Jerry returns - The story of reviving a Commodore 64
Spontaneous meeting
The other day I was up in the attic, looking for something, and while moving boxes around I caught sight of two Commodore 64 machines stacked on top of each other. One was in its box, the other resting in a plastic bag. I remembered that I had received them from an acquaintance more than ten years ago, who no longer needed them. I took them up to the attic, thinking I’d do something with them someday.
At the time I didn’t have the inclination to deal with it, but ten years later the inspiration finally came. Not at the exact moment I found them, but later - because thoughts like these tend to mature gradually in one’s mind. I decided to try to bring them back to life and show my 13-year-old daughter what kind of computer her father used to play on when he was her age.
I immediately gave them names: Tom and Jerry - and with that, a (well, actually two) C64 restoration project began. This is the story I’d like to share with you.
-
Warning! The experiments and repair methods described in this article are based on personal experience. If you choose to try them yourself, you do so entirely at your own risk. The author assumes no responsibility for any damage or malfunction that may occur.
Status assessment
Two Commodore 64C units - the latest revision - were hidden in the package, both of them sticky, dusty, and yellowed. Tom seemed to be in much better condition than Jerry. Their previous owners had covered them with stickers, which had already fallen off; I assume they once belonged to children.
There was even a price tag stuck under Jerry’s function keys, which must have been there for a long time, as its silhouette had been burned into the case by sunlight. Jerry’s anti-slip feet had partially melted, and there were some minor scratches as well - fortunately mostly on the back side. The attic climate over the years probably didn’t do them any favors either.
The first problem that immediately stood out was that neither of them had a power supply, so at this point the chances of even a basic test power-on dropped to zero.
Tom hadn’t been affected by sunlight nearly as much - his keyboard and case weren’t really yellowed, unlike Jerry, who had been thoroughly yellowed everywhere by UV exposure. Some people say that the yellowed color is part of a retro computer’s “makeup.” I can accept that, but when I look at Tom and Jerry, I clearly prefer Tom’s color.
That said, it doesn’t really change the overall judgment.
 |
| Tom |
 |
| Jerry |
After opening the cases, both machines were filled with a significant amount of dust, lint, and dirt - both in the keyboards and on the PCBs. In Tom’s case, the metal shielding that covers the electronics was missing. Jerry, on the other hand, still had it, but it was heavily tarnished and even showed small spots of rust. The same was true for the RF modulator’s casing. Interestingly, Tom’s was still shiny, which is probably related to how it had been stored.
At first glance, I didn’t notice any issues with the keyboards. However, later - when I disassembled them - I realized that four of Tom’s keys had collapsed; their return springs were missing. The SHIFT LOCK key switch had also been soldered before, and it was clear that the keyboard assembly itself had already been taken apart at least once.
In contrast, Jerry’s keyboard had never been disassembled. I concluded this from the intact factory adhesive tape used to secure the connector wires - tape that would necessarily have to be damaged if someone wanted to unscrew the assembly, since a few screws are hidden underneath it. During disassembly, it also became clear that Tom’s and Jerry’s keyboards were not from the same variant: the profile of the keycap mounts differed, meaning they are not interchangeable.
Examining the PCBs made the reality fairly clear: Tom had most likely been used as a donor machine, as the DRAM, CHARACTER ROM, and KERNAL/BASIC ROM integrated circuits were all missing. Honestly, I was surprised that the SID, VIC, and MPU chips were still there.
In Jerry’s case, fortunately all the ICs were present, with one notable exception: instead of the original Commodore chip, a replacement 27128 EPROM had been installed in the KERNAL/BASIC ROM socket. This could have several explanations, but I suspect it was used to replace a faulty chip. This is further supported by the fact that both C64s had probably already undergone major repairs following a power supply failure - most of the key ICs were socketed, and signs of re-soldering were clearly visible.
I think it’s quite fitting that the original Commodore power supply is often referred to as the “death brick.”
