Help reviving a Commodore 64

[Mercenary6502]

HI,

 

I've been in the process of optimizing and de-junking my collection, and that includes repairing my faulty computers. I just got an IBM 5150 revived by rebuilding it's PSU and now it's time to do the same with an original (Rev A) Commodore 64.

 

Background on this computer: It was given to me for free with one of those original epoxy filled PSU's that are prone to fail due to heat. Being inexperienced and unaware that those commonly fail, I went ahead and fired it up. It worked, but only for a couple seconds and the video was very faded. I of course then discovered that the PSU was at fault. Shortly after, I opened and chiseled out the components of the PSU and looked into my options of rebuilding it, but otherwise the computer has been sitting for years without much touching until now.

 

For starters, I needed a new PSU. I happened to still have the original transformer lying around (everything else was damaged from chiseling) and finally got brave enough to test it by wiring a power cord to it. Both outputs looked good (around 9.5VAC) so I went ahead and built a new one around the original transformer with an old-school 7805 regulator (just for the time being at least):

 

 

Needs more optimizing and heat shrink so it looks scary, but I now have the 9VAC and 5VDC voltages I need!

 

Next was the Commodore 64 itself. I fired it up and nothing happened, just an initiation pop of the SID in the TV speaker with no video. Checking voltages I found that it's on board 7805 was garbage so I replaced that and got a step further. I could now get a gray screen by adjusting the pots in the video section of the board. After about an hour of poking around and powering off and on, it suddenly fired up, and after adjusting the pots the video was much brighter and fresher than it's ever looked before!

 

 

 

But of course, that didn't last. It'll go back to black screen (with a single dim vertical color line on the far left of the screen) after being shut off for more than a couple minutes. Poking around some more, I discovered that touching the crystal clock and/or the capacitor near it has a drastic effect on the video when it has the black screen (components marked below):

 

 

It seems like a combination of touching those parts, adjusting the pots, and restarting the computer will eventually get it power up with video like above. I've done it more than once. I've resoldered those components to ensure that there aren't cold joints and also checked the capacitor to find it within tolerance.

 

So this is where I'm at. Does anyone have any clues as to why it's acting like this? I don't have a whole lot of experience with video circuits so I don't really know where to proceed. I understand that the same crystal clock is divided and served to the entire computer, so if there was trouble there it wouldn't even boot. But it does obviously work sometimes. If I had an oscilloscope, I would check for all of those clock frequencies at time of failure but I don't.

 

 

Thanks for your time! (Didn't feel like joining a Commodore community )

[carlsson]

Traditionally, the #1 culprit for occasional black screen on breadbox C64 is the PLA chip, kind of a memory controller. Sometimes the computer will start, most of the time it won't.

 

Perhaps though this is not your fault. Does the speaker say pop every time? You could try entering commands in the blind to see if you get real audio:

 

POKE54296,15

will turn on the volume and causes a click

 

POKE54278,240:POKE54273,10:POKE54276,33

will set up a sustained sawtooth note at a relatively low pitch

 

I understand it can be difficult to know if you're getting all keys right, perhaps check first in a situation where you see what you are typing.

 

Do you have any peripherals like a floppy drive or a game cartridge? Sometimes those can be connected to exclude possible fault sources, but it is a bit crude troubleshooting.

[Keatah]

I am not a C64 specialist. But I can tell you that touching parts like you did would cause all sorts of distortion because your body is changing the reactance and capacitance of a sensitive tuned circuit. That's normal. That's why they are in a shielded box.

[Mercenary6502]

Traditionally, the #1 culprit for occasional black screen on breadbox C64 is the PLA chip, kind of a memory controller. Sometimes the computer will start, most of the time it won't.

 

Perhaps though this is not your fault. Does the speaker say pop every time? You could try entering commands in the blind to see if you get real audio:

 

POKE54296,15

will turn on the volume and causes a click

 

POKE54278,240:POKE54273,10:POKE54276,33

will set up a sustained sawtooth note at a relatively low pitch

 

I understand it can be difficult to know if you're getting all keys right, perhaps check first in a situation where you see what you are typing.

 

Do you have any peripherals like a floppy drive or a game cartridge? Sometimes those can be connected to exclude possible fault sources, but it is a bit crude troubleshooting.

