Can we fix bad chips ... in the oven?

well hello everyone and welcome back to adrian's digital basement on today's video i'm going to try to fix some dead chips i think everyone knows my dead parts bin because i love to throw bad chips in here so i have lots of stuff to try this process on so let's get right to it [Music] ah the dead part's been i love this thing people always ask me like why do i keep these old chips but i have to say there's something cool about the fact that i know that i've fixed so many computers that i've ended up with all of these bad chips here which in and of itself is kind of cool but also maybe there's a way to fix some of these and of course if i'd thrown all these out i wouldn't have the opportunity to try right so first i'm gonna pick out a selection of these chips here to see which are worthy contenders for attempting a fix and these are the chips i've come up with for this little experiment now starting from the left side here we have four commodore 64 roms i think there's three kernels and one basic we have eight pla chips we have seven sid chips and starting at the top here we have three ted chips this is the graphics chip graphics and sound chip from the commodore 16 and the plus four then we have two 8501 processors these are also the processors from the commodore 16 and the commodore plus four um these chips are extremely unreliable as are these and then here is the vic 2e the palvik 2e that's from my commodore 128 the german commodore 128. i recently had a mail call video where i replaced this with a chip a viewer had sent in and that fixed the computer and then the next row here we have two 6510 processors well one's a 6510 one is an 8500 which is sort of the slightly newer version of the same chip they are interchangeable with each other and then down here we have two commodore 64 vic 2 chips now even though all of these came out of my dead parts bin the problems with these chips are kind of varied and they're not all just completely dead some of them have different types of faults so the reason why the commodore 64 is on the bench here the zip 64 is because i went through and i tested all of these chips anytime it says bs here it doesn't mean bowl you know what it actually means black screen and uh because that's a really common fault on the pla and with roms processors things like that i just wrote bs when we're getting that but there were some other interesting faults that popped up and i would like to show you a few of those just as a comparison to see if this process actually changes anything when we do it compared to how it is now so first i would like to start with pla number six so here it is nothing special about it but it does something interesting when you put it in the 64. so i'm going to replace the pla in this machine i currently have the plankton in here and we'll pop in the number six pla into the machine there it is and when i turn on the 64 let's see what we see okay so it appears to be working except for a problem the computer successfully booted into basic without issue but look at the text it's just rainbow colors so that means that there's something going on with this pla where it's not mapping the color ram correctly into the vic2 space or something to do with that let me try out this easy flash three see if it works we might just get a black screen when i try this oh it looks like it's actually working correctly so yeah that's a strange failure mode let me go into this okay yeah everything looks good okay so just sitting here i was about to run donkey garden arcade and slowly but surely we're getting sort of flashing flickery colors until we have this very nice christmas effect on here now which you know isn't totally terrible looking but it's certainly not normal operation on the 64. let's just try donkey kong arcade see what happens there's the rainbow for a split second [Music] oh wow you know what's gonna have to happen [Music] yep eggpid dance party even though it's got flickery colors and other issues [Music] it is interesting that this graphical screen here doesn't have any corruption whatsoever it seems to only be in text mode although donkey kong arcade that graphics splash screen it was having issues and i went to this version of jiffy dash which is a hybrid with dolphin dos and it shows a little rainbow on the startup screen here and instead of the normal solid colors we're getting this crazy effect so i thought that was a little bit of an interesting look at this weird failure mode on this particular pla when all the other ones that i have here are just black screens so this is the only one that does anything that's a little bit different now these here are the bad sid chips and i actually have a couple that act really strange and i want you to hear how they sound so this is the number four chip and i'll just replace my good sid with this number four turn on the computer oh we got a black screen because i forgot to put my pla back in oops okay so over to adrian's tools and we're gonna go to the sid tester which i think if you watch my channel you've heard this before it just runs through a whole bunch of waveforms and modulations and it's perfect for testing all of these sid chips when i push number one it should run through all the tests and you will hear all three voices just play through some normal tones it'll say sine wave triangle wave square wave things like that and then there's some filter test so take a listen to how this sounds [Music] [Music] [Music] so yeah that sounds a little screwed up that was the number four chip and on my list here i wrote that one down as just all screwy so some of the examples of things you have are like this number one sid seems to work except it has bad low pass and bandpass filters so regular voice playbacks of all three voices it works actually fine but then as soon as you go to use the filters i think in this one you don't hear any sound at all so those filters are dead altogether and then number two chip you do hear all the sounds as normal but it's super super quiet you have to turn the volume crazy loud and then you hear a lot of interference so something's wrong on