C-64 Black Screen and 5v Blinking?

[DistantStar001]

So I'm trying to fix my second C-64 motherboard.  I have one working now (out of seven), but with the exception of the PLA, SID, VIC II, and Color RAM, all the chips are soldered in, so swapping ICs isn't really an option.  The board I'm working on right now is an ASSY  NO. 250425, that gives a black screen  when powered.  It used to be fine until the component at Q3 fell off.  I replaced it, but it's been giving problems ever since.

I've tried my Dead Test/Diagnostic Cart, and to make a long story short I ended up socketing and replacing the RAM chips in U9, U11, U12, U13 and u21.  Also did the same for logic chips in U8, U13, U14, U25 and U27.  U15 and U26 were already socketed when I got the board, but I did put fresh chips in each.  Nonetheless, whatever I do, half the time, all I get is one flash error with a dead test, or garbage with the diagnostic (I usually have to hit the reset on the cartridge for that).  The rest of the time, I get a black screen, but the power indicator light (indicating the 5v power) blinks on the cart.  

Now, this isn't a new issue, and dates back to when I put in what I assumed was a damaged PLAtinum PLA replacement, but I've since replaced that with a PLAnkton.  While the board still didn't work, but at least the 5v wasn't blinking anymore.  Until today, that is.  

As the last time this happened, It seemed to be the PLA, I removed it, and decided to verify the chip in my other working board, and again, long story short, it's fine.  So I've left it in my working C-64, but I'm holding off on putting the other good PLAnkton into this board until I have a better idea what's wrong with it.  

So far in tracing this issue, I can eliminate the following: the PLA, the CPU, the VIC II, all the ROMs, and both CIAs, as suspects.  I also know it's not the SID as it doesn't have one at the moment.   I'm at least 95% sure that it's not the CPU (I've it them with a spare, but they get a little warmer than the one in the working board), and I can't imagine that it's any of the socketed RAM as I've replaced them repeatedly and checked all the traces. I've also piggybacked the remaining soldered in RAM with known good chips, so I don't think it's any of them either.  At this point I'm running out of ideas.  I know that I'm missing something, the question is what?

[shoestring]

Regarding the 5v 'blinking' and assuming the power supply is working OK on the other motherboard. Have you a ruled out a dodgy power switch on the c64 ?

 

The power switch in the 64 is known to oxidise internally. Have repaired at least one c64 that either gave me a blank screen, booted to junk or failed to a blank screen randomly during use.

 

Each time I had a blank screen, the LED went off. So that was a hint that I was dealing with a power issue.  After power supply, I'd be looking for bad capacitors and closely examining the power switch.

 

Edited July 18, 2019 by shoestring

[DistantStar001]

The power suply is a simple two brick for the moment with a 5v USB and a 9v AC spliced into the DIN cable from an original Commodore 64 PSU. It is the same supply I've been using on my working C46, and in every test, gives a consistent 5.24v and has never over volted beyond that.

 

I did check the 5v on the cassette port with a multimeter and got a pulsing reading that spikes roughly around 1.6v, and drops to a little over 0.3. This occurs with or without a PLA installed. Also, the pulsing seems to synchronize with the blinking LED.  I also plugged in an LED power light from an extra VIC-20 case I have, and noticed that the light was very dim when flashing.

 

I'm leaning against the power switch being the culprit due to the regular nature of the pulsing and blinking.  If it were the switch, then I'd expect the power to be either constantly low or absent, or the intermittent failure to be irregular and not evenly spaced at around once a second.  This pulsing is consistent and regular, not random.  The steady nature of the pulsing and flashing has me leaning towards an IC or other component that could be pulling the 5v low, and effect the timing of other components.  The question is what can do that?

 

 

[+DrVenkman]

Have you confirmed 5V on the Vcc points of all the chips? It might help  you track down where there's a short or broken/intermittent contact.

 

[shoestring]

5 hours ago, DistantStar001 said:

 

I'm leaning against the power switch being the culprit due to the regular nature of the pulsing and blinking.  If it were the switch, then I'd expect the power to be either constantly low or absent, or the intermittent failure to be irregular and not evenly spaced at around once a second.  This pulsing is consistent and regular, not random.  The steady nature of the pulsing and flashing has me leaning towards an IC or other component that could be pulling the 5v low, and effect the timing of other components.  The question is what can do that?

 

 

I hear ya. That's a weird issue for sure.