 

Thanks for your response. I tried entering the POKE commands but am not getting anything. I think that keyboard has a few touchy keys and needs to be rebuilt though so it's hard to justify. I do happen to have a working floppy drive which I plugged in. Once the computer is powered on, the floppy drive lights, makes some noise and stops. But I can't seem to blindly get it to read a disk directory or anything.

 

And yes, I do get that pop in the speaker every time the computer is power on.

[R.Cade]

Yeah, sometimes works, sometimes doesn't... #1 is PLA, #2 is ROM, #3 is RAM if the power supply failed high. If it failed low/dead, then probably not.

[Mercenary6502]

I am not a C64 specialist. But I can tell you that touching parts like you did would cause all sorts of distortion because your body is changing the reactance and capacitance of a sensitive tuned circuit. That's normal. That's why they are in a shielded box.

I was wondering about that, that makes total sense.

[Mercenary6502]

Yeah, sometimes works, sometimes doesn't... #1 is PLA, #2 is ROM, #3 is RAM if the power supply failed high. If it failed low/dead, then probably not.

When the power supply failed, it was missing the 5v while the 9v was still accounted for and within tolerance. So no high voltages.

 

The PLA does seem to run hotter than many of the other chips, but maybe it runs hot normally?

[R.Cade]

Correct, it runs pretty warm normally. Not as hot as the VIC-II, but hot to the touch. They can be replaced with modern logic (PLAnkton) which runs cool and 1/10th the power.

[Mercenary6502]

If I had another C64 that worked I would swap the PLA but I don't. Is there a procedure to test it by checking pin levels? I have a logic probe.

[Mercenary6502]

Now that I think about it, a couple times when it was working the characters on the screen would turn into garbage after a certain amount of time and the center of the screen (not the border) would start rolling. Could a faulty PLA cause that too?

[carlsson]

 

 

U17 906114-01 (82S100PLA) PLA

Blank screen, no border. It can produce colored screen or flashing color garbage instead of startup screen. It can cause intermittant loss of cursor, screen freeze and/or program crashes after warmup and it can put random characters on screen. This chip normally runs hot. It is the most common chip to fail in the C64 because it runs hot normally, like the SID.

http://personalpages.tds.net/~rcarlsen/cbm/c64/c64-ic.txt

 

So yes, it could be the cause and you touching components to get it to boot being a red herring.

 

e5frog on this forum has co-developed the PLAnkton chip. I believe it sells for $15.50 + $10 shipping, through his North American business partner eslapion. There are a few other PLA replacements too, at varying prices and abilities.

[Mercenary6502]

http://personalpages.tds.net/~rcarlsen/cbm/c64/c64-ic.txt

 

So yes, it could be the cause and you touching components to get it to boot being a red herring.

 

e5frog on this forum has co-developed the PLAnkton chip. I believe it sells for $15.50 + $10 shipping, through his North American business partner eslapion. There are a few other PLA replacements too, at varying prices and abilities.

 

Alright, sounds like I should start looking into a replacement. The PLAnkton looks cool and is reasonably priced aside from shipping costs. I also just read about the method of using an EPROM (if you have one fast enough) to emulate the PLA. Anybody have takes on that? The PLAnkton does looks like the best option, but I have the resources to do an EPROM PLA and it would be free. I use Atari's more anyway.

[carlsson]

The community is awfully split when it comes to the opinion on the EPROM solution. Some think it is one of the worst things you can do, and you really need that EPROM to be very fast, while other people think it is alright and not very different from how the original 82S100 was designed.

 

There has also been discussions about using off-the-shelf GAL chips, but I've only seen solutions for PET computers featuring such chips.

 

Then there are the SuperPLA, the realPLA and possibly a few more which however may not match the PLAnkton on price (except for shipping, that is).

[Mercenary6502]

Alrighty, I appreciate the help everyone! Made short work of that!

 

That one will probably go back on the shelf for awhile before I get a new PLA. I was seriously considering the EPROM method but decided to do it right by getting something professional and proven like the PLAnkton. Being cheap can cost me a lot of time and I'm starting to see that if it's not very educational, it's best not to waste my time and spend a little more. I did get the power supply reconfigured and safer, so at least I was able to get one step closer to getting that computer going again before putting it back on the shelf for now. One day!

 

By the way, all four of my Atari computers are still running strong on the same original components from the 70's and 80's including the PSU's. Why can't Commodore's do that??