the output stage there this one has really quiet output but then it also has dead filters where when the filters play you don't hear anything at all so sort of screwy that was number four we just heard number five here says bad v3 so bad voice three plus dead filter so voice one and two work fine but then it goes to use voice three you don't hear anything and then the filters are all dead you don't hear any sound either and then we have number six which is all screw which i'll play for you in one second and then number seven is working except it's extremely quiet again just like number two so let me pop in number six you can hear what another all screwy one sounds like into the machine powered on okay that gave us a black screen for some reason which is weird because it's okay that time it worked that's weird okay you hear nothing on voice one at all well i hear just noise uh wow caused the computer to crash let's see if that works but it did crash midway through okay yeah something's very wrong here with this sit in the computer yeah i sort of remember having one of my bad sids where it would cause weird issues in the computer not to mention it would sort of work sometimes and then other times it just would completely have no sound or crash and that's exactly what's going on right now but the first time i ran through this it sort of sounded like the number four one except it was all garbled like you'd have tones that were all over the place so it was rather interesting and unfortunately i didn't record it on my initial testing all right well i've talked a lot about these bad chips and what's wrong with them but i haven't talked about what i'm going to try to do to fix these and i'm sure everyone's been screaming at their screen telling me to hurry up and talk about this what prompted me to make this video and try this experiment was i got an email from daniel mantioni he's the man who designed the gal pla project which of course i focused and shown on this channel where it's an easy to make pla replacement for your commodore 64 using very very inexpensive parts from china like a few dollars worth and a pcb you can just get made anywhere and then you have working pla chips he told me he had been reading the atari forums and saw a thread about people talking about reviving chips using a very unorthodox method and it's baking the chips as in yes baking them in an oven at high temperature he actually went ahead and tried the method himself just using his oven in the kitchen and had some interesting but mixed results so before i tell them to you i want to try this out myself and see what kind of results i get with these bad chips because maybe some of these maybe they'll come back to life or maybe none will come back to life who knows right it's definitely an experiment so according to daniel the method that the people on the atari forums used was they put the chips in the oven at 150 degrees celsius for 30 minutes and when they removed the chips they let them cool tested them and they actually had some that seemed to start working again so i actually used my oven in the kitchen for cooking food in and i didn't really feel like putting computer parts in there even though there's probably no danger 150 degrees celsius about 300 degrees fahrenheit and that's not going to melt the chips or anything like that but it still might release flux residue or who knows what in there and i just decided i didn't want to do that so i hit up someone i know in portland that happens to have one of these toaster ovens and yes this is a regular black and decker toaster oven that was originally designed for cooking food in and has now been adapted with the use of this microcontroller that takes over the heating element in here not to mention a little servo that can crack this door open and this is used specifically for reflow soldering now the design of this little door and all this insulation in here is to allow precise temperature control so you set up these profiles on the screen and then that allows you to follow the appropriate temperature curves for doing exact refill soldering but what's really sweet about this oven is you just go into bake i can hit edit and i can set the temperature say to 150 degrees and the duration and the oven will take care of that exactly and because this thing will never be used for food again it doesn't really matter if this thing releases a bunch of toxic chemicals inside of there and it will never contaminate the food also what's pretty sweet is the original grate that this thing had has been replaced with this solid metal sheet or aluminum and that way the ics i put in here won't fall through and down into the bottom of the oven so first things first let's load up all these ics onto the metal sheet so i can put them in the oven i mean this sounds completely ridiculous like how is this possible that this is even gonna work but maybe it will that's why we're doing this experiment i spaced out the ics so they weren't all touching and there we go close the door there they are little chips in there they're about to bake and currently the oven is 22.8 degrees inside there we go 150 30 minutes i assume i push this button here open after bake close when cool use cooling fan once baking is done i mean are these different options i can set i'm going to set it to use the cooling fan when done and there we go bake 150 degrees for 30 minutes start preheat oh man i can't even believe it on the screen here the temperature is rising it's uh now 31 degrees and it's rising pretty rapidly 32 33 so it's it's going up quickly all right now i just wait till it gets to temperature and then the 30 minute countdown will begin so i came to check on this oven and uh sure enough the baking is done it's counted down to zero and this servo has cracked open the door so it's letting the heat out the oven has dropped to about 81 degrees celsius so i'll just let these chips cool all the way to room temperature before i test them here we are the cookies are out of the oven i mean i'm just kidding the chips are out of the oven all of them have cooled completely to room temperature so i'm gonna go through all of