 

For clean power, the VIC-II requires a separate +5v power rail via an internal 7805 which is supplied by the 9v AC.  To the bottom right of that board on a heatsink. I assume Commodore were trying to eliminate the noise / jail bars you typically see on all c64s.

 

It's also vital for the clock generation ( MC4044,LS629, LS74 and LS193 ) and a few transistors in the chroma/luma output need it too. The modulator is also fed with CAN +5v which is the secondary 5V supply.

 

Would be interesting to know if the components tied to the secondary 5V supply are flickering as well.

 

Good luck.

[DistantStar001]

So far every IC I've tested is giving the same low voltages as the cassette port or lower, however the user port gives about 4.9v.  I did replace the rectifier, but that had no effect.  

[shoestring]

That’s quite normal. See attached. I believe that is supplied by VR2 ( have to open my breadbin and check )

 

So you should have roughly 12v at the rectifier CR4, VR1 and 5v at both coils ( L2 and 5 ).

 

If you have 5v at the power input pins 2 and 5 of cn7 and you’ve ruled out the switch then check the output of the coils at L5 and L2. They should both be around 4.9v.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Edited July 19, 2019 by shoestring

[DistantStar001]

I checked the voltage at C19 and got roughly 12v, also the vcc on the VIC II was about 5.1v.  So I don't think the 9v rail has a problem.  However when I checked the power DIN connector, I got the same pulsing (1.6v to 0.3v).  I'm guessing that means the port has failed, since the pulsing and low voltage exists so early in the circuit?

[shoestring]

10 hours ago, DistantStar001 said:

I checked the voltage at C19 and got roughly 12v, also the vcc on the VIC II was about 5.1v.  So I don't think the 9v rail has a problem.  However when I checked the power DIN connector, I got the same pulsing (1.6v to 0.3v).  I'm guessing that means the port has failed, since the pulsing and low voltage exists so early in the circuit?

 

I doubt it but you could feed the 5v upstream of the CN7 to verify. Ideally, after the power switch and with the switch removed.

 

Something is either 1) dragging the voltage down [ bad connection, even the crud in the switch can cause this ] or 2 )  something is shorting to ground. 

 

With the machine off and your power supply disconnected from the c64s power connector. What is the resistance between 5v and gnd pins on the connector ? ( pin 5 and pin 2 ).

On my rev C board  I don't get a reading ( completely open ). If you measure really low resistance then you have a short somewhere.

 

On my cassette port of my rev C board I measure 3.2k ohms & 1.8k ohms on the user port. Yours might be different but if you get a very low resistance closer to 0 then expect you have a shorted component somewhere. 

 

You need a good precise multimeter to identify a short as you get closer to the source, your meter will register a lower resistance as you get closer to it. If your machine isn't socketed, it will be harder to find.

 

If you don't have a dead short then what is the current draw on the 5v line ? ( you'll need to put your ammeter in series and set it to mA  ). A healthy c64 will pull between 800ma and 1000ma. What is your power supply rated at if it exceeds this amount ? ( if it's exceeding this amount then your power supply is shutting down to protect itself ) ?

 

You've mentioned that you replaced Q3 when the issue started, what is the type of transistor you've replaced it with  ?  ( The original part is C1815 or 2SC1815 which is NPN,  plenty of those available ).

 

 

 

Edited July 20, 2019 by shoestring

[DistantStar001]

The voltages check out on the VIC II, but I did find an intermittent short on the power port. I replaced the port (and ripped a large trace in process). I did managed to patch it, and the new port works.  I was advised to swap out the USB charger.  The new one is 5v and 2.5 amps. That got rid of the pulsing, and the new USB works with my working C64 as well. 

I now get a consistent 4.93v on the cassette port, but the RAM Chip at U9 got extremely hot, and I'm relatively sure that the CPU went bad. I'm guessing the result of consistent the under voltage they received while I was tracking pulsing down. I replaced the CPU, and the RAM chips, but when I plugged in my dead test, I got the same one flash error I had before. The Diagnostic Cart, on the other hand, give colorful garbage with a black border. Not sure what that's about, but it could be that replaced U13 and U25 and the only replacements I have are 74s257 chips instead of 74ls257s. I was told that it should work, but I don't know for sure. 

In addition to U9, I've replaced the RAM chips in U10, U12, U21 for good measure. None of the new chips get hot, so I'm guessing that's good, but I still have some sort of fault on the board.

[shoestring]

I would never expect it would be a shorted power connector. Looks like you're getting somewhere. Does your diagnostic cart and dead test cart still work in the good c64 ?