The boards themselves were heavily contaminated with dirt, but - crucially, in my opinion - they were free from corrosion and physical damage. The edge connectors were worn, but that clearly came from regular use. The earlier IC desoldering work had been done neatly, with no lifted traces on the PCB. However, the re-soldering was not particularly clean - likely due to improper technique or simply rushing the job. Interestingly, this was more characteristic of Tom, while in Jerry’s case the work appeared to have been done more professionally.
Clean-up
I decided that before attempting to bring the electronics back to life, I would first thoroughly clean the main components: the case, the keyboard, and the PCB.
Cleaning the case
Cleaning the case turned out to be one of the simplest parts of the restoration process. The casing is made of high-quality plastic, and the original badges and stickers are very durable and resilient. The only thing I had to be careful about was not breaking or snapping anything during cleaning.
First, I dusted the cases off with a brush, then thoroughly scrubbed them inside and out in lukewarm soapy water using a nail brush. After washing, I dried the parts with paper towels, and finally went over everything with isopropyl alcohol - especially the areas that had previously been covered with sticky labels.
I think the result turned out very nice. After cleaning, Tom looks almost like new, while Jerry remained yellowed, but is now completely clean.
 |
| Before (Tom) |
 |
| After (Tom) |
 |
| Before (Tom) |
 |
| After (Tom) |
 |
| Before (Jerry) |
 |
| After (Jerry) |
I also did some research: the yellowing of plastic parts is caused by bromine added to the ABS material during manufacturing, which turns yellow under UV exposure. There is a method called Retrobright that can restore the color of old yellowed plastics, but opinions on its effectiveness are divided. Some say the yellowing can return after a few years, while others claim it can damage the plastic. I decided to spare Jerry from this treatment.
Cleaning the keyboard
The C64 keyboard is well-built and durable, and fortunately it is relatively easy to disassemble and reassemble. I first thoroughly dusted it with a brush, then removed all the keycaps. Each key has a spring underneath it. As mentioned earlier, unfortunately four springs were missing from Tom’s keyboard. Some of the springs also showed signs of rust, likely due to moisture exposure.
I cleaned the keycaps thoroughly in lukewarm soapy water using a nail brush, then let them dry. Since each keyboard has 66 keys, this process was a bit time-consuming - and by the end, my hands were definitely feeling it. Still, it was absolutely worth doing.
Before reassembly, I cleaned the keyboard base using isopropyl alcohol with paper towels and cotton swabs. The springs were lightly sprayed with a small amount of instrument oil, then the keycaps were snapped back into place. I ordered replacements for Tom’s four missing springs from AliExpress. Unfortunately, exact matches weren’t available, but I found very similar ones. With that, the cleaning of both keyboards was complete.
Cleaning the PCBs
Since the PCBs were dirtier than acceptable, I decided that - even though it was a somewhat risky operation - I would wash them as well. Using lukewarm soapy water and a combination of nail and toothbrushes, I carefully but thoroughly cleaned the boards, paying special attention to the connectors. This was followed by rinsing.
First, I rinsed them with tap water, then with distilled water. The purpose of the distilled water rinse is to remove any salts and minerals present in tap water from the board. After that, I rinsed the entire PCB surface once more with isopropyl alcohol. The role of the alcohol is to displace water due to its strong water-repellent properties. Water can especially cause problems in tight or hard-to-reach areas.
 |
| Jerry is taking a bath |
 |
| Tom after bath and IPA rinse |
After rinsing, I dried the PCBs with a hair dryer, then set them aside for a full day to dry. This last step is important and should not be skipped. The drying time must be allowed before powering them up, as it’s essential that all moisture evaporates from every tiny crevice. Experience shows that hot-air drying alone is not sufficient for this.
The next day, I gave them another light wipe with isopropyl alcohol. I’m hopeful that this cleaning process won’t develop any issues over time. By the end of the process, the boards looked almost like new, truly a pleasure to look at. I believe it was absolutely worth doing; in other words, the boys deserved a proper bath.
Revival step by step
After carefully cleaning and drying the PCBs, the moment finally came to try bringing the machines back to life, the most exciting part of the whole process.
