The UGLIEST Repair of a Commodore 64 (Part 1)

hello and welcome to novel's retro lab this may look like a normal intro to you but it's not i am from the future and i'm coming here to warn you about the repair for this commodore 64 that you're about to watch and i just finished and it's going to get ugly really ugly if you're looking for something that involves cleverly figuring out the faults based on observable data cleanly replacing the parts and easily verifying the results go look elsewhere what you're about to watch may be disturbing to some people so watch at your own peril you've been warned [Music] all right it's been a while since i've done any kind of repairs in a commodore history for but it should be really fun getting back into it so let's give this a try this is from my friend javis and i think this is not working the led is missing so before we even power it on let's open it up and look at it and then we'll power it on and measure voltages so this is a 250 407 assembly i think that's one of the earlier ones so it probably has the more complicated clock system here right this is not the simplified version of the clock system everything looks pretty straightforward things are socketed which is good so i'm going to remove this annoying rf shield i usually just not keep those and then let's power it on and see what happens okay let's power it on and see what we get and whether this no signal message goes away or not we seem to get nothing and we get no signal that might make it even easier okay so that means that there's something that is really not working at all so the the graphics is chip is not generating an image at all and that's usually because there is either no voltage getting to it it's completely dead or there's not clock signal those are the most likely options right now so let's check voltages first i'm going to get ground in here this is the input to this 17 dc that's fine it should output about 12. yeah that's good enough and then we'll first find out the negative the input to this should be what nine yeah and output should be five perfect so these are the dc voltages generated from the 9 volt ac we also get 5 volts straight into the rest of the system so i should measure that for example at the cpu let's slide it over there we go and that's pin six and here we get four point eight nine that's four point nine yeah that's probably good enough okay in one last check i'm going to check the ac voltage right here that should be what i said 9 or yeah something like that ac okay so voltages seem fine before we continue i wanted to thank this video's sponsor pcbway pcbway makes really high quality printed circuit boards supposedly they call them prototype boards but really the quality is just as good as any production board i've come across and certainly way better than a lot of these 80s boards that we see so often if you're interested in printing your own boards head over to pcbway.com and check them out so as usual let's start checking the clock signal on the cpu and pin one and it looks like we're getting plenty signal right there that looks totally fine that's slightly under 1 megahertz so that's yeah that's great and then in 39 we have clock out which i think maybe is like the half signal or the inverted clock or something like that and that also looks totally fine so at least the cpu is getting a clock signal which means the vic 2 is also getting a clock signal because it actually generates the clock signal and sends it over to the cpu those were my two most likely candidates for why we're getting no video signal at all so i should probably just look at the video signal itself coming out of the vic2 so we have two signals that we really care about coming out of vague two one of them is sink and luminance and those are that's in 15 and then your color information is in 14. so this is luminance in sync and that looks fine for an empty mostly empty signal and color [Music] so yeah that's surprising that looks like we should be getting a signal but the tv was not identifying anything okay so this is one of those slightly embarrassing moments where i'm double checking things because it makes no sense we seem to be getting a perfectly fine video signal out of the vic 2 even though it looks black but we're not getting anything here and then i suddenly had the thought oh no are we set up for the right input so we weren't we were set up for scart and this is a composite input i told you i hadn't done things with the commodore 64 in a while so it even says it here actually so now we're in the right input and the right source and let's turn it on and okay okay this this makes a lot more sense so we are getting a signal like we already saw on the oscilloscope it's just black screen so the classes for this are going to be different than my original theories immediately i would think pla especially because the voltages are fine it's very like a pla it could even be a memory although those often have some patterns i could pop out the sid chip just to make sure that it's not interfering with things actually let's do that right now so the sit tip that you always need to check this because the pla and the sid will swap places depending on the revision of the board we don't need this anymore and i believe this is the pla let's see the ic ic u17 that is the pla right and so this one is the set let's just remove it just to get rid of one possible thing from going wrong and let's turn it on right and we get nothing the pla is socketed so we might as well try it before we do anything else not even that does cartridge it's so likely for that to be the problem so here have some pla replacements that i've used in other videos before make sure it's off yep and then if this is not it then we'll continue maybe with the bed test cartridge okay so it wasn't just a bad pla although the pla could be bad until we it in a different computer but this is probably the way to go for now so time for a test all right with the test cartridge and we turn it on it's normal that it's blank for a few seconds but after about maybe 10 seconds at most it should get us some information on the screen and if not is that there is something more serious than just some ram or it could be a catastrophic failure of a ram that is interfering with the bus as well that's