Can we fix bad chips ... in the oven?
ฝัง
- เผยแพร่เมื่อ 30 ก.ย. 2024
- I was given some information about a very unusual way to possibly fix bad chips. Let's try some experiments and see if this actually works to revive some bad Commodore/MOS chips.
--- Video Links
Controleo3 with Reflow Oven Build Kit
whizoo.com
Toaster Oven:
“BLACK+DECKER 4-Slice Toaster Oven, TO1303SB”
PAL 128 bad VIC-IIe:
• Waiting 35 years for a...
Link to Daniel's GAL PLA project
• Can you replace your C...
www.freepascal...
Electromigration Blog Post and information:
kirchhof.tech/Electromigration.html
en.wikipedia.o...
Time for another Macintosh repair video. This Mac Color Classic has a big problem -- too big perhaps?
--- Tools
Deoxit D5:
amzn.to/2VvOKy1
store.caig.com/...
Jonard Tools EX-2 Chip Extractor:
amzn.to/2VazxDS
www.jonard.com...
Wiha Chip Lifter:
amzn.to/3a9ftWw
www.wihatools....
O-Ring Pick Set: (I use these to lift chips off boards)
amzn.to/3a9x54J
Elenco Electronics LP-560 Logic Probe:
amzn.to/2VrT5lW
Hakko FR301 Desoldering Iron:
amzn.to/2ye6xC0
Rigol DS1054Z Four Channel Oscilloscope:
www.rigolna.co...
Head Worn Magnifying Goggles / Dual Lens Flip-In Head Magnifier:
amzn.to/3adRbuy
TL866II Plus Chip Tester and EPROM programmer: (The MiniPro)
amzn.to/2wG4tlP
www.aliexpress...
TS100 Soldering Iron:
amzn.to/2K36dJ5
www.ebay.com/i...
EEVBlog 121GW Multimeter:
www.eevblog.co...
DSLogic Basic Logic Analyzer:
amzn.to/2RDSDQw
www.ebay.com/i...
Magnetic Screw Holder:
amzn.to/3b8LOhG
www.harborfrei...
Universal ZIP sockets: (clones, used on my ZIF-64 test machine)
www.ebay.com/i...
RetroTink 2X Upconverter: (to hook up something like a C64 to HDMI)
www.retrotink.com/
Plato (Clone) Side Cutters: (order five)
www.ebay.com/i...
Heat Sinks:
www.aliexpress...
Little squeezy bottles: (available elsewhere too)
amzn.to/3b8LOOI
--- Links
My GitHub repository:
github.com/mis...
Commodore Computer Club / Vancouver, WA - Portland, OR - PDX Commodore Users Group
www.commodorec...
--- Instructional videos
My video on damage-free chip removal:
• How to remove chips wi...
--- Music
Intro music and other tracks by:
Nathan Divino
@itsnathandivino
Outro Music:
Abyss by | e s c p | escp-music.ban...
Music promoted by www.free-stock...
Attribution 4.0 International (CC BY 4.0)
creativecommon...
I'd keep the rainbow text chip for Xmas and birthdays. :)
You could modify the BASIC rom to do that, or make a 4066 module that screws it up. Or replace video ram with a microcontroller..
I bet that chip went over to the LGBTQ side...
I'd say that while baking them for longer might've been a good idea, increasing the temperature most likely wasn't. From the point of view of electronics, the difference between 150*C and 200*C is that the latter is above the melting point of solder. Now, I'm not sure if these old-school ICs contain any solder internally, however even if not, it's not impossible there are components in there that are affected negatively by such a high temperature. If amongst the not yet baked chips you still have some that are even remotely worthwhile, I would try baking them longer, but not going above 150*C.
When performing experiments like this, you should only change one variable at a time, as well. I would have kept the temperature the same, and just baked them for longer.
The Enrichment Center is required to remind you that chips will be baked and then there will be cake.
The cake is a lie
The cake is a lie
The cake is a lie
The cake is a lie
I bake chips in the oven all the time, they are just usually the potato kind.
You should listen to that Weird Al Yankovic song. It's all about the Pentiums, baby.
I think he says I have a 1 GHz CPU. What kind of chip do you have under the hood, a Doritos?
Any flashy effects from those?
Angry Louis Rossmann hisses on the background
So you’re saying his PPBus is hot?
Well these arent flipchips (which was louise's main argument against baking cards) so it should be ok.
The dead parts bin for you is like a hunter's prize trophy room. Each chip has a story to tell.
I shove my RAM and Cache DIPS into my drywall in my basement above my workbench. There's even a 386 CPU in there. One day, if I get enough, I want to arrange them in a 12x12 tile and install it somewhere on my floor.
Mmm, baked chips. You even have the DIP as well!
I'll see myself out now
pun intended here, I guess
Ba-dum tssssssss
SID will happily eat those chips with DIP
This comment deserves more than 38 likes surely?
I knew that was coming sooner or later!
How about chip off the old block as well.
We chip you chip we all chip for something.
I'll chip in on that one.
When the chips are down.
When all the chips are on the table or in the oven in this case.
Ziff socket da a chip a da.
Not so obscure Disney reference there.
I know there could be haters on that one but just I think what you want it's a matter of personal opinion of course just like any comment should be taken there's always going to be those people or that person I don't think I've seen you at least one channel and seen one of those people at least a few times!
