Hey all... I have a quick question here that I cant seem to find the answer to. Perhaps its nothing.
I was looking at the Konami VRC mappers. I notice how the VRC2 and 4 'versions' are dependent on its PCB board, because the VRCs pin 3 and 4 are different.
I see that, for example, Bio Miracle, Gradius 2 and Japan Cup use PCB 351406 , which indicates VRC4b, but on the VRC, Gradius 2 and Japan cup are 052785, and Bio Miracle is 054003.
For VRC4e, Kid Drac, Crisis Force and Tiny Toons all use VRC4 "052784" and Parodius is also VRC4e but uses "052785".
Anyone know what these numbers mean? Does that mean for VRC4, on TOP of the PCB differences, there are 3 variants? 054003, 052784 and 052785?
BTW these are NOT date codes. Thanks!
The PCBs appear to have uniquely identifying numbers. Sometimes multiple map to the same functionality (e.g. 350603, 350636, and 351179 all are the mapper 23 VRC2b). It's just that Konami used numbers while Nintendo used words ("HVC-S?ROM-nn")
Yeah but I dont think you addressed my question. Within the same bank of VRC4e's, there are numbers stamped ON the VRC and they're different. For VRC4 theres 3 different numbers. Im not sure if that means the VRC mappers themselves are different or they're just the same. IE VRC4e "052784" is different somehow from VRC4e "052785"
Most likely, different manufacturing runs.
I suppose it's possible there's a die revision inside, too, but we've never found any behavior differences (unlike, say, MMC1A vs MMC1B)
Yeah, manufacturing runs is the most likely bet, or some kind of internal circuitry revision (but the end result is the same, i.e. no mapper or board behavioural differences). Let's not forget that it was pretty common to switch brands (manufacturers) of ICs during this time period too, as focus on cost savings was extremely prevalent (combined with availability of chips).
So, to answer your question: aside from board identification (in an emulator for wiring/compatibility needs), along with mapper number (handled by the NES or NES 2.0 header), it is indeed "nothing" (to worry about) (re: "perhaps it's nothing"). :-)
Ok cool, thanks!
I thought that too, then I was helping someone make a Parodius. The PRG was 256k so i told him to wire A17 to the A17 line of the VRC4 and it still wasn't working. The wiring of everything else looked good but the VRC he was using ended in '84' instead of the '85' in bootgod and I thought that might have been the issue, cuz nothing else made sense.
About two years ago I made some experiments, patching several VRC2 and VRC4 games to run on the TMNT board (
http://bootgod.dyndns.org:7777/profile.php?id=1773). I only needed to change the mapper control writes, nothing else were changed on the games.
What vary between each of those board types is how the ROMs and the mapper are connected to each other and to the famicom slot.
Also, VRC4 is a superset of VRC2 so to make Contra operate on a VRC4 board I only had to patch the protection out (microwire eeprom interface feedback loop) and make sure the writes were going to the right addresses for the board I am using. Twinbee had a oddball VRAM address configuration (one address line is skipped on the mapper registers) so I had to delete a rotate instruction on the program.
Now making VRC4 games run on VRC2 is not really feasible because it lacks hardware (IRQ timer). Some pirates did it by implementing some form of external timer using TTL chips.
I suggest you have a look at the wiki VRC2/4 pages and give it a good read.