these test them out on the 64 and of course the c16 see if these things work testing is complete and here are the results so everywhere you see a red dot that means that there was no change no improvement basically it had no effect on all these chips but take a look right here i actually have one green check mark that was a pla that had a black screen and it now boots the computer although it's not perfect it has glitching color which actually matches the number six pla which still works and is completely unchanged from before but yeah number four is working this is how a working pla looks on my system 38 911 bytes no rainbow text so this is the number four pla the one that's seemingly fixed i'll just pop this into the machine and now when i power it on as you see it's working but we're getting this flashing rainbow text and also notice 30 719 bytes free which is a little unusual i'm not quite sure how the pla can be having that effect but clearly this is so it's intriguing that this chip actually improved after the baking process but it's still not completely fixed so what mechanism could be the explanation for even any improvement i mean one out of all these chips is not a great percentage but it's something danielle who originally alerted me to this phenomenon of possibly fixing chips with heat sent me this link to an article about electromigration now i know a little bit about electromigration and this is something that definitely causes a lot of failures in ics especially when they're just sitting there unused and they die but it all is also sped up by temperature as far as i'm aware and a lot of design considerations go into the design of integrated circuits to allow them to not suffer from premature failure from electro migration this seems to be a blog from someone who worked in an ic manufacturing facility in silicon valley known as zymos i'm unsure of what that is but he was a reliability technician he says one of the most feared things about ic chips were ovens hundreds and even thousands were placed in chip carrier trays and put into ovens at temperatures as high as 150 c for days or even weeks on end and eventually they all died but the story doesn't end there if i return these same dead chips back to the oven this time without any bias which is basically no power applied to them after a time they'd miraculously be restored to life and he goes on to talk about that he thinks it's the electromigration that somehow the heat without the bias is actually reversing some of that electromigration i'll put a link to this article so you can read it for yourself but he kind of says that if you have a chip that kind of comes back to life when you put it in the oven you should put it back in the oven and do it for longer and that hopefully will help restore it more there are other failure modes on ics of course besides this that won't be reversed by this phenomenon if this is a real thing even at that but he seems to have said he got some success with this and of course there's that original atari post that daniel had mentioned which i haven't read personally that talked about other people having success and 100 this number four pla here was not working before the oven and now it sort of works so it it at least partially was recovered now of course i've also been talking to frank aka isaiah 8 dwf about this potential phenomenon if this could actually work and he's skeptical as as i was i when i first tried this but i've already shared with him the results that one of these chips has recovered at least partially and he suggests that perhaps i try this in the oven again at a higher temperature and for much longer so because i don't really have anything to lose i'm gonna try that i think i'm gonna do 200 degrees celsius for two hours in the oven and then we'll see if that pla that got came back to life gets any better or maybe some other chips also are revived it's worth trying right incidentally i talked to the person i borrowed this from and he said this oven is a kit and it wasn't inexpensive but um i will put a link to the description for this particular kit and i think there are probably other kits available at this time so let's just reposition these chips so i don't want any of them touching each other we're going to bake we're going to edit and turn this up to 200c and we're going to do two hours i'm going to change this to leave oven door closed after baking and don't use cooling fan so it's just going to slowly ramp down the temperature when this is done and there we go 200c for two hours hit start and now we wait and it's actually the next day these did stay in the oven for a full two hours at 200 degrees and then whatever it took to cool i had the door stay closed while i was cooling so these chips have definitely been baked now to see are any of them improved before i test any of them let's turn on the xiv64 make sure everything is working come on retro tank there we go looks good no funny text 38 911 bytes i am going to remove the plankton and i'm going to stick the number four pla back in if i can find it so it seems like one side effect of 200 degree baking is that the sharpie writing is now harder to read you can still see the x but the number i wrote on this chip for instance can't really see it of the plas only the number five chip is actually hard to read the rest are still readable they're fainter than they were but they're readable so i can go right to the number four chip which is the one that had sort of improved on the last round at 150 degrees and i am going to turn this on for the first time put your comments in the comment section below to say if this chip is any better or any worse than it was before i did the 200 degree baking here we go oh wow we're back to the black screen so it actually made it worse now i'm just checking the connections definitely no pins are bent or anything like that and we actually have worse all right so 200 degrees that was not good let's try the number six chip which is the one that worked with rainbow text from the very beginning let's see if it has degraded um wow well um this one actually appears to be working now what uh okay i'm