Disclosure: Note that as an eBay Partner Network Affiliate, I may earn from qualifying purchases.
Introduction
I purchased this nice 250407 Revision C motherboard as part of a lot of boards. It was missing the PLA (U17) and the VIC-II video chip (U19) as shown below. A new GAL-PLA (e.g. from eBay) and Kawari VIC-II replacement (e.g. from eBay) brought this machine back to life. I liked this board so much that I decided to socket all the ICs on the board to be able to use it for troubleshooting other machines by making it easy to swap any chip. The board got flaky and eventually stopped working after that. Desoldering chips is always risky and, long story short, I got in a hurry and damaged a trace that I caught and one that I didn’t and had to troubleshoot. I document these fixes below.
Case Study
As mentioned in the introduction, this board stopped working shortly after I socketed all the ICs (e.g. CIAs, ROMs, RAM, logic chips, multiplexers, etc.) that did not already have sockets. During the desoldering process I noticed a lifted trace at U14 (74LS258). This is shown below by the yellow arrow. You can see that this general area is not representative of my best work. The problem was that I got in a hurry given so much desoldering that I did not keep up with cleaning out my desoldering iron. As a result, the suction was not performing optimally and thus not removing all the solder from the pin holes. Thus, there was more than the usual solder on top of the board which resulted in the lifted trace even thought I used hot air. You can see the residual solder remaining in some of the holes (e.g. pin 1 of U14). Fortunately, it was easy to solder in a bodge wire on the bottom of the board to connect these two pins (see next photo below). That restored connectivity and fixed the damaged trace.
Unfortunately, what I missed was a damaged trace on pin 10 of U26 (74LS373) that eliminated continuity with the ground plane. This was hidden from view (and not obvious anyway) because I installed a socket over it. It took a lot of work to track that down including going pin by pin with my multimeter in continuity mode testing each pin and their connections. I eventually fund the damaged connection (orange arrow above). Fortunately, there was a via with solder in the hole from the manufacturing process that was located right next to that pin under the board. I was able to restore the connection with a solder blob (above to the right). This brought the computer back to life.
Below is the final board with the trace fixes, the new GAL-PLA, and a new Kawari VIC-II replacement.
Comments
This repair was a great lesson in keeping the desoldering iron clean with optimal suction. It was also a good lesson in taking my time. Very happy to have at least one board that is fully socketed for testing purposes. I also recapped this board with new electrolytic capacitors because there were several that looked like they were out of range. In addition, I replaced the RF module with a new RF board from the Retro Channel. This RF replacement board, in combination with the Kawari, provides a nice clean signal (see second image below). This is rarely necessary, but I did it just in case. Note that I tinkered with this board on an off for several weeks before tracking down the bad trace. Took a while!
Note that I later dropped this board into a newly made clear case from Retro Fuzion. The clear case is sharp, but it also makes it easier to take the case apart to remove chips as the screws are in metal guides. Repeatedly assembling and disassembling the original bread bin cases risks cracking the plastic or breaking the standoffs.
