I recently bought a set of Robotron boards complete with the Clay Cowgill Multigame, clearly described by the seller as not knowing if they were working or not. I assumed that they wouldn't work but was very happy for the bargain price I paid. Boards duly arrived in good condition. They had a sticker on them showing they'd been to Leopardcats back in 2004, so I was happy that any repair done then would have been of high quality. And the boards showed no signs of nasty hacks.
Rather than powering the whole thing up on my test bench (kitchen worktop), I started off just with the CPU board. Not good news, the switcher shut down immediately power was applied. Checked for shorts, none. So removed all socketed chips (CPU, 2 x decoders & 24 RAM chips). Re-applied power and it stayed up, which was good news, showed one of the removed chips was faulty. Re-fitted CPU and decoders, still happy. Then RAM chips 4 at a time (to narrow down the fault), all ok until the last 4 at which point the switcher shut down again. Removed the final 4 one at a time, and bingo, found a faulty one, replaced it and powered up, the switcher stayed up.
Plugged in my trusty Joust ROM board (didn't want to risk killing the multigame at this stage) but got an odd rug pattern and the dreaded 131 error. Although good news that the board was trying to run.
131 indicates a RAM fault, but isn't necessarily the RAM chip that is faulty. So checked the pins of the RAM indicated by the fault code. Nothing on the Do output pin. Not high, not low. Nothing. Checked the other RAM chips. The same on all 23 (the one I replaced earlier was showing pulsing output). So all the RAM chips are faulty, must have been something nasty hit the board, lack of -5V is usually guaranteed to kill all the 4116's. Replaced all the faulty chips, now getting an output but still the 131 error (but I don't believe that RAM chip is faulty).
Next up, check the other pins on the RAM. All looked ok except pins 5,6 & 7 (A0, A1 & A2) which were either stuck low or nothing at all. All the address lines are fed from 4 x 74153 multiplexers, which are known to regularly go wrong. Checked the relevant outputs and sure enough, 2 of the IC's were giving the wrong outputs. One had already been swapped out in a previous repair and was socketed, so pulled it and tested in my tester - it showed to be faulty
Swapped this one out, and desoldered and swapped out the other one that was showing wrong outputs. Fired the game up and it works!
Final step, connect up the multigame.
If this doesn't work then there's not a lot of troubleshooting I can do as it's FPGA-based, so fingers crossed...
Needn't have worried, it works! Likelihood is that the Dallas NVRAM module will need replacing as it has a date code of 2006 and the battery is likely flat, but I'm very happy now it's working. Another Williams board brought back to life.
Rather than powering the whole thing up on my test bench (kitchen worktop), I started off just with the CPU board. Not good news, the switcher shut down immediately power was applied. Checked for shorts, none. So removed all socketed chips (CPU, 2 x decoders & 24 RAM chips). Re-applied power and it stayed up, which was good news, showed one of the removed chips was faulty. Re-fitted CPU and decoders, still happy. Then RAM chips 4 at a time (to narrow down the fault), all ok until the last 4 at which point the switcher shut down again. Removed the final 4 one at a time, and bingo, found a faulty one, replaced it and powered up, the switcher stayed up.
Plugged in my trusty Joust ROM board (didn't want to risk killing the multigame at this stage) but got an odd rug pattern and the dreaded 131 error. Although good news that the board was trying to run.
131 indicates a RAM fault, but isn't necessarily the RAM chip that is faulty. So checked the pins of the RAM indicated by the fault code. Nothing on the Do output pin. Not high, not low. Nothing. Checked the other RAM chips. The same on all 23 (the one I replaced earlier was showing pulsing output). So all the RAM chips are faulty, must have been something nasty hit the board, lack of -5V is usually guaranteed to kill all the 4116's. Replaced all the faulty chips, now getting an output but still the 131 error (but I don't believe that RAM chip is faulty).
Next up, check the other pins on the RAM. All looked ok except pins 5,6 & 7 (A0, A1 & A2) which were either stuck low or nothing at all. All the address lines are fed from 4 x 74153 multiplexers, which are known to regularly go wrong. Checked the relevant outputs and sure enough, 2 of the IC's were giving the wrong outputs. One had already been swapped out in a previous repair and was socketed, so pulled it and tested in my tester - it showed to be faulty
Swapped this one out, and desoldered and swapped out the other one that was showing wrong outputs. Fired the game up and it works!
Final step, connect up the multigame.
If this doesn't work then there's not a lot of troubleshooting I can do as it's FPGA-based, so fingers crossed...
Needn't have worried, it works! Likelihood is that the Dallas NVRAM module will need replacing as it has a date code of 2006 and the battery is likely flat, but I'm very happy now it's working. Another Williams board brought back to life.
