Here's some pictures of our 1980 Williams Firepower pinball machine:
Humble beginnings:
Just got it home! The playfield...
...the backglass...
...and the backbox.
Finally working! The finished product.
The working playfield...
...and the backbox.
Ever wonder what's under the playfield?
The backbox, without the backglass.
The brains behind the displays...
The solenoid driver board
The MPU board
The sound board
The power supply
The story behind the machine, or how NOT to repair a broken pinball machine:
While looking through the classified ads one day, I stumbled across an ad for the machine. It turned out to be in partly working condition, and after some quick web research, I decided it could be saved...
After dragging it home (if you take it apart, it will fit in the back of a Pathfinder...barely!) and reassembling it, I started to assess the damage.
The previous owner said that it was working fine until one day when some of the solenoids stuck 'on', and then it blew a fuse after a few minutes. He had unplugged one of the wiring plugs from the driver board, and that stopped the problem solenoids from kicking in. With that plug still undone, and the glass off, everything seemed to be working fine except that some of the solenoids were not working. Assuming (!) that this was just due to the unplugged wiring, I went on to play with the machine for a game. After the game was over, the displays went blank...and it wouldn't do anything at all. Great...I worked on it, and now it's even worse than when I started.
After posting some questions to rec.games.pinball, I got some great diagnostic advice and info, and dove in to see what had gone wrong. The driver board had two dead 2Nxxx transistors, and one of the 7408 chips was giving weird readings on the ohmmeter. I replaced the transistors and the 7408.
The MPU board is supposed to blink some LEDs on the board to indicate what (if anything) is wrong with the board when you press a special diagnostic switch. I was just getting two solid 'on' LEDs, which meant it wasn't doing *anything* at all. One of the first things that the MPU board does on reset is to turn those LEDs off. It wasn't doing that. It could have been the PROM chips had gone bad and the CPU was executing nonsense, or it could be just an I/O problem.
One common problem for Williams boards of this vintage is bad sockets on the PROM and CPU chips. I replaced them hoping that would bring it back to life, but no luck.
I tested the address and chip select lines with an oscilloscope, and found some believable activity there. I ordered some parts from Jameco, and waited. I had tried powering up the board without anything but the power connected, without any change in behavior. That is supposed to isolate the problem to the MPU board if that happens.
The only thing between the CPU and those LEDs is a 6820 PIA chip. Either the CPU wasn't able to read the program from the PROM chips, or it wasn't able to talk through the PIA chip. I replaced the PIA chip, and voila! The LEDs started blinking the correct diagnostic codes.
I put it back together, and it booted up and lit the displays again. I tried the repaired driver board, and it seemed to have stopped the 'stuck on' solenoid problem. I played it for a few minutes, and then one of the pop bumpers stuck on...and then the displays started flaking out again...and then it all went blank...again. Back to Jameco for another order. A handful of everything that looked like it could be dead. PIAs, transistors, 74xx chips, more sockets, the works.
Putting another PIA chip on the MPU board didn't help this time. I checked some of the signals on the driver board, and determined that one of the PIA chips there had gone this time. I replaced it, and got it working again, although some of the displays weren't working correctly. Thinking that it had something to do with the solenoid voltage getting shorted through to some of the TTL chips, I finally got a clue and pulled the fuse on the solenoid power circuit. I tried running the machine through some of the self-test modes, and it worked OK (except that the solenoids didn't have any power) I put the fuse back in (MISTAKE!) and did more tests but after a few iterations, it all went dead again. No displays, no bootup, no nothing. Killed it AGAIN!
I went back to the driver board, and the same transistors were shot again. This time I replaced the 2Nxxx transistors, the associated TIP122 transistors, as well as the 7408 and 7402 chips associated with those signals. I also replaced all 3 PIA chips on the driver board. Then I replaced the PIA chip on the MPU board as well. Running the MPU board by itself got me back to the blinking LEDs stage at least.
After killing this thing so many times, I decided it was time to step back and try to find the root cause of the problem. (Now why didn't I think of that sooner!) Each time it had worked fine, I had problems after using the solenoids, even with the wiring unplugged for the 'problem' solenoids. The solenoid firing circuitry includes a diode that is supposed to stop the voltage spikes from killing the rest of the circuitry. I did some checking of the diodes that were installed on the solenoids, and found that several of them were 1N4001 diodes, a much lower power rating than is supposed to be used there. That spike could easily be killing the transistors and shorting them out! I then replaced ALL of the solenoid diodes with the stronger 1N4004 diodes. At this point I noticed that the fuses that were installed on the machine were not the right ones. In particular, the 2.5A solenoid power fuse had been replaced with an 8A one. No wonder things were cooking once there was a short!
After replacing all of the diodes (and all of the other things that got cooked) it finally started to behave again. Yeeha!
The rest of the story is the usual cleaning and adjusting, and of course, lots and lots of play testing the completed machine. I think it needs more testing...
Note: Unless otherwise specified, all text and images on this page are © Copyright 2000, 2001 by Pat Mancuso.