PCB Repair Logs Zippy Race Bootleg
Zippy Race (Bootleg)
 |
|
| PCB Image | Reserved |
|---|---|
| Pin Out | Reserved |
Repairer: cal2
Forum Thread: Zippy Race (Bootleg) PCB Repair
Symptoms:
bad text / road layer. Displays as if the data is not read or written to the proper location: Lots of garbage, flickering text, and so on...
Repair:
The text / road layer is handled by the top most pcb (the game pcb is a stack of 3 sandwiched pcb). The top most also handles the sound. Compared to the original, the sound and text layer have been merged to a single one.
My first thought was about ram. There are two 2128 on it. Both tested good after desoldering. Since the amount of IC is limited on the text board, Il desoldered the two first rows of ICs to test them. All tested good.
I put sockets on the pcb and put the ICs back. While poking on the ICs pins, I found that the signals between the 74LS374 @ H1 and 74LS283 @ J1 and K1 were noisy. Between them is a place for a resistor network (pull ups) which has not been installed. So I gave it a try and soldered a resistor network. Signals were a lot cleaner, but it didn't help at all. Same symptoms.
Under the pcb, near the ribbon connectors of the text layer pcb, I found a lot of wires. Repair attempt or original ? I will never know Anyway, I had to find some schematics to understand how this system works.
Of course, zippy race schematics are not available, so I had to use the schematics for the closest game I could find. Moon matrol is nearly the same hardware so I used it.
After a lot of time understaning the schematics, I found an interesting place. There are 3 74LS157 @ 1A,1B,1C.
They are used to multiplex the vram address bus. One address bus is used by the display hardware. The other comes from the cpu board and is used to have write access to the vram.
Compared to moon patrol, there are no 74LS367 connected to these chips on the text layer pcb. The 3 74LS157 are connected directly, through a ribbon, to 3 74L367 on the *cpu* board. Fortunately, the cpu board is the same as for moon patrol so it was easy to find the related chips.
Using a 34 channel usb oscilloscope I found that the vram adress bus used for display was OK. The address counter was doing its job. However the cpu adress bus was strange. But since the cpu makes random accesses to vram it may be normal.
I then compared the cpu address bus on the outputs of the 74LS367 on the cpu board to the inputs of the 74LS157 on the text pcb. All signals were the same except for A3,A4 and A5.
I tested the continuity of the ribbon cable for this signals and found they were open.
Fixing these connexions also fixed the graphic problems.
Lesson learned today: never underestimate the ribbon connectors.
