Amstrad Mega PC motherboard with a blue NiCad battery and blue-green deposits across nearby circuitry.

Amstrad Mega PC motherboard with a blue NiCad battery and blue-green deposits across nearby circuitry.

Close-up of blue-green corrosion around the Mega PC's battery, nearby chips and circuit board tracks.

The NiCad battery left a rather impressive blue mess behind. The motherboard was less impressed.

Close-up of two black diodes, nearby chips and scattered blue corrosion on the motherboard.

Those copper tracks are not going to be happy!

Empty battery footprint with blue residue near two black diodes and exposed copper tracks.

Well at least the battery disappeared from the board easily enough.

Such a wonderful shade of blue. Unfortunately it has eaten the tracks on the motherboard. That’s the damage a leaking NiCad battery can do. The electrolyte is alkaline potassium hydroxide rather than acid, but it can still make an impressive mess of a motherboard.

Discoloured copper trace beneath the floppy-drive connector on the Mega PC motherboard.

I don’t like the way the vias near the floppy drive have bright blue in them…

Wider view of the motherboard showing the empty battery footprint, exposed copper and floppy-drive headers.

Or the way the copper tracks are now exposed and no longer shielded.

Underside of the Mega PC motherboard showing discoloured tracks, vias and rows of solder joints.

Just as messy, if not worse on the underside.

I cleaned everything up as best I could with some isopropyl alcohol, then mapped out the damage. The alcohol cleaned the surface, although it wasn’t chemically neutralising the alkaline electrolyte. Multimeter time.

Annotated underside of the Mega PC motherboard, with coloured lines tracing connections around the damaged area.

I mapped the affected tracks on the underside…

Annotated component side of the Mega PC motherboard, with red and green lines tracing connections near the former battery.

…and on the topside before the soldering started.

From the maps above it looked as though most of the damage affected at least one of the serial ports. Serial? Pffft, I have PS/2 connectors.

Unfortunately, one of the main power tracks from the power connector was damaged as well. I handed the board over to a guy with a red shirt and he did some soldering while I stood around eating candy.

Close-up of the cleaned battery area, nearby diodes and exposed trace on the Mega PC motherboard.

D12 to C127 if I’m reading that right!

Underside of the motherboard with blue, orange and red jumper wires soldered across damaged tracks.

Thick traces need thicker wire.

Underside of the motherboard showing repair wires held in place beneath yellow Kapton tape.

A generous amount of Kapton tape to keep our shiny new wires away from the case.

Shiny new wires on the underside of the board.

Blue BIOS setup screen with onboard IDE enabled and manual cylinder, head and sector settings for the first hard drive.

Hard drive auto-detect didn’t work, so I manually entered the drive geometry.

Seagate ST3120A hard drive identification label, shown during attempts to get the Mega PC's storage working.

The 106 MB Seagate was happy in another machine. The Mega PC had other ideas.

Quadtel Extended BIOS Setup screen showing enabled onboard IDE and a user-defined hard-drive geometry.

Quadtel Extended BIOS Setup really felt modern!

After reassembling everything, I found that the PC side wouldn’t recognise either of the hard drives I tried. I started with a roomy 106MB Seagate, then moved on to a beefy 365MB IBM drive. Both worked fine in another machine.

Either I’m really bad at entering drive geometry or something hard-drive-controller-like on the motherboard is fried.

FDISK drive-selection screen showing disk 2 with 102 MB free and disk 1 reporting zero megabytes.

FDISK could see 102 MB free. Actually partitioning it was another matter.

MaxBlast Plus warning dialog reading 'BIOS Read Error 80 on drive 81'.

MaxBlast Plus was not particularly optimistic either: BIOS Read Error 80.

Quadtel BIOS setup screen showing another user-defined hard-drive geometry with onboard IDE enabled.

Can’t get enough of that sweet Quadtel BIOS setup!

I thought I was close with that FDISK result above, but partitioning still failed and the system kept throwing errors. There is only one hard-drive connector on the motherboard, so I’ll have to wait for the other motherboard I ordered to arrive and start comparing.

Overall, though, a good result. The Sega side works perfectly and the PC side works somewhat. I’m very much looking forward to receiving the other Amstrad motherboard so I can finally work out what is going on with the hard drive.

Sources