Some time ago, I was sent a Barcrest Double Up fruit machine program cartridge to dump. At first glance, it seemed fairly straightforward, but it didn’t take long for things to get a little more interesting.
The cartridge appeared to be running version 1.3, which immediately stood out. Every reference I could find online pointed to version 1.5 as the commonly found release, with no mention of earlier revisions. That alone made this cartridge worth a closer look.
Inside the cartridge was a standard 27C512 EPROM, complete with a handwritten label reading “DU 1.3 P3”.
A handwritten label can be a sign of a prototype/location test, upgrade, conversion or hack so my interest was piqued even more. The EPROM read cleanly with no issues, and I soon had a binary image ready for inspection.
That’s where things took an unexpected turn.
On examining the contents of the dumped binary, it quickly became apparent that this wasn’t Double Up at all. The code ASCII text points towards this being from Centrepoint, another Barcrest title.
Further searching revealed that the version 1.3 did, in fact, line up with several files already floating around online. Encouraged by this, I downloaded a handful of them for comparison. However, none of them matched my dump exactly. A binary comparison showed five bytes differed across the files. Just five—but enough to raise more questions than answers.
At this point, my lack of hands-on experience with fruit machines became a limitation. I don’t currently have any hardware or emulation setup to test the effects of those changes, so for now I can only speculate. It’s entirely possible that the differences are insignificant—region flags, payout tables, protection tweaks—or they could represent something more meaningful.
Looked at my dump of this title in Ghidra and comparing it, I can see that the internet file has the ROM check NOP’d out.
Adding another layer of complexity is the presence of a PAL “characteriser” IC on the board. The device fitted to this cartridge is marked “M515”. Again, this raises further questions, as I’m no longer entirely confident that this is the correct PAL for this particular title or version. I’ve started reversing the PAL to better understand its role, but that investigation is still ongoing.
As with many things in this space, what began as a simple EPROM dump has turned into a small mystery and one that I’ll be revisiting once I have the right tools and a bit more background knowledge.
Leave a Reply