 

Not sure about the 74LS vs S. But generally speaking, timing might be off when substituting 74LS with S so you have to be careful there.

 

What does the c64 do without the cart ?

 

Any screenshots ?

 

Since you had a short to ground, it would be good to check the PLA again, 

 

But booting to junk generally means 1. bad ram / kernal rom or 2. bad pla. 3. bad ram address multiplexers ( U13, U25 ) 4. bad cpu or 5. CIA ( but a cart should normally work even with a bad CIA chip ).

 

This is a great guide  btw.

 

https://retrocomputerverzamelaar.nl/commodore-64-problems/ 

 

Edited July 20, 2019 by shoestring

[DistantStar001]

The cartridge boots fine in my working C46, in both Dead Test and Diagnostic mode.  

 

Without the cart I get a black screen (both through the video out port and RF modulator).

 

Screen shots below.

 

For the most part, I had yanked the PLA while tracked down the fault, but from what I hear the PLAnkton's are more resilient than the originals.  

 

Most of the RAM has been replaced.  I can try another kernal ROM.  The PLA as mentioned before is a PLAnkton and received only minimal exposure (the one that was in the board when this started is in my working C46 at the moment, and shows no problems).  The CPU was replaced as the original only gave me a black screen on the Dead Test (no flashing), the new one gets the flashing.  The fault screen from the guide you posted, my garbage screen is similar, however, I only get it with the diagnostic cart.  It also resembles one of the RAM fault screens as well.  At the moment the CIAs have been yanked.

[+DrVenkman]

So if I understand, the Dead Test Cart gives one flash? Try swapping your 74S175 chips and see if you get a different number of flashes. 

 

Do you have an oscilloscope or logic probe? You can use either one to verify activity on all the address and data lines, plus proper Vcc and Vss on them and the logic chips. I’d also examine all the sockets you installed under bright light and magnification to ensure no broken traces, verified with continuity measurements between all the reworked areas and the adjacent components. 

[DistantStar001]

Swapping the 74S275s in U13 and U25 had no effect, and every pin other than ground reads low with a logic probe.  Sadly my oscilloscope does not work at the moment (a project for another time).  I'm getting activity from pin 14 on U9, U21, U22 and U23, but the rest of the RAM reads low on that pin.  Also the VCC on the active chips reads a little over 2v and closer to 1v on the inactive ones.  I'm guessing that there is still a short somewhere pulling that low.  Also, on the VBB, I'm not getting any voltage off the VBB lines.  The 5v on the 74S275s in U13 and U25 and the CPU are fine.  

 

Oh, and I've checked all the traces on the sockets and they seem fine.

Edited July 20, 2019 by DistantStar001

[shoestring]

2 hours ago, DistantStar001 said:

Swapping the 74S275s in U13 and U25 had no effect, and every pin other than ground reads low with a logic probe.  

 

Hmm. Is your 6510 even getting a valid reset signal signal ? Check pin 40 of the CPU with your logic probe while the computer is off and then on again. You should see a low for about 1 - 2 seconds then toggle to high.

 

If you do not get a valid reset check the 556 timer ( U20 ). On the 6510, /IRQ line should pulse ( pin 3 ). On pin 4 ( /NMI ) you should have a high reading. If the reading on pin 3 of the 6510 is not pulsing then you most likely have a bad CIA chip at U1. And if pin 4 is not held high then the other CIA is stuffed ( U2 ).

 

Yes, a bad CIA can result in booting to junk or black screen as those provide some of the signals to the 6510 ( /IRQ and /NMI ).

 

 

[shoestring]

Other things to check are pulses on pin 38 of the 6510, that's the / read write enable.  It should pulse.

 

What's the voltage on pin 6 of the 6510 ?

 

If something is still shorted then something should be getting hot ? ( with that much voltage drop your power supply should too )

 

At this point it sounds as though your CPU is not running at all and stuck waiting for a valid reset, hence the 2 x ram address multiplexers sitting around waiting for something to do.

 

The address pins of the 6510 are most likely stuck in a high state.. ( pin 7 - 20, 22 and 23 ).

 

[+DrVenkman]

I'm concerned about that crazy-low reading for Vcc. Personally I'd go kind of "brute-force" in my analysis: I'd power up the board, then start testing voltages at every point on the +5V rail from the power DIN all the way to each chip. 

[shoestring]

Me also. But it just doesn't make sense why he be getting different voltages on the same +5v rail from the DIN source. 

