Mapper 172 (images from MLX):
Attachment:
DumaFront.JPG [ 1.58 MiB | Viewed 746 times ]
Attachment:
DumaBack.JPG [ 1.62 MiB | Viewed 746 times ]
According to my trace attempt (correct me if I'm wrong), I identify the following JV001 pin functions:
Code:
Pin# Mapper 136 connection Mapper 172 connection General pin function
----------------------------------------------------------------------------
1 PPU A13 PPU A13 PA13
2 CPU D5 CPU D0 Input D5
3 CPU D4 CPU D1 Input D4
4 CPU D3 CPU D2 Input D3
5 CPU D2 CPU D3 Input D2
6 CPU D1 CPU D4 Input D1
7 CPU D0 CPU D5 Input D0
8 CPU A0 CPU A0 ($0001) A0
9 CPU A1 CPU A1 ($0002) A1
10 CPU A8 CPU A8 ($0100) A8
11 CPU M2 M2 M2
12 CPU /ROMSEL CPU /ROMSEL ($8000) /ROMSEL
13 CPU R/W CPU R/W R/W
14 +5V +5V +5V
15 CPU A14 ($4000) A14
16 CPU A13 ($2000) A13
17 CHR /OE CHR /OE CHR /OE
18 Output D5
19 Output D4
20 Output D3
21 Output D2
22 CHR A14 Output D1
23 CHR A13 Output D0
24 PPU A10 PA10
25 PPU A11 PA11
26 CIRAM A10 CIRAM A10
27 PPU R/W PPU R/W
28 GND/+5V GND/+5V
Which is odd, because that would mean that Mapper 172 has mapper-controlled mirroring, which nobody emulates. When MLX redumped Mahjong Block (Mapper 172 version), the Kazzo identified the mirroring type incorrectly, which would not happen in the case of fixed mirroring, so it's not entirely out of the question. I would not have any idea what write pattern selected the mirroring type though.
I found it easier to trace the Mapper 172 images, which is why the Mapper 136 column is rather empty. Yes, Mapper 172 has CPU D0..D5 connected in reverse, and the three Mapper 172 games indeed have code to reverse the desired six bits before they output the data to $4102.
@krzysiobal: I only noticed this now in your otherwise excellent schematic, but F-15 City War is Mapper 173, not 136. (The thread title was originally "Sachen TCU-2/Mapper 136" because that one started the thread. I have since renamed the thread title to reflect the expanded scope of the discussion.)