another possibility suppose i should take out the roms as well yeah this is not coming back so let's take out the roms since they're socketed we don't need any of them for the test and they often go bad as we've seen in past videos oh we get a white screen this time doesn't mean we're going to get some cartridge output doesn't seem to be the case but that's interesting i hadn't seen the sort of gray output before does it happen consistently i'm going to turn it off and on again it's not consistent interesting oh okay this is very good so it looks like we're probably having some memory issues the test is trying to do things i don't know if it's going to even be able to identify which of them is wrong this looks like maybe some kind of video memory is wrong i'm not even sure it's making any progress it seems to be stuck in this loop this probably means it's a memory problem right now maybe in addition to the pla and maybe in addition to the rom so once we fix this we can try adding the other chips in i tried turning this on and off a few more times and i didn't get any output so i think that something is really clobbering probably the databus potentially the address bus so let's have a look at them maybe we'll be lucky and we'll find out that one of the data chips is holding data out always at ground or always high or something like that so let's check him out that looks really bad because that's not even i mean that's just noise that is that those are not very fast clock signals that's just really loud spiky noise so yeah that looks really bad i'm surprised that anything loaded whatsoever oh and then what is that that's mostly high i mean that's more believable although i would expect the output not to be high so so long very similar very similar okay and then this oh this is not just noise so it's just generally ugly something like that so what's that noise yeah this is noise but those are not those are something and that is nice problem with this is that we don't know where that's coming from is it coming from the ram i see is it coming from really anything else connected to the data bus what about the address bus itself let's look at it on the cpu because there's the easiest always high that's not a good start okay that's general activity that's fine [Music] and then suddenly one of them high like that yeah that's not good so i'm not seeing any single ram i see that i'm pointing to it and saying that is most likely faulty so as i'm looking at the board something happened and then suddenly i stopped seeing signals so start tracing it back and i notice there's no clock signal like what's going on and then i notice that i'm having problems with the voltage generated by these two in particular this voltage regulator and the vic is not getting any any power so let's see what's going on that it's probably an unrelated problem to what we're having earlier so let's see i'm going to turn it on [Music] so that seems okay 5 volts yeah and this 21 [Music] i mean this is great so i don't know why it dropped to nothing i checked the power supply and the power supply is fine i even tried a different power supply and it's also fine i even have one of those guard things is not on right now but it's just checking that you're getting nine voltage ac and five volts and yeah all of that is fine but for some reason the power here just um went away which is not good let's try it again [Music] oh what what is this look at that that's yeah so i turn it off and on and we're getting nothing there and we're getting two and what about the ac input supposed to get not so that's fine but we are getting oh now we're getting something with this is crazy a second ago we're getting nothing now we're getting so this is even higher than it was a second ago so there is definitely something weird going on in here i could try swapping the voltage regulator it's a relatively easy thing to do they do fail sometimes i'm not 100 convinced this is the problem because i've seen that typically when they get overloaded and you know you need to leave them 20 30 seconds especially without a heat sink and they shut down but i guess if it's faulty the shutdown part could come in and and they could um they could fail like that it's the best theory that i have right now i mean i don't think there's gonna be any of the capacitors along the way that could be causing this or the coil and there isn't much else so you know the other interesting thing would be does that happen with this removed so before we even put another one in just to isolate what's upstream from this so yeah let's remove this and turn it on and see if we get a reliable what was it supposed to be getting like 9 or 10 volts in here with the voltage regulator removed let's check what the voltage is at the input and yeah it's still really low 1.3 and dropping okay so i suspect that it's not that voltage regulator it didn't really behave like a fault in there but it's just one of the few things that was in the way so there's that there's this other voltage regulator the 12 volt one that usually even has a much lower current draw so i doubt that's the problem i removed the 12 volt voltage regulator because really there wasn't much else to try in here and let's see if i made a difference now we're still almost no voltage there and what about the input to interesting so the input to the 12 volt regulator is perfect it's a 21 [Music] volt and obviously there's a little higher because we're not drawing any current from there but the 5 volt one is abysmal let's look at the difference between the two circuits the commodore 64 power supply provides two different kinds of voltages five volts dc on one side and nine volts ac on the other we're having issues with the nine volts ac so let's focus on that side after it goes through some capacitors and coils the nine volts ac goes through a fuse on one side and the switch on the other and then immediately after that the parts of the system that need nine volts ac tap into that and that will be mostly the expansion slot i think following that we have one path that goes to the 7805 voltage regulator to generate 5 volts dc and another one that goes to the 7812 to generate 12 volts dc both of these voltages are used in the vic-2 and sid chips i imagine it's designed this way to avoid noise