And then there's that person that tries to find something bad every single comment even when there is one word or even a? Or something like that don't be that guy the world is bad enough now as it is if you know what I'm saying need I say more that answer is no
Didn't know about this method for fixing chips, interesting! What I can recommend for destroying them is putting them in a microwave. ;)
Better yet if all else fails get a bigger hammer or a hammermill either one.
Either or a piece of drill Rod or even just a metal rod about 1/8 inch or so in diameter lamp in a drill press preferably more like a high-speed router motor or Dremel or perhaps a PC board drill machine of some description lower into the center of the ship housing keep going until said metal rod comes out of other side there you go non-working chip very effective for destroying a ship in a piece of equipment that you have to go gree to destroy before salvaging and also take multiple cuts at random angles across circuit boards and anything left in a piece more than a few inches It's Hammer Time. Pretty much yes that would do it yeah I had a few of those Savages more than a few over the days that I used to more than a few over over the old years I used to do stuff like that.
Yes had to sign an agreement that whatever it was had to be destroyed in a way that it could never be resold put back together excetera or at least anything that could be reused in that equipment to repair 04 whatever or be reused in its original application yep sometimes had a little fun with it to of course.
One-time put about 50 some odd bowls possibly closer to do you need to a hundred or more I crossed the B 4.5 volts or upwards of a voltage at quite a few amps yep nice big let out the Magic Smoke.
What else is fun take any component that's read 12 volts as long as it does not provide a direct short when it goes 220 / 208 instead of anywhere from lesson 5 to Route 44 voles or maybe more DC nice Pop Fizz bang possibly even a topless chip yep that happened once that was kind of cool didn't expect it again it was something that had to be destroyed first made non operational permanently did this of course after Salvage what was allowed to be that is.
Also I do remember sometimes if you run enough voltage sometimes even as little as 9 or 12 volts to some LEDs of course without pumping resistors sometimes either you get a nice physical a pop or crack little crack in the ID and of course Magic Smoke and other times half the LED would be blown off or the end of the old lady would be actually blown off completely.
I don't exactly remember how I discovered this thing happening once but I discovered if a magnetic Reed switch a magnet was placed just right spot I think I was trying to do a latching magnetic switch with the opposite polarity to release it if I'm not mistaken when I ran across this?
But under the right conditions in the circuit although this won't go for long obviously because of disintegrated and white of her contacts of course what if it's up just right and the right conditions happen godanriver how or what or how or whatever the contacts would start oscillating what are off rapidly and yes quite a bit of blue light as well but basically would act like just a regular relay as a buzzer except there was no coil involved it was just because of magnetism and all it's possible that the contacts got hot enough that they temporarily lose their magnetism for Split Second but I don't know exactly what was causing it but it would act like a relay or a buzzard and oh yeah a lot a lot of radio frequency interference add just up until the contacts were no good was one of those for the weird things that I discovered by accident it's possible I had problems with the circuitry and some of that contri don't know if I was able to reproduce this or not didn't have me Maniac switches I want the waste trying it though so yeah it was kind of expensive at that time. I only had a handful of them probably.
might destroy the microwave! seriously, pins will get hot, possibly spark and melt through plastic in the interior. Or so I was told...
Good way to destroy SD cards when you hear heavy knocks at your door... lol
Be careful... If there is not enough wave absorbing material in the microwave too much of the rays may go back into the emitter and destroy it. Melting through the bottom - I think that's no danger as usually the turntable is made of glass and the objects are far away from the walls which are usually made of steel too (often painted white).
Always at least put a glass of water with a spoon in it into the microwave when running it.
The spoon is to prevent delayed cooking. When there is no object in the water it may "explode" as the boiling bubbles may not form at 100°C and the boiling may actually start at a much higher temperature.
Yeah you were talking about superheated water in the microwave that Yoda II that gets Disturbed basically explodes instead of just boiling.
Yes can confirm this is a thing I know someone that it happened to fortunately no injury they happened to jar to microwave and that's what caused the water to go instead of when they tried to take it out why that happened to someone instead of what could've under the right circumstances.
Also this is covered in MythBusters as well we actually had a thing about microwave ovens in science class as well as well as I did some Demos in class you know experiments for extra credit although I didn't need it!
I did get a lot of extra credit in a lot of classes sometimes just talk to the teacher but something interesting and they said you know that could be an extra credit project even where you would normally not do projects or things like that in school classes.
Also found out in shop class when I was doing personal projects which I had access to the facilities and or repairs for anyting even if Donna shop or late in Hindi I was getting extra credit kind of surprised but still cool.
The very first (2708) EPROM's could only be programmed, and erased a couple of times, after which erasing (with UV light) didn't work (sufficiently). The solution was to bake them in an oven, after which erasing them worked again (for a couple of times).
Supposedly Intel considered doing the erase process with x rays and then a high temp bake to reverse that damage, but it was decided the quartz window and a germicidal lamp was easier to deal with.
I recently read about flash memory that incorporates built-in heating elements to anneal the memory cells, making them work again after they've worn out. I think they got something like a 10× increase in lifespan from that. It's not the same failure mode (crystal defects in the insulator vs. electromigration), but the same fix works on it.
@@IDoNotLikeHandlesOnYT I read about thermally erased flash memory as a concept, but to my knowledge it's never been commercialized. Planned obsolescence much?
