Namco 50xx/51xx/52xx/53xx/54xx replacement project
Another set of Namco ICs I am dealing with, start with 5xxx. These chips are based on a 4bit Microcontroller from Fujitsu called the "MB8840 series", see the picture on the right side. They are used on the Midway/Namco Galaga Board, the Bally/Midway Super Pacman board and the Atari/Namco Xevious board I own. The datasheet for the "plain" core is still available, I got my copy from a sharp MZ user page.
The required ROM firmware for these custom implementations of the MB884x can be found in recent MAME distributions, they are made available by the Guru's de-capping project. So what was left for me to do was to reconstruct the mb88 core based on the datasheet. Quite easy - as long as the datasheet is not full of typos and missing information - which was the case of course :-)
So again it was necessary to do a lot of analysis of the custom IC behaviour in its "natural environment", e.g. the Galaga board.
Unfortunately the 54xx is a noise generator. Means, we need to produce the same "garbage" on our replica as we see on the original.
Really a challenge to check, I have to say...
The 51xx/52xx and 53xx IC setup was easier to verify, as it has a much more deterministic behavior. I used my FPGA-DIL replacement board and was able to run the replacements on a Galaga, Bosconian and Dig Dug board without problems!
What makes it complicated is the fact, that a fairly complex Microcontroller can be verified only based on the behaviour of a few I/O pins - so the chances to miss some (probably only minor) thing is fairly high. This simply means, that even if it works perfect finally in the above environments, there are still chances to fail in an other environments (like Pole Position or Xevious), just when initialized and used differently. But lets see. I also have a Xevious board but didn't used and check out that one for now. So there is enough to do...
I can provide a basic Xilinx project including the VHDL code which emulates the 51xx, 52xx, 53xx IC sucessfully
on my FDIL setup (xc3s100e-4, see tables on the right) on a Galaga, Bosconian
and Dig Dug board (not all boards include all chips, though). The 54xx IC did work as well, but I am actually
restructuring the code to separate the proprietary firmware from my code to distribute it without this binaries.
For now only the 42pin devices (51/52/53xx) are available as download. Nevertheless, the gate statistics below
is still valid for all setups.
Click here to proceed to the code request page
The synthesis estimates a max. clock speed above 90MHz for the code, although the ICs are expected to run at 1.5MHz.