pretty surprised um let's put the diagnostic cartridge in because this one was always failing with the diags this is the regular diagnostic test it would get to oh there we go it's still screwed up how funny this was working in basic and then one power cycle later and we got the rainbow text again and it shows pla test is bad which is exactly how this number six chip was acting before i never showed the diagnostic test with this cartridge but this is exactly what it does and it will freeze up on the color ram here so if i pop this cartridge out and we turn it on i assume that we're going to see rainbow text look we got rainbow text again so it worked one time very briefly and then it failed again so these results are interesting i think i just need to go through and test the rest of these chips and we'll see if any others have changed from the first time i tested them so the results are not encouraging everything on this column was unaffected with both the first and the second bake that's why there's just two red dots down here with the roms i thought there might be some changes but there actually were no changes i didn't have the cpu in properly when i was testing so all of the results are exactly the same there was a good chip number three i forgot if i mentioned that already it was in my dead parts bin obviously that was erroneous because this worked even on my very first initial test but it's still unchanged so it still works it just it didn't get worse or better blue screen black screen black screen so yeah really no changes between all the bakings there the sid chips were completely unaffected by all of the baking in both rounds and then these are the plas and we had number four that actually started working by glitchy after the first bake and it's now totally dead there was one change in that this one right here um it was black screen originally and that's the first dot right there on the left well this slash there means that it now does something different now just gives a brown screen so it's still not working but it is different so that 200 degree bake at least changed it didn't fix it and then the number six chip which was always rainbow text from the beginning did work perfectly which is why i put a slash there initially it worked and then after the second power cycle it went back to its failure mode before i talk about conclusions there is one more thing i wanted to show flipped upside down here are all the plas that i tested from one to eight and i just wanted to show one thing that was interesting is all of these bad chips come from the korea factory commodore was making ships in lots of places taiwan hong kong i think even the u.s potentially but all of these bad ones are from korea so this clearly is not a big enough sample size to determine that it's all the korean ships that are bad in fact i asked isaiah 8 dwf to look at his bad place he's got three and i think only one of his bad ones come from korea and the other two were from another factory so anyways but all of these are from 83 and 84 and that is what i have found that is this date range of pla chip that are the most unreliable and maybe the korea thing has something to do with it too but who knows all right time for some conclusions did baking chips in the toaster oven actually help well according to my data it didn't i mean it seems to have had some effects because we had a little bit of improvement in one chip but then that chip actually died on the second bake although i ramped up the temperature it might be possible that it's commodore's fabrication process the way they made their chips that doesn't respond well to the heat treatment and that other manufacturers perhaps would but unfortunately i don't have enough bad chips to test that particular hypothesis both daniel and i had very limited well basically no success in baking our chips and reviving any so i'm not really going to recommend that you go out and try this with your own dead chips of course they're already dead so you're not going to lose anything by putting them in the oven other than maybe stinking up your house and potentially contaminating the inside of your oven so keep that in mind don't i'm not advocating and putting any chips in an oven you cook food in so you might want to run off to a thrift store and pick up a very cheap toaster oven even one that without a fancy temperature control like this and then just use a simple oven thermometer to figure out the internal temperature before you put your chips in there and then use your watch to time the duration considering how many bad commodore chips i have especially the fact that i have seven bad sid chips i was really trying to be optimistic with this process to hope that it would actually revive a couple of my chips but unfortunately as we've seen didn't really work for me so if you go ahead and try this process yourself and you have more success i would love to hear about it in the comments section below or of course maybe you can post a video about it yourself to show this process working to maybe help other people but for now i'm going to call this particular process not effective on commodore chips and your mileage may vary and finally before i end this video i want to thank danielle for sending me the original idea isaiah did dwf for letting me bounce some ideas off him and of course tyler here in portland for letting me borrow his awesome temperature controlled toaster oven modded reflow oven thingamajiggy here so that's gonna be it if you enjoyed this video i'd appreciate a thumbs up but if you didn't you know what to do hit that thumbs down button subscribe to my channel and hit that little notification icon if you want to get notified on your phone when i post new videos and finally put your comments and your suggestions in the comment section below i really appreciate it when you do that and that's gonna be it stay healthy stay safe and i'll see you next time goodbye [Music] you

Channel: Adrian's Digital Basement

Views: 88,597

Keywords: retro computer, sid chip, reflow oven, c64, commodore 64, bad pla

Id: QxeUJbIubHA

Length: 26min 0sec (1560 seconds)

Published: Sat Dec 05 2020