@@nyanpasu64 maybe they couldn’t get it to work?
There are some ICs that are classified as moisture sensitive, and which have time limits from the time they come out of the sealed packaging to the time they are soldered to the board, and if that window is exceeded, then they must be baked for 24 hours at 125C / 250F and returned to sealed packaging.
That is a bunch of BS! (black screens) LOL
LMFAO
@securitycountercheck how about this either one or the other it still works out so but still give you a chuckle and seeing both of these so you're my book either way it's still would be BS.
Or as I always said BC is I didn't use that word I was a good kid.
Raised very well but not spoiled even though I was an only child!
My parents were excellent and kind gentle what parents should be!
@securitycountercheck Bull? The Bailiff in Night Court?
Have a look at the troubleshooting guide from Commodore ..... most of the sympthoms are BS LOL and the purposed solution is to to check 70% of the ICs. Lovely.
Can we all admire the adaption of that oven? That is pretty cool!
my views certainly always been that if you're writing a chip off as dead anyway, chucking them in the oven and seeing wth happens is no problem, what have you got to lose? they get more dead xD
Yeah exactly my thoughts as well.
Just like trying to revive dead hard drives at least two point recover data or a CD that's scratched Beyond usability xcetera because chances are there's no hope in recovery it professionally if you don't have the budget but you might give it a try sometime there's all kinds of methods summer crazy and others not so much have you ever heard about trying to freeze both CDs and hard drives but never tried any of it but if it's dead anyways or it's a CD that's scratched do you know where why not
@@aaronbrandenburg2441 well with discs i have a polishing machine, i just polish it repeatedly until it either works or theirs nothing left to be sure.
@@SparksNZeros CDs you need to fill the scratches and then polish, which are like grooves if viewed under a microscope. There is a putty intended for eyeglasses that works great with fixing scratched CD's, it's been 20 years but there should be a similar product available still. If the reflective (label) side is scratched though, it's game over.
Have you tried with cinnamon?. Everything gets better with cinnamon ;-)
Yuck
How about cinnamon and spice even better
I’ve heard cayenne and nutmeg are other well known keys to perfection. 😁
Especially superglue. th-cam.com/video/7XYPnHODt1E/w-d-xo.html
@@mal2ksc Awesome
You got a free extra voice on your SID chip 😂
i have actually repeared some motherboards of laptops with the oven indeed 170/180C for half an hour and they are running now for couple years faultless :D
There's a really interesting video here on TH-cam by Bil Herd (one of the engineers who worked at Commodore on the C64) called "Component Aging and Heat" where he talks about the effects of heat on chip reliability, but he also goes into a lot of detail about how chips can be damaged by improper ESD procedures. The damage doesn't necessarily mean that the chip will be immediately broken but it may make it more fragile in a way that shortens its lifetime. So it's entirely possible that those Korean chips were manufactured in exactly the same way but weren't handled properly during manufacturing. Check the video out if you haven't seen it, you'll probably find it interesting.
You really think the Koreans knows all the complex steps of how to manufacture these ICs, but at the same time don't know how to handle them properly? Speaking about Bill Herd, MOS/Commodore had an unusually primitive process and problems with contamination and low precision, that's the primary reason why many ICs in Commodore computers frequently fail.
@@herrbonk3635 I’m assuming that’s how he knew how they could be made more fragile/brittle by mishandling!
And no, of course as a matter of course they knew that they were doing, but it’s always possible a small batch had an accident, but then tested good and was sold. Failing after 30-40 years is nowhere near “defective” for sales purposes, given the projected lifetimes of the machines.
@@herrbonk3635 There's a large cost associated with proper handling of ESD sensitive devices, it could be some upper management people that though that skimping on handling the chips properly was a good idea, as long as the chips handles the QA test, they can break later...
Computer guys: You need to keep heat from your chips to keep them from damaging them!!!
Too late, the chips are already dead...
Computer guys: Well, you better heat them up to save them then!!!
:-) Great vid... Really interesting results.
I guess it's sort of like putting yellowed plastic cases in the sun for bleaching.
It does sound weird, when you leave out the "biased" / "non biased" (electrical current / charge applied or not applied) info tid bit. Funny how 'details' can be important to one's perspective or understanding.
Hello Adrian, I was wondering if it would make a difference if you put the chips in with the legs up so they are electrically isolated?
My guess is that most of those chips got bad due to a short circuit which i think you can't fix by baking them. Also it's better to understand what problems baking can cure. So you also need to know what has happened with those chips in the first place.
@@Madrrrrrrrrrrr I don't expect it to make a difference at all, but it never hurts to try
Today on Adrian’s Digital Kitchen:
It’s freshly baked chocolate SID and PLA chips!
Soon on Adrian's Digital 420 Kitchen - Edible chips with some secret sauce...
I wonder if it's possible to decap those dead chips and see where the damage is that causes them to fail?
I doubt it. If i understand correctly its an alloying process, so instead of semiconductor-semiconductor, you get an alloy at the junction and that disrupts the function of the part until its unusable. Thats why silicon diodes fail in a short circuit because you have a P and a N touching but when they alloy you get a neutral piece of silicon so it just conducts because it cant generate the depletion layer to stop the flow.