 

 

Thanks again

[Oge]

You removed the epoxy from a Commodore 64 Power Supply... Respect to you

[cjameslv]

You removed the epoxy from a Commodore 64 Power Supply... Respect to you

Yeah true that! When my c64ps died i just got a c128ps and added in the wire from my c64ps and bam it can do both (granted not at the same time).

Edited April 10, 2016 by cjameslv

[Mercenary6502]

You removed the epoxy from a Commodore 64 Power Supply... Respect to you

 

Yeah true that! When my c64ps died i just got a c128ps and added in the wire from my c64ps and bam it can do both (granted not at the same time).

 

If you aren't trying to save anything but the transformer than it isn't too bad. What I ended up doing was drill a small hole through the top of the power supply box and hammered a round chisel into it. That pushed all of the contents out of the box leaving me with one big epoxy brick that I could hammer away at without the box getting in the way. The hole now serves as a place for a nice LED power indicator (right in the center of the Commodore emblem).

 

 

I like to try and save transformers when I can. Depending on the application, they can be pricey and sometimes hard to find. I also love the fact that this transformer has duel outputs so a separate one isn't needed.

[Mercenary6502]

Update:

 

I went a head and tried the EPROM method just because I could even though I've been considering the PLAnkton. The PLA was defiantly at fault because the computer almost always fires up to it's blue screen now which is great. The downside is that I'm having timing issues so it doesn't boot properly. At first I was getting a screen full of garbage characters, but I started playing with capacitors between ground and CASRAM as suggested by a helpful Youtube video. This has made improvements, but it still glitches and locks up. The value of the capacitor used in the video is 68pf, but that seems to be too low for my C64 (the C64 in the video is a newer revision). I've been salvaging good caps from junk boards and have had the best results so far at 78pf (achieved by combining two of the few caps I have on hand). But it still glitches and becomes unresponsive (as can be seen with the falsely reported ram size):

 

 

Any ideas on how to get the timing dialed in correctly? I'm using an AT27C512R 45ns EPROM and as stated before, I believe this is an original REV A C64.

 

Thanks

[R.Cade]

It's very difficult to get an EPROM PLA running in an early REV board. Those were made for the 82S100 instead of the later CBM-produced 906114 so the timings are very different. Ray Carlsen may have some info on this on his site, but it's very hit/miss/experimental/your mileage may vary. Every board is different due to different RAM, TTL chips, etc.

 

Easy way on these is to find a real NOS 82S100 with PLA equations burned to it. arcadecomponents.com usually has them...

Edited April 26, 2016 by R.Cade

[7800fan]

Or just get PLAnkton and save headaches.

[R.Cade]

Or just get PLAnkton and save headaches.

Is that known to work on all rev boards, including the Rev A?

 

How is replacing with the originally-designed parts a headache? The only known 100% replacement is the original part... These were mostly shipped with 82s100's or the occasional odd equivalent.

Edited April 26, 2016 by R.Cade

[7800fan]

http://www.lemon64.com/forum/viewtopic.php?t=57265&postdays=0&postorder=asc&start=0&sid=768efb7c9850f9076bb746240aa70067

 

7 pages of comments about this and the person selling one. It runs cooler than a stock PLA, doesn't blow out, uses less power, and seems to be 100% compatible with all flavors of C64 excluding the later short board using 64-pins memory manager that integrated PLA and a couple chips. Those 64 pins chip almost never go bad.

[Mercenary6502]

Yes the plan is the to buy a PLA like the PLAnton but trying to get the EPROM to work in a REV A C64 is just an interesting challenge. That's primarily why I tired it. According to some information from Ray Carlsen's website (thanks R.Cade) the early C64's like mine have a very narrow range of acceptable timing. Not too slow, not too fast (like I pretty much discovered on my own playing with delay capacitors).

 

I also read something about creating the CASRAM delay with some slow inverters, but that probably wouldn't make a difference though.

Edited April 27, 2016 by Mercenary6502

[carlsson]

Well, the good thing now we know what was the cause for the computer not starting without touching the crystal, however that could salvage a bad PLA chip.

 

It seems the "old" SuperPLA no longer is produced (?), the realPLA is in low stock/unavailable, Jim Brain's PLA project hasn't reached his store so in practise the PLAnkton might be the only commercially available solution right now, besides using a fast EPROM and make adjustments. Personally, I've got a spare SuperPLA Multi with fewer and fewer machines to use it on, but I'm sure some day it might come in handy.