I asked myself what the absolute minimum requirements are, from an electronics perspective, for a C64 to boot and display the startup screen on the monitor. This is what I came up with:
- The power rail must not be shorted, neither on the passive nor the active side (e.g., due to a bad solder joint or faulty component)
- A functioning clock generator
- A working RESET circuit
- And the following integrated circuits: VIC, MPU, Memory Controller (PLA), CHARACTER ROM*, BASIC/KERNAL ROM, DRAM
If the above circuit elements are all working properly, then a C64 should boot. Without them, it won’t do anything at all. Reaching this point essentially means victory.
(* Without a CHARACTER ROM, the C64 will probably boot, but you will see a flickering mess on the monitor.)
I performed the power-up using a laboratory power supply. This has two major advantages: the maximum current draw can be limited, and the supply continuously measures the actual current and power consumption, which provides valuable diagnostic information. For proper operation, the Commodore 64 requires two supply voltages: 5V DC and 9V AC. Since, among the onboard components, only the SID chip requires the 9V supply, and the SID is not necessary for the machine to start, it was sufficient to provide just the 5V DC for initial testing. To do this, I soldered a wire to the input of the power switch and connected its other end to the positive terminal of the power supply.
I started the revival attempt with Tom. Before applying power, however, I inspected both sides of the PCB using a magnifying glass and a microscope, looking for faults and issues such as tiny shorts between IC pins or barely visible damage to the traces. During this inspection, I noticed the rusted top of capacitor C66, a 470 μF / 35V electrolytic capacitor - likely caused by leaking electrolyte. I desoldered it and measured its capacitance, which turned out to be only 47 μF instead of 470 μF, so I replaced it. I also found tiny solder droplets left over from previous repairs, which I carefully removed with a needle.
Before the first power-up, I measured the resistance between +5V and GND on the board while it was powered off. I got a reading of about 9.2 ohms, which initially confused me, but after checking the schematic, I realized this was due to resistor R19, a 10-ohm resistor. Its role is to discharge the stored charge in the capacitors when the machine is powered off, ensuring the system becomes de-energized as quickly as possible - a very useful design choice, in my opinion.
After switching on the main power, the resistance increased to around 850 ohms, indicating that there was no short circuit, and that the PCB could be safely powered.
 |
| The faulty C66 capacitor |
 |
| The R19 resistor |
Before the first power-on, I removed all ICs from their sockets. I powered up the board while monitoring Tom’s current draw. It was around 150 mA, which can be considered perfectly normal.
The second step of the revival process was to verify the operation of the clock generator. I inserted the 8701L6 IC into its socket and powered up the board again. Using an oscilloscope, I measured the signals on pins 6 and 8 of the IC. If the clock generator is functioning correctly, the COLOR clock should appear on pin 8, and the DOT clock on pin 6. In a PAL video system Commodore 64, the COLOR clock runs at 17.734472 MHz, while the DOT clock is 7.88 MHz. The signals were present and the frequencies matched, which meant the clock generator was working. So far, so good. Time to move on.
 |
| Testing the clock generator |
From this point on, each step of the revival consisted of inserting one more IC. I would insert a chip, power on the board, and observe how much the current consumption increased. At the same time, I checked each IC by touch to see how much it was heating up. I started with the VIC chip, since it generates the clock for the 8500 MPU. At this stage, I used the oscilloscope to probe pin 1 of the MPU to confirm the presence of the 985 kHz (PAL) clock signal - this checked out.
Next came the MPU, the two DRAM chips, the CHARACTER and KERNAL/BASIC ROMs, and finally the two CIA chips, one after the other. Since Tom was missing several ICs, I “borrowed” the ROMs from Jerry, and replaced the DRAM chips from my own stock. In every case, the current consumption increased slightly as each new IC was added, but always within a normal range. I deliberately left out the SID chip, since - as mentioned earlier - the C64 should boot without it. Of course, during the entire process, a monitor was connected to the composite video output.
Unfortunately, Tom did not come to life. ☹️
I tried probing signals with the oscilloscope. Some were present, there was activity on the address and data buses, and the MPU was receiving its 985 kHz clock, yet the system still wouldn’t start. At this point, I set Tom aside and moved on to Jerry.