Electromigration causes atoms to shift, so overtime they mix with whatever was at the other side causing problems like leakage in diodes or making transistors work outside their specifications
@@laharl2k Damn. That's a bummer man. But it's interesting to know
@@laharl2k So basically an internal short.
@@parad0xheart
sure, but you are still talking about 10nm process chips....like its quite hard to reverse engineer that without super high precision equipment. And you shouldnt see much of a difference but everything just being a few atoms thick, not sure if that would hold true at that scale. Maybe it so prominent vs normal transistors that the issue becomes aparent at "plain sight"
@@laharl2k at this point in history (late 70’s early 80’s) all that was required for reverse engineering was an optical microscope and submicron tech was a dream, and a distant one at that. You are correct in that Modern chips are practically impossible to reverse engineer. Hopefully I understood your comment properly, feel free to correct me if I didn’t!
I love the Christmas Effect PLA.. its like a secret built in feature !!!
Well, I've had chips cooked in the oven before, but I don't think I'll try this recipe, they look a tad too crunchy... :P
They call crisps chips in the US. So it's more likely they would dip them in salsa :)
@@6581punk by the way that brings up something interesting and not having to do with computer chips actually.
but actual quote on quote what people would call chips.
The sea is most chips here in the United States are actually called chips as far as I'm aware at least they can be cold chips.
But where it gets interesting is Pringles.
Because of how they are made dehydrated potatoes excetera excetera excetera and so forth.
Evidently they do not meet the definition of a chip according to the regulations in some way or another.
Because late they are produced and what they're made from evidently they cannot be legally called chips.
So that's why Pringles are actually referred to as potato crisps.
Of course I'm glossing over a lot of things and just summarizing but that's from what I understand the reason why and that's why they're not called chips on the packaging.
My guess is if they did it would be called false advertising but it's weird how things are with the step thing.
At least sometimes.
It's kind of like an old some adult beverages cannot be legally called one thing unless you're made a certain way if you know what I'm saying.
I don't know enough about that and to actually even say anything about it have any accuracy since I'm not a connoisseur of that stuff and never even had one drop personal choice you know.
Also along the lines of computer chips years and years and years ago even well before anyone had internet at least at home or even access to it anywhere I'd say maybe if you're military up which is where it came from.
Then. Worse not internet as we know it or probably even before Phil knew what was what sign
There was some stuff out there about the potential to repair computer chips Autumn molecular level using some process that could actually go in and somehow be able to normally make or break connections Etc it was so long ago I don't even remember but I just remember hearing about this possibly an article somewhere who knows this was so long ago I barely even can remember it at all. Just remember here is seeing it somewhere or hearing it actually probably both.
I'm pretty sure it was some sort of laser process but is that may be inaccurate but evidently that was a thing at one time I'm pretty sure was just potential I don't know if it actually have been done but who knows it was so long ago when even know and never leak or anything about it after that so who knows
@@6581punk I feel bad for the poor British man who ordered a Burger and Chips in a US bar and they have him a large bowl of corn chips with his burger.
The roof of your mouth thanks you. You thought Doritos were bad that way?
not for Bender
Disappointing. It would have been amazing to see the SID chips come back to life since they are getting so rare.
Check out Jan Beta's latest vid, he tries out a modern replacement with encouraging results ... th-cam.com/video/ZtpRdVTuHQw/w-d-xo.html
www.retrocomp.cz/produkt?id=2
@@gb7767 perhaps some wave put a link two. Video in here somewhere please I'm sure other people would appreciate that because as you know newest does not stay new for a walk and I course daily by the time somebody sees his comments or yours Toby long long time ago if you know I'm saying so wink would be appreciated I'm sure I'm not the only one that would appreciate it
@@aaronbrandenburg2441 done, the ARMSID is a available at www.retrocomp.cz
Maybe if the tray was electrically insulated and the atmosphere was flooded with nitrogen, more testing required. Paint on a thin layer of "waterglass" (aka Sodium Silicate) and you have a heatproof electrical isolation.
This was a really neat experiment. As I'm sure you know, successful chip companies revise their manufacturing process periodically. Intel famously calls this their tick-tock cycle, where they alternate between process improvements and new chip designs. Commodore did not do that. MOS Technology had a process that was pretty advanced for the late 70s (at least it had one manufacturing advantage that Intel and Motorola lacked), but they only modernized their process once. 6500-series chips could potentially behave differently from 8500-series chips because the 8500-series chips were made on a newer manufacturing process. So it wouldn't surprise me if chips made by other random companies in 1983-84 responded a bit differently from Commodore chips. I love Commodore, but chemistry wasn't their strong suit and ultimately I think it was their undoing. Atari bought chips from pretty much any company who had capacity to make the chips they needed, so I would expect dead chips that came from Atari machines to potentially behave differently from chips from Commodores.
Also, regarding your PLAs. While Commodore did packaging lots of places, the wafers only ever came from one of two places. They had a fab in California that they closed around 1985, and they had their notorious fab in Pennsylvania that was active from the 1970s until their bankruptcy, and that briefly came back from the dead after Commodore went under. The biggest problem with the PLA was the chemistry Commodore used to make them, so I wouldn't expect the factory where the packaging occurred to make much difference. Maybe a factory that did sloppy work would make the chips fail sooner, but since the wafers themselves were fatally flawed, eventually any PLA can fail just from entropy.
