PCB Repair: Mortal Kombat II


Severe graphical corruption. RAM UA11 and UA12 reported bad during self-test:

(There was also a sound issue but I’ll cover that another time).


UA11 and UA12 store palette-data for each frame-buffer pixel (and are not to be confused with the actual palette RAM at the final stage of the video output). It’s feasible that both RAMs are defective but it’s also possible there’s a problem with the buses/signals that interface the CPU to the video RAM. What common signals do UA11 and UA12 share? Well, they both reside on the upper half of the 16-bit color data bus:

Video RAM banks 1 and 3, with palette data RAM on CD[15:08]

There are two other components on the CD[15:08] bus that could potentially be at fault:

The buses on the left hand side originate from the CPU (LAD0-15)


Probing the CD bus with an oscilloscope showed that CD13 was stuck at around VCC. Furthermore, the resistance measured between CD13 and VCC was very low: around 15Ω. Presumably this low-resistance path had developed inside one of the four ICs on the CD[15:08] bus. Removing each IC until this ‘short’ disappeared seemed to be my only option.

Hoping that I wouldn’t have to mess with any of the SMD RAMs, I chose to remove UE15 (74ALS245) and UC15 (74LS374) first. Unfortunately the low resistance path was still present without them.

I removed UA11 with the aid of some Fast Chip removal alloy (i.e. cheapo Chip Quik). Bingo! The low resistance path was no more and I congratulated myself for having avoided removing UA12 instead (which now showed as good during the self-test).

The video RAM used by both T-Unit and Wolf-Unit is a 128kx8 multiport DRAM. There are at least three viable replacement parts that I’m aware of:

  • Micron MT42C8128DJ-8
  • Hitachi HM538123AJ-8
  • Toshiba TC528128BJ-80


I ended up replacing UA11 with a slightly faster 70ns part:

Yes, it bothers me that they no longer match. Also, this is my first attempt at drag-soldering an SOJ IC…


Replace RAM at UA11

%d bloggers like this: