We’re attempting to dump the good old BIOS from the older of the two Japanese Sega Channel adapters. Unfortunately, getting a readable file and getting a reliable dump aren’t necessarily the same thing. As we’ll soon discover!

First up, I pulled the older adapter apart to see what was hiding inside. The plastic had a July 1994 marking, and there were some interesting differences compared with the other unit, including a few different chips and a wire running across part of the circuit board.

The two adapters are rather similar at first glance, but looking more closely reveals a few changes between them. The older one also has slightly different component markings, which makes me wonder what Sega was changing between the revisions. At least having both machines makes comparing the hardware rather straightforward.

The backup battery wasn’t looking particularly healthy either. I also noticed what appeared to be evidence of previous soldering around the BIOS chip, although I couldn’t say for certain whether that was factory work or something somebody had done later.

Out came the desoldering equipment, and after a little encouragement the ROM was free. Fortunately, the pins survived intact! There was some stubborn solder hiding around the legs, but nothing that couldn’t be persuaded to let go eventually.

One side of the chip came free more easily than the other. I eventually managed to work out where the remaining solder was holding it, using a plastic tool to ease the pins loose. It’s always nice when a job like this ends with all the pins still attached to the component!

The chip was marked NEC D23C4000SCZ, so I started experimenting with compatible device selections in the GQ-4X programmer. After a few unsuccessful attempts, selecting D27C4000 with the 16-bit adapter gave me something readable.

I had to remember the correct socket arrangement on the GQ-4X too, since the 16-bit adaptor adds another little complication. Once the programmer was configured and the chip was properly seated, the reading process was remarkably straightforward. It’s everything that happens afterwards that gets interesting!

Using the byte-swap function revealed some recognisable information, including a July 1994 date and BIOS version 2.11J. That was encouraging! The problem was that the dumps weren’t producing consistent checksums, so I couldn’t yet trust the resulting file as an accurate copy.

I even gave the resulting file a quick look with Fusion, but that didn’t establish a successful working BIOS image either. Still, it’s rather lovely being able to see the Sega Channel version information inside the ROM data after all that desoldering.

So we’ve learnt a little more about this older adapter and managed to read some interesting data from its BIOS. Unfortunately, the checksum problem remains. Looks like there’s another mystery to investigate!

If I end up fitting a socket to this adapter, I’ll need to get permission from its owner before making any further permanent changes. For now, I’d rather get a reliable dump before doing anything else.

Sega Channel Japan Series