Good points and I didn't realize the wafers were from two places only. Good to know! It's odd how most MOS chips made at the same time as the PLA are fine but the PLA is so unbelievably unreliable now.
@@adriansdigitalbasement in researching this myself (I found this fascinating), I found guidance from your least favorite vintage chipmaker, Micron Technology, on baking chips to fix them. I thought you'd find that amusing. Interestingly, they repeat JEDEC standards, which suggests baking chips at 125 degrees C for 8 hours in most cases, or in some rare cases, 24 hours. The reason for this is to bake off accumulated moisture. Time to extract those hated MT memory chips from the dead parts bin?
Back in the Late 70's and early 80's You could restore Eprom chips (2708). These were ultraviolet eraseable Programmable Roms. 450 degrees fro 30 minutes. Most were restored.
www.classiccmp.org/pipermail/cctalk/2017-December/036588.html This is not my post.
They are usign Fahrenheit in that link, 450 deg C is likely deadly for a chip.
Yes, it is 450 F.
Interesting video!!! Funny thing afterwards, I got youtube suggestions to watch baking christmas cookies! 😁
Silver sharpy markers will stay visible after baking them.
In the early 90s acquired (for little or no money - I was a poor student) a whole bunch of 20MB Seagate ST125 IDE drives. They were notorious for what was called "sticktion" where the heads stuck to the platters. The drives usually got tossed - I got hold of them and baked them in the oven. That usually solved the "sticktion" problem and I had a collection of these drives that kept working for many years - some still work in my vintage collection!
You should have made this a collab with Retro Recipes... 😂
Baking chips is usually done if you want to solder them professionally ("reflowing"). Chips are protected against moisture but if you get the chips our of the packaging too soon you should do that to gently remove moisture otherwise it could create some pockets of steam inside and damage the chip. But the procedure is much less violent but longer: you gently increase the temperature to 105C for hours then finish for 24h at 125C...
@4:19, don't bake that chip. That's no faulty chip, THAT'S A FEATURE!
Hello, I'm working in an electronic manufacturer. I can confirm that a lot of time heat shock can "cure" bad components. Part of my job is to investigate test failed products, and it almost involves cross-check, which means we dismount a suspected component, replaced a new one, and the suspected component mounted into a known good board. So a few times, both sample (the failed and the good one) is working fine after the cross-check. We don't know the reason why. Also we don't know how long it will remains good, maybe after a few days of use, it would fail again. The hot air station is set at around 350-400C because of the PCB heat dissipation. So you might try to go higher temperature, but just for 5 minutes or so.
they go great with BBQ sauce
No, chips go better with salsa
@@eatportchops *vinegar and fish
MOS (Mostly Obsolete Semiconductor).
I always thought it stood for Metal Oxide Semiconductor
@@officermeowmeowfuzzyface4408 Joke, It was a joke
Great video Adrian. I’ve seen this work on motherboards before but not chips.
This is a very reliable fix for the PS3 red ring of death..
As a motherboard reflux, I had place a motherboard in a preheated oven, then brought temp up to about 400° then allow it to cool in the oven overnight.. it was successful but repair didn't last
Yeah i can imagine that it might work on whole circuit boareds where some of the solder joints are a bit iffy; solitary chips on the other hand, not so much.
Bring an ASUS Motherboard and a Grafikcard XFX back to life again. For 1 Hour at 110 C ! 2 of 6 tries in the early 2000 !!!
I’ve seen few YT vids with people hairdrying the motherPCB of dead smartphones (or bootup loop). They fixed it for half an hour, just enough time to pull the data.
Yeah, that tends to be related to solder, very different from many IC problems.
Always been wondering what was going to happen with the dead parts bin!
I second that and third it and so on I was wondering if he had more than one Dead part Ben even to tell you the truth.
Any adult Lego fans out there might know about that one person on TH-cam brick tzar and the Boneyard AKA broken Lego or damaged Lego parts he handles same kind of thing.
But no not in the oven that would be a very very bad idea but I was referring to the dead Parts Ben.
Yes I know I spelled that wrong but Google before you say anything about it.
But but he did say it is videos brakes aren't any of said he never throws away a bad or damaged broken Xterra Lego Park although sometimes they can be useful for some things occasionally or if you're not purest modifications which I'm not 100% of purest but I tried to be closed when it can be done with what I have in terms of parts and what exists in the Lego Universe but there are some things you're out price range or other things that would not be practical but maybe in the future would be possible either with you parts or perhaps getting a hold of Parts they don't make anymore AKA perhaps Bricklink might be of use and no I do not have an account yet waiting for approval.
BEST EASY BAKE OVEN EVER!
very interesting video. I watched it without scrolling through to keep the thrill up for me. Somehow i was hoping to get a surprise at the end that this works somehow. But deep inside myself I expected no improvement. very entertaining video. Thank you! Cheers, Peter
Damn, I was really looking forward to the RIFA caps being put in the oven. What a letdown. :D
All they need is a good camp fire cooking...
Hotboxing with Rifa
You like to live dangerously, don't you Catriona?
@@BertGrink Totally :)
@@catriona_drummond Kewl! 😁
I had what I totally thought was the Christmas tree PLA on my '84 64, but it turned out to be the color ram failing in a weird way. In the meantime I had swapped out the PLA with a PLAtinum PLA replacement. So I replaced the color ram and actually have an extra good in case my Silver Label '82 C64 has a problem. I really hope it never does though, since it has all original '82 dated chips.
