Quoted from pincoder:
Don't worry about hijacking a thread.. especially this one. Working out an issue in public is good for everyone, including future readers.
Here are a few notable facts about these boards:
1 - Sometimes, when the CPU chip crashes, the PIAs let out one last quick blast of outputs through to the connected devices. For all of the PIAs except the solenoid PIA this is not a problem. However, since the sound card is linked to the solenoid PIA outputs on the driverboard, it detects that last blast and because the sound board doesn't crash when the MPU board does, it goes ahead and plays whatever sound it thinks it was told to play. Sometimes the "code" it picks up is invalid and so it doesnt play anything. Sometimes it is valid and thats when you hear a sound. Let's not worry too much about this for now.
2 - 0.5 seconds after power up, the CPU chip is constantly executing instructions. It boots from the IC17 socket, and in the case of the Pincoder Adapter all of the instructions it will need to execute are located solely in the IC17 address space. If for some reason it can't read the next instruction correctly from whatever address it's looking at, the CPU will crash. Moments after this happens, the blanking circuit kicks out and essentially cuts power to all of the outputs (displays, lamps, solenoids, flippers,..).
3 - When working correctly, all of the devices/chips on the MPU and driver board that are connected to the address and data bus lines of the CPU chip should pull themselves off of those busses when they are not being asked to read or write on the busses. If any one of those other devices doesnt "let go" of the busses, the CPU will again crash. So if you're having a boot problem, or a frequently crashing problem and you know your power is good, then pulling as many ROMs, RAM, PIAs, and the 5101 chip out should help reduce and hopefully eliminate the boot/frequent crash issue. Running the 01a-leds test is the best test to experiment with in this case, because it uses the least number of those components possible so removing those components and running the test again should help you narrow down which component is hanging the busses.
4 - You can remove the driver board to run tests 01 through 05. PIA1 cannot be removed for these tests as it is required to run the leds on the MPU board. If you suspect PIA1 you can swap it with another PIA and try again. Most of the time the PIAs are not in a socket, so there's an additional hurdle: remove the PIA without damaging it, then put in a socket and install the PIA.
5 - The "tilt" error is typically caused by IC15 through IC18 being faulty, as they report erroneous switch closures to the PIA, one of them being "tilt".
To me, it sounds like your game possibly:
- Has power problems (you have enough volts, but maybe not enough current, or perhaps the power is noisy or has too much ripple).
- Has bad ROM chip(s), or sockets.
- Has a bad switch PIA (IC11), or 4049 (IC15, IC16), or 7406 (IC17, IC18).
- Has a bad lamp PIA (IC10), or 7406 (IC12, IC19)
- Has one or more bad pins on the 40 pin connector, (I know it was changed by someone else previously)
Don't panic. These are just mere possiblities. No, ditching these boards and replacing them with a third party board isnt necessarily the way to go. You'd be throwing away perfectly good boards, and you will have left all of these questions to the unkown, and where would that get anybody? It's all good.
Assuming good power, let's check for chips pulling down the address and data bus. Here's how to do that:
Remove ROMS/RAMS/5101 from MPU board. Remove as many PIAs from the driver board as possible (if they're not in sockets just leave them for now). Disconnect the driver board completely. Run the 01a-leds test, watch the flash rate of the LEDs and take note of how steady the switch from up to down is. Watch for a stutter in this pattern. If the CPU is crashing at all and immediately restarting, the flash rate of the LEDs will jump for a moment and continue again. You should be able to run it without any stuttering for at least 5 minutes. If there is no stutter and you believe it could run forever, then power down and re-insert a ROM chip. There should be no change. Power down again and Continue (one at a time) with the rest of the chips you've removed from the MPU board. Eventually you should find which chip is causing the crashes, if there is one. When you are done with the chips on the MPU board, connect the driver board (keeping as many chips removed as possible) and try again. Again don't worry about the false sound being triggered during a crash. Keep adding chips back to the driver board and run the test. (of course, always power down between tests).
If, instead of stuttering, the LEDs stop moving altogether, it's still a crash problem.
We need to get you to a place where you are confident on the power, and confident on the MPU board. After that we can deal with driver board issues, and game booting etc. For now, lets get a steady MPU.
Just remembered system 4 could still have tranceivers (IC9, IC10). Those could be the culprit too, and, please (anyone) chime in and correct me if I'm wrong, but I believe they can be removed in system 4 and jumpered, just like they did in system 6a. Is that correct?
Also, I'm not the best for troubleshooting power issues without a scope or a logic analyzer. Does anyone care to provide any tips on verifying good power with limited tools?