[DistantStar001]

U11 had an intermittent failure causing a slight short between pins 10 and 11.  It's been replaced, and I'm now getting between 4.9 and 5v from the VCC on all chips.  

 

I still get a black screen with no cart, one flash on the Dead Test, but a new kind of garbage on the Diagnostic cart.  I wasn't able to take a picture as I haven't gotten it to repeat (most of the time the screen is just black), but it was reminiscent of a failure at U13.  The 74S257s in U13 and U25 seem to get a little warm, so I'm guessing that they really aren't as interchangeable with the 74LS257s as I was led to believe.  

 

Pin 6 on the CPU reads at 4.96v

 

The 566 is new, but was installed while I was having the under voltage issues.  Still, it is giving a proper reset to the CPU. I also think that my probe may have been wired wrong, as it should be flashing green for high and red for low, but appears to do the opposite.  That, and off is on, and on is off for this thing.

 

I also replaced my kernel ROM.  

 

Pin 38 on the CPU does pulse rapidly between high and low.

Edited July 21, 2019 by DistantStar001

[shoestring]

46 minutes ago, DistantStar001 said:

 The 74S257s in U13 and U25 seem to get a little warm, so I'm guessing that they really aren't as interchangeable with the 74LS257s as I was led to believe.  

 

That's fine for them to get a little warm but not hot to the touch that it burns your finger.

 

How about the /NMI and /IRQ signals to the CPU and the address and data lines ( they should pulse ) ?

 

If you don't have anything, check the CIAs. I'm not familiar with the two carts but generally I'd imagine these should work without the kernal rom, in any case without the cart plugged in, I would be checking the enable lines from the PLA to the Kernal rom, Character rom and Basic rom ( pin 20 ) on each mask rom. Basic rom -> high , Kernal rom -> pulse , Character rom -> pulse. If these signals are wrong then U17 has crapped itself.

 

 

Regarding your probe. Is it showing green instead of red on pin 6 of the CPU or when you measure VCC anywhere else  ? If so then I'd have to agree with you... otherwise the reset signal is inverted and is wrong.

 

 

 

Edited July 21, 2019 by shoestring

[DistantStar001]

I get no detectible activity on the MNI OR IRQ on the CPU (with or without the cart), it just holds high.  I got similar results with the CIAs, but I'm not sure if they work.

 

Just re-read.

The probe shows red on the vcc.  So my signal is inverted, but the questions are how and why?

Edited July 21, 2019 by DistantStar001

[shoestring]

8 minutes ago, DistantStar001 said:

Yeah, it shows red on the vcc.  I guess that means that every time I said low before I found out, that means high.

 

Yes. Red = high, green = low. So I guess reset circuit is good then.

 

8 minutes ago, DistantStar001 said:

I get no detectible activity on the MNI OR IRQ on the CPU (with or without the cart), it just holds high.  I got similar results with the CIAs, but I'm not sure if they work.

 

Yeah, it shows red on the vcc.  I guess that means that every time I said low before I found out, that means high.

 

I think U1 is toast. Without those signals, especially without /IRQ your CPU will not do anything. CIAs are sensitive and won't take much punishment...   even just touching the joystick pins with machine turned on can damage them. 

 

U2 sounds like it might be OK. You can try swapping them around ( U1 and U2) and if the other CIA is working then you might have different symptoms and maybe even a booting machine!

 

Good luck.

 

[DistantStar001]

1 minute ago, shoestring said:

Yes. Red = high, green = low. So I guess reset circuit is good then.

Ok, good then.  I was just going through the 566 pin by pin.

[shoestring]

So to be specific. 

 

CIA U2 IRQ connects to NMI pin 4 on the 6510 ( should be held high )

CIA U1 IRQ connects to IRQ pin 3 on the 6510 ( should pulse )

 

If you get no activity on pin 3 then replace U1.

[DistantStar001]

I had two good CIAs in another board, now I have one since pin one broke off.  But it did test good, so maybe I can solder a new leg on it?

 

In the interim I'm not messing with the CIAs, since they're not needed to boot, or for a dead test.  Also, I may have only have one good one left.

 

With or without ROMs or CIAs, I still get a single flash on my dead test, and black screen with no cartridge.  I did get a picture of the new garbage screen on the Diagnostic below.  I've replaced the RAM repeatedly, and I can't believe that they're all bad.  

Edited July 21, 2019 by DistantStar001
typo