Very interesting experiment! But to be honest, high temperature just deteriorates the semi-conductor structure, I think. As far as I know, (pre)baking is needed only for re-flow process because moisture in package possibly explodes if you do re-flow solder with sudden, massive temperature change.
Zymos was chip and chipset manufacture for intel 8086 to 80286 computers. I have a 80286 system with one of their chipsets.
I wonder if that Zymos was the same company which was involved in the development of pacemaker chips and which is somewhat related to the development of the Amiga.
I'm sort of new to this degree of fixing bad chips, can someone explain in a quick tl;dr why baking works? My only reference point is for CPUs/GPUs with bad or faulty solder points.
I actually have a GAL PLA. I don't know if it was necessary though. My issue was the system was running at 3.5v with the U2+ installed. Turns out a MIDI cable is not suitable for powering a C64. Fortunately I bypassed the issue by externally powering the U2+. I'm going to buy a black brick of death just to salvage the cable (I refuse to solder a DIN). But yeah my colors were all screwy not unlike that PLA you showed (but not as bad). I replaced the PLA and the color ram. No change. Then I noticed the voltage was super low. I'm surprised it worked at all. lol So I still have the MOS PLA. It's probably fine. I'll leave the GAL in there for now. lol
99.8% of the time, the substrate of the chip itself is damaged. And baking at around 120°C can make them work again, for a few weeks, before they fail once again.
If 120°C does not work, and you have to go to ~220°C and if it actually works, congrats. You won the lottery because it was a 0.2% soldering issue.
People need to understand, 99.8% of the time, baking is pointless because its NOT a soldering issue! The chip itself was faulty from the beginning and is damaged goods, now and in the future. It cannot be repaired, not even by the Chip Manufacturer itself! And only shady eBay sellers bake their broken cards to sell them off!
This is one reason why i will never buy a used $300 GPU! Its just stupid and a waste to invest so much money on a gamble.
I have to mention that this "fix" while sometimes successful isn't truly a fix. Most of the time the process of heating the chip up temporarily repairs the internal traces inside the chip, and will almost certainly die within days or a few weeks, at most a couple of months. This reminds me of the botch repairs on the Xbox 360 consoles to fix RROD issue. I repair newer consoles for a living and I would never recommend baking a chip to fix anything no matter how small the issue is
@Mr Guru no people "think" they are fixing the problem by reflowing the solder under the chip, but in reality it is the "baking" process which temporarily fixes the problem, which is why wrapping a console in a towel used to work. The solder never got to melting point when people used to wrap the console in a towel and simply hit roughly 100⁰C - 120⁰C (I never checked exact temps the 360 reached with this method). That problem was hardly ever down to the solder balls, and was a defective chip because of the manufacturing process
Back in the day I got an extra 6 months of life from my dead 2008 Macbook Pro by baking the motherboard at 190°C. It was used regularly for After Effects rendering so I guess the solder joints had become brittle and cracked from the constant temperature fluctuations. A question re: your process - Isn't the oven supposed to be empty when preheating? You put the chips in once the correct temperature's been reached? Or is that just when making cookies :)
200 degrees is too hot. You're probably letting other things diffuse inside the chips, besides any small impurities or defects that electro-migrated before. You're probably killing the IC transistor junctions by letting any dopants diffuse too much -- you're "blurring" the interface between the junctions. Electro-migration occurs because you have a voltage driving the impurities and defects to move, but heat without any bias does a different thing, which is just ordinary thermal diffusion, which because it's a random process tends to partly nullify the non-random migration caused by the electric fields inside the chip during operation. But too high a temperature may cause too strong a thermal diffusion, therefore causing damage of a different sort from that produced by electro-migration.
(I'm a physicist, so this is just my too cents -- my explanation is just a tentative one.)
I'm curious if placing the chips with the pins touching a conductive surface like your aluminum plate might have had any effects on the outcome?
I wondered about that too
I mean, theoretically that could be an issue, but it's really not likely. The bigger concern would be static discharge into the pins, but I honestly think a few of these died due to that, so...
The aluminium plate means that all pins have the same potential, which fits that description of undoing electro migration. So I think the method was fine.
@@MikeStavola Where are you getting static discharge from in this process?
@@jong2359 It was a figure of speech. It's like saying, "I'm more worried about being pecked to death by chickens than being struck by a meteor."
An actual concern I'd have with this process is a lack of drying before the oven. If you ever watch an encapsulated chip explode in a reflow oven due to trapped moisture, you'll know why.
No joke... many years ago a client of mine had bought a cheap Dell laptop from ... Walmart. Well, it lasted about 3 months and stopped booting. After a bit of Googling I found out that someone had noticed that bad soldering was leading to the heat from the machine warping the board and opening the connections. Someone had mentioned they had luck wrapping the laptop in a blanket and putting it in an oven for an hour at 150º. We did't have much to lose as it was already not booting, so I gave it a shot, and it worked... for a few weeks anyhow.
@Mr Guru As you said in another comment chain: Solder doesn’t melt at 150°C.
@Jeffery Hahn Baking chips is a temporary fix so I’m not surprised that it broke again after a few weeks.
