C64 With Black Screen

[C64Person]

I have good news! I socketed one RAM on the good board so that I could swap in RAMs from the bad board, and I found five dead RAM chips on the bad board! I unfortunately only had four extra RAMs, so I am going to have to order more before I will know if that was the only problem, but at least I got somewhere.

[C64Person]

My new RAM set arrived, so now I know that at least the bad board's RAM are fine. I also swapped all three ROMS and the CPU into the good machine, and these chips are good! I replaced both LS257 multiplexers at U13 and 25 on the bad board, so they are both known good now too. Finally, I checked logic chips U28, 16 and 14 of the bad board as well, and none of those produce a black screen on the good board either.

Given this, there are only 6 more logic chips on the bad board that I could swap. I suppose I will swap these as well, unless there is any better advice.

[rmzalbar]

Wow! five bad RAM. That probably means the C64 has been exposed to a failed power supply that fed it too much voltage on the 5V side (or maybe just a surge/lightning strike event.) The RAM chips are the most sensitive to this; they will usually always die, while the Commodore custom chips are more tolerant of high voltage. So that means that every chip with the exception of the VIC-II saw high voltage. You've now tested everything except the remaining logic chips. Wow you've been through everything on this poor board! Nope.. pretty much down to checking those logic chips.

If you wanted to try to find a bad one without removing them.. they could be anywhere from totally failed (which a logic probe and pulser could tell you) to working logically but weak output (which you could see with oscilloscope)

[C64Person]

Okay, I swapped the last 6 logics, and they all seem fine. Now this becomes harder.

Either the fault is related to something else on the board other than chips, or it is related to one of the 4 smaller logic chips inside the VIC-II area (I can't swap these because the good board is the newer assembly 250425, where these chips were combined into 1). Both my power supplies are working now so I can start comparing signals if you have any ideas where I would start.

[rmzalbar]

All those extra chips in the VIC area do is generate two clock signals, color clock and dot clock, and we did check those early in troubleshooting. They are available on pins 21 and 22 on the VIC, 14.318 and 8.18MHz respectively. If you have clock there at those frequencies, it's working.

First of all, when you tested the RAM, are you running a ram test program of some sort? If you are only swapping the chips into that single socket and testing to see if the machine starts, that may not be a thorough enough test. You can have bad RAM cells that still allows the computer to start if it's not in a memory location that runs the exec and basic, yet when you put them back into the other board they are now plugged into a socket where that bad cell is critical. If you have all new RAM, try it.

Second, if all chips are good in the other machine, then we need to reassure ourselves that all memory and address traces are going where they are supposed to, on all chips and the ram sockets. I know you had to work on the RAM area a lot, but you may need to recheck.

Another thing is, if you plug in a disk drive and power on, does it reset the drive? What if you blindly type in the command to load from a disk, does it start running? Maybe the video output is broken somewhere.

[C64Person]

I tested the RAM in two different sockets on the good board, and all the 5 bad ones showed faults (mostly black screens, a few garbled character screens). All the RAMs in the bad machine are new now, so they are all fine. How do you recommend testing traces? I can check for continuity, but will I be able to follow every trace and see if it's continuous?

Clocks on VIC-II are good, so I'll rule out those chips. I swapped every other chip into the good machine or replaced them, so I am confident that the bad machine's chips are all good now.

I have 2 disk drive boards, but one functioning drive component. I don't know if the boards actually work for two reasons - I have no floppy disks, and I have not had a working C64 to use them with until now. If I connect one of the drive boards to the working drive, it does spin up when I start the computers. I really have no experience with the drives, those were going to be the next project once I got the computer working.

[C64Person]

I fixed it!!! Both C64's are running now! The problem was ultimately two traces that had become shorted to nearby traces during rework: one by the U9 RAM (more U9 problems) and one by a CIA. I reheated the solder and broke those connections, and it's now there's a normal start screen! Unfortunately, it took my new U9 with it, but I found a working extra RAM to replace it with. It looks like my troubles aren't quite through with yet though...