crossover between this part of the circuit which is probably quite sensitive to noise and the rest of the system so that way they're completely isolated the 5 volts part of the circuit which is the one that we're having so much trouble with starts with a bridge rectifier that converts the ac voltage into a dc1 probably around eight or nine volts dc and then it goes straight into the voltage regulator so maybe that's where the problem is this is a rectifier normally you may see those with like four diodes but this has pretty much like the four diodes built in i'm inclined to think that somehow this randomly failed us in the middle of the repair and by replacing it we may be able to get our power back it's easy enough to test i don't have any spare ones so i'll have to take it from a different commodore 64 board but that's not a problem i'll go ahead and order some spare ones it's probably a good thing to have on hand anyway okay the rectifier is out of the board and let's see if that makes any difference i'll check that and that well that should be nothing right because nothing gets to here so i guess i can check the input of that one that one is still fine and i can check the ac which was dropping earlier so this should be about nine or ten yeah so that seems reasonable so maybe it was the rectifier after all so this was actually pretty difficult to remove from the board it has really thick legs and i'm not looking forward to doing that with another commodore 64. so i would rather check to see if i can at least make sure that this is not working and so a rectifier pretty much is just four diodes arranged in a particular configuration so here we have plus and minus the that's the output and then the other two pins are the ac current so we should have if i flip this so we have plus minus in the inputs and we should have diodes that are pointing from here to here from here to here and then from here to here from here to here and we could at least test that those diodes are working that way and the current is not going the other way around so let's do that really quickly so i'll set the multimeter to diode testing and just to double check this is the positive this is the negative that's good that's good that's good that's good and then the other way around we shouldn't get anything nothing nothing nothing so on paper this looks fine it may start failing once you actually apply some current that's certainly possible so there's something weird going on because i'm testing things and voltages seem to come in and out and i unplug things and i'm getting different results so i'm thinking maybe it's something you know just not even one of those components so i've hooked up the oscilloscope to the input of the voltage regulator this is the should be getting like 10 12. and so yeah so i've turned it on and we get nothing and so i'm just wondering is there something else going on like this has a fair amount of play this is the ac voltage or half of the ac voltage coming through the switch and does something change as i move this i mean it seems to move a little bit in response to me pushing it and it really moves a lot the um the connector so suppose we could have some dry solder joints but um i didn't see anything earlier so let's look at them a little bit more closely because i can't make any sense of why we wouldn't be getting the full voltage otherwise so these are the solder joints underneath the power connector and i mean they look okay but it definitely has a lot of play or some play anyway and looking at it closely maybe they're moving a little bit so let's have a closer look with the microscope so look at that that is a crack that's a crack and that's a crack those are definitely bad joints they're called dry joints or cold solder joints and if i move it oh look at that look at that i'm barely touching it and you can definitely see how it's moving oh yeah look at that so this is really amazing that on the microscope it's very obvious but when i was looking at it on the board they looked okay they look shiny and yeah look at that like how can make them wiggle so yeah we definitely need to reflow those and hopefully that's the full source of our problems and then we'll have power again so to fix this i will just melt the solder around the joint and then add some extra fresh solder and i'll do that with all of them definitely with the three that were wiggling but i'll do it with the other ones in the connector as well oh yeah now it's not moving at all so that was probably it all right let's test it again and see if it makes a difference look at that that's the input voltage into the voltage regulator first try and this one over here looks fine i mean the shape is a little weird but that's totally fine so yeah and i'm sure we're getting five volts yep that's for 5 volts exactly and 12 volts over there [Music] so wow all of this digging and really was the power connector so really the the lesson to be learned here is that on commodores if this wiggles at all look at it very very carefully even with magnifying tools because chances are some of those there could be some dry joints and you may not even be able to see it with a naked eye all right now that we've finished all that weirdness with the power connector we can go back to the problem we're trying to diagnose and see what's going on with the main board at least now we're not going to get any weird interference and power cutting in and out like we're getting before and we can focus on the problem itself if you enjoyed this video please consider supporting noel's retro lab on patreon or joining the membership on youtube not only is that the best way to support this channel and allow me to continue making more videos but you also get some extra perks like early access ad free videos and more thank you again to all the supporters see you next time you

Channel: Noel's Retro Lab

Views: 30,578

Keywords: commodore, commodore 64, c64, black screen, pla, MOS, 6510 CPU, basic ROM, kernal ROM, color RAM, 4164 RAM, faulty ram, dead test cartridge, video signal, vic2, graphics chip, cold joints, dry joints, power connector, voltage regulator, bridge rectifier, intermittent voltage, clock signal, bud signals, oscilloscope

Id: 8Bu_VMoiEIg

Length: 25min 31sec (1531 seconds)

Published: Thu Feb 11 2021