FYI for tape media: Baking is SOP for remediating magnetic tape media. I used to do so all the time when turning 40-50 year old 9-track tapes. The binder between the film and the ferrous layer would become sticky. Red label Memorex were the worst offenders as the film and the magnetic media would often delaminate entirely when passing over the read heads. www.heeltoe.com/retro/tapes/
Try baking a good chip. The baking might kill it. Imagine stressing a good chip to that kind of heat. I can’t see why this baking theory could possibly help.
Ya, I'll just go to the grocery store and get some Lays Baked Chips, much easier and they taste better
still don't believe it will fix, "might" fix some specific fails but not all, not by a long shot, keep in mind that we are talking about silica and gold
That's interesting. I was also thinking maybe it wasn't hot enough, but I guess that wasn't the case. No real effect. Maybe it would be an interesting follow-up to try freezing them instead, for no apparent reason other than to experiment.
Oh, the flashing rainbow effect was rather cool, btw. Keep it :)
I once did the opposite to rescue some precious files from a dead harddrive - placed it in a freezer, then had ~4 minutes to boot from it and save the files. had to do it a couple of times to save all the important stuff. but it worked.
The only reason to bake ICs is to remove moisture before machine soldering so the packages aren't damaged by steam. I guess it was the PS2/Xbox 360 oven fad that started this but stop baking your electronics and learn to make cookies if you want to use your oven!
according to bill herd, they used too much boron for the faulty PLAs, if i remember correct.
alot of them are not 100% dead as you found out. They are more 'analog' now I guess. Some are slow with slow rising or falling edges at the output.
with a scope and different baking methods you could find out what would be the best method by comparing the signals :-)
I tried 180 ºC and worked for me in some chips like Via brand. So, I think 200 ºC is too much.
Back in the '90 sometime me and my friends used to freeze some hard drives and then they back to life and allow us to make a quick backup from whatever is inside.
Perhaps freeze the chips should have some effect either.
Or perhaps you could make another video about this cold method to revive HDD.
Thanks for this excellent video.
Would there be any advantage in putting the chip die side down on the plate, rather than the pins? Also does the method specify the chips should be heated at the same rate as the oven, rather than preheating the oven, then putting them in (as one would when cooking)
I'm noticing a trend with your dead parts bin, sounds like MOS stuff was sort of crap. I understand why they're no longer in business. Commodore probably would have been better off going with a third-party chip fab. So many dead SIDs and PLAs.
Well we are seeing thirty years later after unknown usesge and or storage conditions.
MOS had way higher yields and thus lower costs than other fabs. And MOS survived Commodore -- what they couldn't survive were the cleanup costs of all the groundwater contamination at their main facility.
Interesting concept, I would've thought temperatures that high would only be a detriment to the chips. However I just checked a couple of datasheets and they list the maximum storage temperature of 150°C, so I guess they can handle it for at least some time.
I know people bake their graphics cards and such, to reflow the solder and/or fix broken bits inside of the main graphics unit. ICs are in similar design (As far as i know) but miniatured, so i could kind of understand why people would try that. But alas it doesn't work. :P
I kind of want the rainbow text programmable logic array (pla) chip
Interesting experiment, but no. My company build equipment for the telephone industry we would run our systems in a heat room at 50DegC for 72 hours and that would attempt to find chips that had "infant mortality" there was a formula for heat and time that would map back to normal operating temp and years of service.
have you tried the effct on a good chip ? i know somebody repaired a mobo this way, after baking the thing worked for 2 years, the next time again but lived shorter, he tested this for 5 years if i remember well. Upto the point it did not work anymore. I was thinking the heat does something with the bondwires pressing them back in place. i once reflowed a BGA on a Fluke 192 and after that it worked again but I think that was restoring a bad solder ball.
Very interesting, thanks will do some tess my self. I wish I had the gurs to decap ICs and then look under the microscoop what the problem is. like burned bondwires, holes in the die etc. because that can not be cured in any way
Ive never baked chips,but sometimes ive freezed them with cooling spray to fix them. Worked most times with broken tsop chips
Hello Adrian. Thanks for the interesting experiment -- as a retired chip designer (ex PMC-Sierra) I'm a little skeptical about the baking, especially for electromigration. Low temperature solder on the bond wires might get fixed up but my guess at a mechanism for change would be hot carrier injection. Transistors will fail over time because every now and again an electron (or charge carrier) will be especially energetic and slam into the gate material getting trapped and displacing the crystal structure. Heating up the device and then slowly cooling it down would emulate an annealing cycle which *could* free these charge carriers and perhaps allow the crystal structure to reform.
Can "cleaner" current (so from better power supply) slow this process?
@@warrax111 Sorry for taking so long to get back to you! I don't think hot carrier injection (the process) is really affected by variations in the potential across the gate. The process is more that some charge carriers are "hot" in that they have more thermal energy and that there is a distribution of energy across all the charge carriers. Think Maxell-Boltzmann style curves (although I think this is more of a Fermi-Dirac distribution -- details are a bit irrelevant) where some electrons (or ions) are just moving quickly. The odds of this occurring are reduced at lower temperatures, of course.
@@andrewhughes7195 So they should be fine if you just operate the chips at liquid helium temperatures?
@@brianclimbs1509 elevated temperatures are known to shorten device lifetimes. HCI is one of those mechanisms that is temperature dependent. There have been attempts to operate CMOS devices at cryogenic temperatures (liquid nitrogen) but this was for performance not reliability. Liquid Helium would seem … extreme. I suspect that some other effects would come in to play that would prevent to devices from functioning at all. Also, I don’t really think anything can repair damage that is already done.