The old machine's keyboard is not working, as some keys are not working unless I really push them hard, and any time I try to run a line, the system crashes. It's not the computer, because I swapped in the new machine's keyboard, and that one has no issues. Not sure if this is going to be an impossible repair, but I figured I'd ask.

I also don't have any programs for it, but I will try to get some cartridges and/or floppies. I do, however, want to make sure my 1541 disk drive works. Will it show any signs of life without a floppy inserted (other than resetting when the computer starts), or will I need to get disks to use it.

Lots of new questions of course, now that I can really start using my C64 for the first time, but it's great to finally have it working after a year and a half of nothing.

Thanks for all of the help!

[C64Person]

Guess I spoke too soon...

I fixed the keyboard; it just needed a deep cleansing after years of dust/grime buildup.

On the old machine, any time I try to use the CTRL color change command, it completely throws off the system. All keys accept letter keys are printing strange characters on the screen after trying to change color (which itself causes strange characters), and the computer is not useable once this happens. I have swapped all the ROMs, CPU, PLA, and CIAs to see if it stops, but no luck. It happens with both keyboards as well, so it is definitely the computer.

[rmzalbar]

Yes, some kind of logic failure is happening there. It would be great to plug in a test cart at this point and let it run through its battery of tests. Are you getting "38911 basic bytes free" or some different number? Can you post a video of the failure sequence? Does any of what you're experiencing match up with the Pictorial Fault Guide? https://derbian.webs.com/c64diag/

The keyboard does respond well to cleaning. If any of the keys are still unreliable after cleaning, remove the plunger and rub its conductive carbon silicone "foot" on a normal piece of paper just a little bit for a few seconds, as if it were a pencil eraser. It will leave a little skid mark. Paper is mildly abrasive and will give you a fresh conductive surface.

The most common disk command issued:

LOAD"*",8,1

Will get the disk drive to attempt loading. That should start the motor and the red activity light, and step the head around a bit as it looks for data. With no disk in the drive, you'll end up with a flashing red light, indicating an error.

This happens to be the usual command for booting a commercial disk with a game or program that uses the entire disk. Here's a breakdown of that load command:

LOAD [tells DOS to load from a data storage device]
"*" [filename in quotes, in this case asterisk as a wildcard which loads the first file on the disk.]
8 [device number on the serial bus. Disk drives start at 8, if you had four drives plugged in they would be numbered 8,9,10,11..]
1 [specifies the data is to be loaded directly into the memory location as specified in the file, as opposed to loading it into BASIC memory. You'd use this for normal games and programs that are not in BASIC. BASIC programs you would leave off the ,1]

After loading, some programs autostart, others require that you type RUN.

LOAD"HELLO",8
Loads a BASIC program named "hello" from the disk. You could then RUN the program or LIST it.

LOAD"$",8 [the $ character loads the disk directory as a special basic program. You then view the directory by typing LIST.]

[C64Person]

I attached 3 pictures. The second shows the CTRL + numbers 1 through 8 when entered into a string variable, but I noticed that the color change works fine when entered before anything else is typed. I noticed that the color change actually went into effect after the line was executed, as seen in the third picture.

I was just looking at the pictorial fault guide, but there are so many pictures that resemble strange characters that its hard to say if any one chip looks guilty.

At one point, the computer started doing things like the first picture with on startup. After removing the keyboard, this stopped, so I wonder if there are seating issues with chips, or if this is a logic failure as you said.

LOAD"*",8,1 originally produced a DEVICE NOT PRESENT ERROR, though now it says SEARCHING FOR * indefinitely. When it says that, there is no response from the drive, and it just says searching forever, unless I disconnect the drive. Pressing stop does not terminate the search.

As seen in the pictures, I get the normal 38911 bytes free.