Whirlwind Pinball Repair - Crackling/Missing Sound Issue
Author: Wayne Eggert
Crackling and/or intermittent sound on a Williams Whirlwind Pinball. The MPU board had battery corrosion damage which had already been addressed in a previous repair. Several 6821 PIA chips directly under the battery holder were replaced as part of that repair. Sometimes the sound would work fine, sometimes it would break up slightly, sometimes it was entirley missing or went dead during the game. Not all sounds were affected, but speech/sound effects like the thunder at the beginning, the cellar kickout voice, etc were exhibiting these issues. Over the course of several months the game either worked perfectly or sound effects would crackle or not work at all.
Intermittent issues are some of the hardest to diagnose, it's often easier when something is either working or not-working. With something like this I usually jiggle any cables or connectors that might be part of the problem to see if it might be an issue with the wiring or a broken solder joint. I also push down slightly on some of the socketed chips in the game. After jiggling some things to try to reproduce the sound issues I was still coming up short -- just when I thought I found a wire bundle or connector that caused sound to go in or out the sound would start working perfectly again regardless of how much I jiggled anything.
I let it go for a few months as I had other things to repair. The blower fan in the machine was also having intermittent issues but I assumed it was all part of the same issue with the sound since from what I understand the blower gets its signals from the sound hardware.
Finally decided to revisit the issue recently and found the culprit.. it was U24 on the MPU board, which is a 6802 CPU for the sound. Pushing up from the right-hand side of the bottom of the chip caused the sound to mess up. As I would poke or jiggle connectors around on the board, it must have flexed the PCB enough at times to allow the bad socket to break the connection.
Here's a Youtube video of the issue:
Picture: Sound CPU (6802) with Bad Socket; Arrow points to location of chip I pushed up on
I didn't see any corrosion around this area, but determined the socket most likely needs to be replaced. I pulled the 6820 chip to look at the socket and legs of the chip further.
Picture: Closeup of Socket
It's hard to tell but in the picture above there looks like some dirt/junk in the socket.. or bad pins. Won't really know for sure until we pull the plastic off.
I usually cut the plastic socket in a few places with a wire cutter and then grab those pieces and pull them up over the pins. You don't want to pull with too much force, if it's being really stubborn better to cut it into even smaller pieces. What this does (hopefully) is give you something to grip pliers around as you heat up each pin from the back of the board and pull it out with pliers from the front.
Picture: Socket Pins
The pins didn't look too bad actually. Although a few seemed brittle and broke right off. Anyway, these will all need to come out. Unsoldering 40 pins is a pain if it's your first time, but after doing these enough times you get much quicker.
Picture: Unsoldered Pins
The picture above shows the pins unsoldered from the PCB. Thru-holes were cleared using a desoldering gun. I usually do some continuity checks at this point and visually look for any pads that might have lifted. Unfortunatley found one pad that had lifted..
Picture: Lifted Pad; If you look closely you can see almost 3/4 of the circle was lifted
Sometimes you'll lift a pad. Usually this happens if you're using too high of a temperature soldering iron or don't have the solder heated up enough when you pull the pin. Sometimes it seems to happen no matter how careful you are. In any case, this is why it's good to visually check the pads & also use continuity test. Better to know if you have broken connections now than after the socket's all back together.
There's several ways I could have fixed this -- either "stitch" in some wire from the back of the board to the front & solder onto what's left of the trace, or use a jumper wire. I chose a jumper wire since it's not a very big trace to solder too & then I can visually see the connection.
If you go the way of the jumper wire you'll need to identify where the trace goes. You DON'T want to connect to the wrong chip or you will either have wierd game behavior or possibly short a connection and kill some chips in the process. I sanded the trace slightly at the thru-hole and then checked continuity between it & the chip where the trace lead to and found the other solder pad I needed to jumper to.
Picture: Jumper Wire To Fix Lifted Pad
I didn't get pictures of it but like many other people I usually use machine pin socket strips here since you can cut them at length. So there's new machine pin sockets on the front of the board and the picture above is showing the jumper wire used to fix the connection on the lifted pad.
I put the 6802 IC in the new socket and tested the game and so far sound has been working just fine!
No comments have yet been made.