goal with every one of these: not just “it turns on,” but a machine you can actually rely on for the next decade.

Have an X68000 (or Anything Like It)?

Japanese retro hardware, the X68000, PC-98, FM Towns and the rest, is some of the most rewarding to restore and some of the most punishing to neglect. The capacitors don't care how rare the machine is. If you've got an X68000 sitting in a cupboard, the single best thing you can do is not power it on and get it assessed. We work on these Australia-wide via mail-in, recapping, leakage repair and SCSI storage conversions included.

Got one that needs attention? Get in touch and we'll give you an honest assessment of exactly what it needs.