Hell, yet another interesting damn pirate cartridge of MMC3!
:
Here is the pinout of the unknown chip :
Any idea what it can be?!
It seems TTLs are responsible for these pins :
IRQ
CIRAM
CHR A16
And the whole circuit lacks these unnecessary pins comparing to a real MMC3 :
D6
D7
CHR A17
PRG A17
PRG A18
WRAM +CE
WRAM /CE
WRAM /WE
Is it a knockoff Namco 108 clone? That would explain having to implement the MMC3 IRQ (the 4020), the ability to swap left/right pattern tables and $8000/$c000 in PRG (the 74'151), and needing to implement extra bits to get 128 KiB of CHR.
Here is a similar case :
NTDEC8701
Ok, the 74'259 (addressable 8x1 latch, parallel out)+74'151(1 of 8 multiplexer) together make CHR A16, increasing CHR from the 64 KiB supported by the N108 to 128KiB.
The 74'161 latches the data written to $8000: the register address.
The 74'138 decodes four of the eight registers the MMC3 nominally has. I can definitely allocate for $E000, $8000, and $8001, I'm not really certain what the last one would be. Probably $E001.
The 74'74 both contains the IRQ enable bit and the "IRQ fired" bit. Interestingly, both halves seems to be latching CPU A1, not just a fixed value.
The 74'02 both generates CHR/CE with NOT(NOR(PPU A13,PPU /RD) and ... something else, I can't see what those traces are doing.
The 74'4020 seems to connect its Q9 bit to the 74'74's 1/CLK input, or an interrupt after 1024 somethings ... probably M2 cycles. (almost exactly 9 scanlines, FWIW)
I can't figure out what the last IC is. Its silkscreening is really hard to see in the photo, but it appears to be connected to CPU D0..D3.
Does it contain a functional scanline counter?
Hahah, look at the completely random brand selection on those chips! It's like they wanted to be sure to use one of everyone's chips - I see Signetics, NEC, Mitsubishi, Goldstar, Hitachi, one I can't place, and of course the unlabeled packages.
( "M in a rounded box" is Matsushita/Panasonic. )