This goes back to 1988 for me.
My own testing never got above 15% fix, and that was with 74xx chips.
Mostly 7406 out of Commodore 1541 diskdrive would work.
We used to put the chips upside down! While heating.
Reason, is that chips would often die 'hot' with gravity pulling down.
By heating the chips upside down it would reflow back.
Again it is not 100%, but we seem to have had some result.
Love the video.
I have a small pile of dead DRAMs, I'll have to give this a try and see if any of them change.
Yes please do, but stick to 150C.
No idea where my failed 4116 ram chips are or I'd try it as well, as I already have the reflow oven.
I keep thinking you should build a c64 with all partially functioning chips and see what you get.
C64: Misfit Toys Edition
Cuts on the knuckles reveals who's spent time ripping apart computers
You need some fish with those oven chips! Cool episode!
You need to bake the smoke in the chips Back.
Good call not to use an oven for food to bake parts that once had solder and flux on them. Better safe than sorry when it comes to health.
Flux comes from tree sap, I thought. Rosin flux has a danger label as an irritant only, not a poison. Lead in an oven isn't a big deal, just vent the fumes with the range hood. It's the manufacturing chemicals leaching out that is the problem. I used to bake trays of cast "projectiles" in a modified cookie tin (with a heavily perforated base and a bucket handle made from steel wire), hoist them out and dunk the whole tin in a sink of cold water for tempering. Never had any lead poisoning symptoms in the least. But I knew exactly what was in the alloy and the worst was perhaps the arsenic.
Maybe commodore chips would respond better to sous vide?
Ah but was it the baking or your random ESD events that changed the behaviour lol
come on , chips should be fried , not baked
I don't know about individual chips but I have baked dead iMac graphics cards before to great success. Apple, in their wisdom, used rubbish solder in various graphics cards which develop micro cracks over time from being heated and cooled. Popping them in the oven for half an hour at 200 C can often revive them because said solder reflows. I discovered this when I had a poorly iMac and once I'd got back up off the floor after a quote to get it repaired and done some Googling I found out about this process. Unfortunately it doesn't work forever, the cards often fail again because of course the reason for the initial failure is still there - the bad solder, but you just pop it back in the oven and off you go again. the last one I did has lasted over 18 months so far and counting. Many say it doesn't work, but several dead iMacs later I'd beg to differ.
It's like CRT rejuvenation. People say it's temporary and isn't worth doing but I have one Mac CRT still working well after doing this process. So I guess YMMV!
reflowing might be more effective in ram chips. maybe try that before returning the reflow oven? (the reflowinator)
I baked chips in my oven last night. Results were good.
They were yummy.
(Chips are fries in the UK)
And next time we cool Down the Chips 😂
Thanks for the intresdting Video 👍
Thanks for trying it. What happens if you freeze one with CO2 and test it while cold or then room temp?
That's actually good idea. IIRC there was a chip that was getting worse when heated.
Interesting experiment! I love this type of thing.
I think you had the right idea to speculate about what failure modes baking at a high temperature might have an effect on. The failure modes that might occur after this much time include internal corrosion due to moisture wicking in through the plastic case, corrosion due to chemical reactions between the chemicals used in adjoining layers, over voltage, etc. I don't think baking will have any affect on these types of failures.
I did have a PLA once that would start working properly once it warmed up. I guessed that a bonding wire was loose and the increased temp caused enough thermal expansion to make a good enough connection. I think maybe that baking could temporarily 'fix' such a problem but that after after a few heating/cooling cycles it would break again.
Maybe flip the chips so the pin are up ways 🤔
"Fixing" integrated circuits by baking them is actually a somewhat well-known phenomenon -- you can see examples at th-cam.com/video/1AcEt073Uds/w-d-xo.html
* Edit: I should mention that this is at best a temporary fix
I sold IC's for many years, we did something called drypak in much lower temps. New chips baked to dry out all moisture slowly over time, so when the PCB's get wave soldered no moisture means no steam which would escape the ic's; which would crack the lids and destroy very expensive rad tolerant IC's (electro migration repair is a different issue and can also work)
Sell that PLA as the "Christmas PLA", or "Acid PLA" on ebay. 😂🤣
Flip them over? They likely spent their whole time operating in an upright position... maybe get a gravity assist with the heat for the opposite migration?
Might be bad idea. The outside pins are connected with small wires to the silicon in the middle of the ceramic body. If solder is been used for connecting them, at these temperatures it may melt and the wires may fall off.
They might potentially work when cooled down to -20°C or so
intresting idea
Yes, as another experiment, Adrian could use component cooling spray or an inverted can of dust-off.
That toaster oven mod is awesome
I'd love to see if there is any visual change to the dead chips vs the working ones under a microscope. And if heating them also visually changed them. Fascinating video though :)
Adrian,
Great video as always sir.
It would be interesting with your chips to see what happens if you were to try the 150°C for 24 hours to see how that would work out.
What, no chocolate chips?
Next, try deep-frying them!
I feel that baking might reconnect golden wires to the chip.
2:52 I haven't seen that much BS since (insert reference here)
The 2020 election lol
Bs is bull sh*t
... since the last time Biden tried -explaining- saying anything