I followed exactly the same revival procedure with Jerry. However, to my great disappointment, Jerry didn’t start either. I began to fear that the 64-pin 251715-01 memory controller chip had failed because if that was the case, it would be a serious problem. I had spare MPU, VIC, CIA, SID, DRAM, and ROM chips but no replacement memory controller.
After a long series of persistent attempts and measurements, I suddenly noticed something: if I pressed down firmly on the left side of the VIC IC in its socket, the missing signals would suddenly appear on the oscilloscope. Bingo - got it! The culprit was the leaf-spring IC socket.
I desoldered it and replaced it with a precision socket, and voilà, like magic, JERRY CAME BACK TO LIFE! The long-awaited blue C64 BASIC startup screen appeared on the monitor. I probably don’t need to explain how satisfying that moment is after putting so much work into such a project.
 |
| The desoldered place of the VIC chip |
 | ||
One of the first signs of happiness
|
There were still a few minor things to take care of. For example, the power switch had a contact issue. It was re-soldered and treated with contact spray. I also reflowed a few IC pins using flux and fresh solder to prevent future contact problems. The ultimate, perfect repair would of course be to desolder and replace all existing leaf-spring IC sockets with precision ones, but unfortunately I don’t have a proper desoldering station. To be honest, I don’t dare attempt such a large-scale operation without one, as I don’t want to risk damaging the PCB.
I did take on replacing the 40-pin VIC socket, which I managed to do cleanly without damaging the board but it was quite a struggle and took a long time. If I get a desoldering station in the future, I plan to replace all the low-quality IC sockets on both Tom’s and Jerry’s boards with precision ones properly finishing the restoration, and hopefully even bringing Tom back to life.
Along the way, it also became clear that Tom’s keyboard is unfortunately completely unusable. The keys are totally unresponsive, which explains the earlier signs of disassembly - someone had already tried to repair it once. This will also need attention when Tom’s revival comes back into focus.
Testing the SID
I left the SID for last. Naturally, the first step was to check the presence of the 9V DC supply voltage on pin 28 of the IC before inserting it into its socket. The Commodore 64C already uses chips manufactured with HMOS technology, which consume significantly less power than their NMOS counterparts. As a result, the HMOS-based 8580 SID chip’s Vdd supply line only requires 9 volts instead of 12, while the Vcc line still provides 5V for the digital sections.
 |
| The SID chip |
To test the SID, I used the following BASIC program:
POKE 54296,15
POKE 54273,32
POKE 54272,0
POKE 54277,0
POKE 54278,255
POKE 54276,17
POKE 54276,16
RUN
Summary
In total, about one week passed between moving Tom and Jerry from attic “storage” to workshop “storage” and bringing Jerry back to life which, I think, exceeded all my expectations, especially considering I mostly worked on it in my spare time during the evenings. If someone had told me a week earlier that I would revive a Commodore 64 in the coming weeks, I probably would have laughed. In my view, this is simply a manifestation of spontaneity, something that quietly shapes our lives. But most importantly, it feels great for someone like me.
 |
It also helped a lot that Jerry didn’t have any serious issues, replacing a single IC socket was enough to bring it back to life. I’m really happy that I managed to save a forgotten Commodore 64, and I’m hopeful that Tom will soon be up and running as well. These machines remind people of our generation where we started from.
It should also be taken into account that Jerry’s hardware has not yet been fully tested. Just because it has video, sound, and a working keyboard does not necessarily mean that everything is functioning properly. The peripheral ports still need to be tested as well, such as the USER, CONTROL, CASSETTE, SERIAL, and JOYSTICK ports, not to mention thorough memory testing. Proper peripherals will be required to carry out these tests.
 |
| Functional status diagram |
That said, I now have a working C64 where I can write and run BASIC programs, and for now, that’s about as far as the possibilities go. The next question arises: how can I load and run programs, mostly games? I don’t have a Datasette, a Floppy Disk Drive (1541), or any kind of cartridge to take things further. But that’s exactly where this new adventure continues. The next challenge is to find a creative way to load and run programs on this machine.
Stay tuned for the continuation...
Comments
Post a Comment