Embedded Files
Commodore 64C Restoration - Part 2
March 4, 2026
In the previous part of the series we took one of the barn stored Commodore 64C units and gave it a thorough cleaning, then went about placing some of the missing keycaps and springs. In this part we stare into the void that is the black screen we get when powering the system on.
There are a few components which take part in the video signal for the Commodore 64 including the on board crystal, the MOS8701 clock generator IC, the VIC-II and then the signal routes through the RF modulator.
On the 250469 short board, the 8562 version of the VIC-II only requires 5VDC with VCC on Pin 40, and VDD on Pin 13. which comes in on Pin 40. The earlier 6567 versions of the VIC-II required both a 12VDC and 5VDC input.
Checking the voltages on VCC, VDD and confirming VSS all looked good, so the IC was getting the power it needed.
Knowing that there is power to the VIC-II, the question then becomes do we have a clock for the generation of the video signal? Actually the VIC-II requires two different clock signals be present. On Pin 21 there is input for the Color Clock, and Pin 22 brings the Dot clock into the IC. The Dot clock is crucial for there to be a display, whereas I expect without a Color clock the system would just be monochrome.
While checking for the Dot clock on Pin 22 with an oscilloscope I can see that there is no clocking signal coming into the VIC-II. Sliding down to Pin 21 and I can see there is no signal there either. So the black screen makes sense, as without the VIC-II receiving a clock signal it cannot possibly generate an image.
On the 250469 short board, the clock signals for the VIC-II come from U20 which is the MOS8701 IC. This IC requires 5V on pins 12 & 15, and I confirmed that the voltages are present. The IC takes a crystal input on pin 14, which in the case of NTSC is at 14.31818 MHz. From here the IC divides the frequency to produce the Dot clock at 8.1818 MHz and it also outputs the Color clock at 14.31818 MHz both of which go to the VIC-II. There is a trimmer on the circuit which can be used to fine tune the input frequency.
The frequency being output was pretty close to the NTSC standard but slightly off so I tuned the trimmer and was able to dial it in a little closer. I checked the clock outputs on the MOS8701 but there still wasn't anything coming out of it. My best guess was that the 8701 had an internal failure. This is a custom IC which is now only available through second hand markets such as Bay and occasionally also on AliExpress. I found some online, but.....
Yes I could have certainly found and purchased a replacement 8701, but it seemed more fun to see if I could build a replacement, and someone had already done fantastic work in designing a solution. Over at Lemmini.de, Volker Oth has an article which outlines the functions of the 8701 and their replacement project. The design revolves around the ICS525-01 which is a User Configurable Clock Generator, datasheet here. There are a few variations depending upon if you are working with a longboard or shortboard, and if you are replacing for an NTSC or PAL system, and whether you want to use the onboard crystal or not.
Unfortunately the ICS525-01 is also becoming hard to find, and is not longer stocked at retailers like Digikey or Mouser. While newer clock generation ICs have been released, many of these require a software stack to program initially, and in some cases require a microcontroller to be in the circuit design. In the future I may see if I can put together a replacement solution using a newer IC that is as elegantly simple as Volker's design.
Volker Oth's MOS8701 replacement circuit using the ICS525-01
Assembled MOS8701 replacement
Using Volker's files I produced a few PCBs, ordered the components and assembled the replacement, all for a total BOM cost of about $4 CAD. I installed the replacement and powered up the C64 and saw an image of the Basic screen. It wasn't perfect, there seemed to be some static like interference scrolling through the screen, and the colour would shift out to monochrome and come back in. I checked the Color clock reaching the VIC-II and the frequency was solid, so the issue was somewhere else. That would need to wait for Part 3 - Chroma Coaster.
Report abuse