Galaga – PCB Repair Logs

*** Newest Entries always added to the bottom ***

Had some good timing on this one.  I while back an operator I do some work for now and then was planning on sending some Galaga boards, so I built up a bench harness. They never showed. My Galaga decided to stop working and I had picked a couple of boards sets up just to have. But I never really worked on any of them. Fast forward and one came in for repair.

Board #1 – In for repair

The board itself is in nice shape and the sockets are single wipe, but not the really bad Scan Bee sockets. On the right are all the Namco custom chips that are heavily tarnished. I swish them around in Tarn-X to get that off and then clean the legs as usual. 

Once I got it all together, the board would not boot. Not surprising. One thing I do with boards with lots of customs is test them one at a time in known working boards. In this case I did not have a working Galaga. I could have tested in another board, but didn’t at this point.

I set up the CATBOX and did some testing, had not done it before on Galaga. Turns out the architecture has similarities to DigDug. All 3 CPUs share the same RAM address space, so all RAM tested clean from all 3 positions. The program ROM for CPU#1 was good. I’ll have to make configs for CPU#2 and #3. 

Put it all back again, watchdogging.. Now I started the part swap Tetris and moved parts back and forth from my board to this one. Finally I got the PCB to boot. GREAT! Last item was No explosion sound – which is controlled by custom 54xx. I swapped a few of those around and the explosion showed up. Then I put the original back and the explosion stayed. 

Moral of the story – the tarnished chips may need to be inserted/pulled a number of times to clear any small tarnish in the socket. They are pretty decent and not worth replacing.. 

The other factor was I determined two of the Z80’s were legitimately bad.

Testing on the bench – everything worked great. Put it in the cabinet and…

This board sounded great on the bench, but in the cabinet it had a very loud bong sound. Capacitor C21 in the audio amp section tested very different from my reference board. Replaced it to fix the problem. Bench power and the game power was obviously just different enough.

After about 20 hours of run time, one of the 08xx customs died causing the game to crash. Replaced.

Board works!

Board #2 – One of my extra Galaga PCBs

This one turned out to be easy. Same tarnish issue as above. Cleaned all the chips, Deox-it.

Booted but the star field was messed up. Reseated the 05xx custom.

Board works!

Board #3 – Board in for repair

Board initially would not boot. After some testing, I moved and reseated some RAM and got to a normal screen. After removing, cleaning and Deoxit all of the chips – the RAM was still a bit flaky and the game was not reliable.

Replaced all of the RAM sockets and it has been stable ever since.

Put it through a multi day test cycle – no issues.

Board works!

Board #4 – Board in for repair

It happed to be with me at the lake and I did all the board maintenance, cleaning and inspection there. Once I got it on the bench.. it was dead. It was a relatively clean looking set however..

I got into it and there were a number of issues. The Namco customs 54xx, 04xx and 08xx all had weak or broken legs. I had to repair a number of those. There was no clock.. I hunted around for a while and determined there was extensive prior work.. The well meaning repair person replaced all the sockets and some other items, but created a number of issues that had to be individually tracked down..

Here are a couple of solder bridges shorted to other traces.. These were killing the clock – but only when the socketed chips were plugged in..

Here is a lifted trace – there was another elsewhere.. Here is another short. There were MANY of these solder bridges to find and fix one at a time.. Minimum of 10 of these.. Prior work issues take far longer to find than actual chip failures.. Lots of time on this one.

Once I got it all going – there was a color RAM@1C on the video board that was legitimately bad. Lots of burn in hours later..

Board works!

Leave a comment