Embedded Files
Commodore 64C Restoration - Part 4
May 10, 2026
Here in Part 4 of the Commodore 64C Restoration we are going to continue getting this barn stored system in tip-top shape. If you missed the previous posts, we have already extensively cleaned the unit, replaced a failed MOS8701, and replaced the RF stage to solve a video issue. The system turns on, and gives a display however while running the C-64 Diagnostics a few issues were encountered.
Despite all of the test harness cables being connected, the system was not detecting the keyboard, it reported an issue with the Control Port and also suggested that the SID 6581 was bad. I can confirm that when typing on the Basic screen the keyboard registered some keypresses correctly, but many were misinterpreted. So why would the keyboard, control port and sound chip all show as being bad?
If the CPU is the brain of the Commodore 64C, the CIA 1 (Complex Interface Adapter, MOS 6526) is its nervous system. When this single chip goes bad, it doesn't just fail quietly—it throws an absolute tantrum for diagnostics. During my initial testing, the diagnostic cartridge flagged failures on the keyboard, both control ports, and even the SID sound chip. At first glance, you might think you are dealing with three or four separate, hardware failures.
To understand why a single chip causes such widespread chaos, we have to look at how the keyboard actually talks to the computer. The C64’s keyboard doesn’t have individual lines for every single key. Instead, it relies on an 8x8 matrix grid. The CIA 1 chip is directly responsible for scanning this grid using its Port A and Port B pins.
Because the game controller ports share these exact same input and output lines on the chip, any failure inside the CIA 1’s registers means the computer can lose its ability to "see" key presses or joystick directions. When the gatekeeper goes offline, the entire communication line collapses, making the diagnostic cartridge assume both the keyboard and the joystick ports are physically broken.
This sharing of resources is also why the SID chip gets caught in the crossfire, which is a detail that trips up a lot of hobbyists. The SID chip handles more than just synthesizer music; it also contains the analog-to-digital converters used to read paddle controllers. The pathway consists of the Control Ports to a CD4066 Bilateral Switch to the SID.
If the CIA 1 is fried the SID chip never receives the data it needs to process the paddle inputs, leading the diagnostic cartridge to report that the SID itself has kicked the bucket.
The original CIA was soldered directly to the board so it needed to be removed first, and then a socket put in for ease of troubleshooting / later repair. I then installed a known good working CIA into the socket and re-ran the diagnostics. This time, everything came back clean proving that the failure in the CIA was responsible for all of the errors showing in the original test.
A failed CIA isn't a death sentence, but it isn't great news either. The CIA was a proprietary IC, and while there are a couple of projects that have developed over time for modern replacements there are still challenges. The 'plug-n-play' option is the J-CIA which appears to be a drop in replacement that has been reported by some in the community to be quite good. The main drawback at this point is pricing which is over $100 CAD delivered, a hefty investment for replacing a single IC.
Another project which is growing is the reDIP by daglem on GitHub. This one is open source and available to be constructed by anyone though it's function as a 6526 replacement is not as thoroughly tested as the J-CIA. Also the assembly will require some skill as there are some pretty tiny components.
While considering producing some boards for the reDIP I did some searching on eBay and I was able to find a seller in Hungary who had a stock of 6526 ICs available at a reasonable landed cost. I had already built replacements for the MOS8701, and the RF modulator for this machine so I took the shortcut and purchased the 6526 CIA rather than build.
But the reDIP remains an interesting project for another day!
I installed the replacement CIA when it arrived and confirmed through diagnostics that everything was working as it should. We are approaching the end of this series with only a few finishing touches to be covered in the final part!
Report abuse