Chips a la Antoine: an IC chip de-capping recipe without chemicals
ฝัง
- เผยแพร่เมื่อ 7 มี.ค. 2020
- Antoine is back, and demonstrates his simple method to decap plastic encapsulated chips, which does not use any nasty chemicals.
Antoine posts new photos of chips on Twitter, he is @Siliconinsid
Our sponsor for PCBs: www.pcbway.com
Support the team on Patreon: / curiousmarc
Buy shirts on Teespring: teespring.com/stores/curiousm...
Learn more on companion site: www.curiousmarc.com
Contact info: th-cam.com/users/curiousmarca...
Music credits: Fire Prelude by Kevin MacLeod, incompetech.com/music/. At least that’s where I downloaded it originally. But now it downloads something else! Commenters have found it again googling CXNXR - EON. Which is apparently the right song, but not the original author either. So I’m not really sure whom to credit. Let me know if you find out. - วิทยาศาสตร์และเทคโนโลยี
![[LIVE] : ONE ลุมพินี 69 | คู่เอก "กุหลาบดำ vs นาบิล"](http://i.ytimg.com/vi/PNM6U8e_HMo/mqdefault.jpg)
In before copyright strike due to picture of ROM "showing the bits"
Were they the "Naughty bits"?
Also a lot of images we can use to reverse-engineer and replicate those ICs! Great, time to start routing... not :P
This is an unbelievable revelation. The painstaking hours of hot acid digestion, sanding, etc. I've seen people go through...
Everyone's house is going to smell like hot epoxy now.
3....2....1
I just tried this and it works. I can't believe how easy it is! I have tried decapping before using nitric acid and couldn't get it to work. Thank you very much!
i've actually done it with those blob chips on cheap boards, these ones are a lot more softer than these packages, they almost turn into a paste.
When I was a kid my parents owned a telecommunications business. When equipment would become obsolete or could not be repaired my dad would sometimes give me the circuit boards. My mom would have killed me if I did this in the kitchen so I would go outside and use a heat gun to remove the epoxy package off the board. My personal experience is with the right technique with a propane torch you can remove a manufacturers die from the epoxy package. I successfully removed some of Intel's dies as well as Motorola, national semiconductor, Texas instruments, NEC, Hitachi, and many others. I also collected an extensive number of eprom I like the fact they had a window and you could see the die without removing the package. I like watching these videos they bring back a lot of memories.
Propane torch was my initial way but it damages the die significantly and produce very nasty fumes.
Long time ago I tried "decapping" by just hammering the chips, and see if I was able to get the die within, with very little success. For the little pieces I was able to get, I look under my microscope and wondered.
This is more than genius, I can decap again!!!
Little did the inventor of the vacuum tube know that the glass enclosure itself would be the basis of the transistor that later replaced it!
Press X to pay respect to the PlayStation that donated its brains to this amazing demonstration!
I would have LOVED to see the main processor done in this way
I had to try it out even before the video was over. Went 2 out of three on some old Realtek audio chips. Awesome!
When it fails, epoxy does not separate from the die surface... but at least you've removed the bulk of the package which will make chemical decaping a lot easier.
same here ! :D
Wife: "Why does it smell of burnt plastic in here?", "Why are there dribbles of metal on my nice hob?" =)
Another brilliant video. You never go wrong Marc. I can't thank you enough for all your hard work.
Thank you! I have been searching for a good way to do this for YEARS!
Beautiful!! Exact what I have been looking for for a long while
As many have said in the last video, it would be amazing if Antoine uploaded the pictures of the chips on the internet !
I think they may fall under copyright, is the problem.
@@PilkScientist Really? Didn't even know that was a thing
@@RaoulB. some people were mentioning it, they contain ROM data which could be back-engineered from photos. He may also just not have thought about it, though, dunno
This is a cool process and the results are spectacular!
Thank you, its been maddening to find the solution to this problem!
Awesome. I'm already taking photos of mysterious chips for which I can't find datasheets.
I never thought just heat would be enough.
I didn't have a heat gun so I put mine in a fan forced oven set on 260°C. It works a charm, putting them back in the oven a couple of times when they cool down too much.
It makes sense though... it is basically glass/sand packaged in organic stuff :D
One word, even being an electronics techy myself... Amazing! Thank you for sharing.
Great video! I watched/listened to it while...
...annotating a completely different microchip!
That is fascinating! Its like looking in on a tiny, hidden world. Which if course, it is! I've seen close ups of ic internals in the past(many years ago, in electronics magazines, of the 90's).. they were nothing like the clarity here!
WOW! Excited for the next episode :D
Normally fuming nitric acid is used... requires a fume hood, protective gear, safety squints etc. How you convert OTP chips to MTP!
Funny, we used to reclaim ICs from boards by putting them on the hob, turned down low and with an IC extractor, then yank those suckers out. That was over 30 years ago and we were utterly destitute electronics students that hadn't turned to crime or paid work to get components from reputable suppliers 🤪 We noticed that if we held an IC in the burner, it would often crack and a simple twist would reveal the innards. Unfortuately, we didn't have access to a decent microscope at the time, so the best we could see was around 50x I think, I did crack open a Sinclair Spectrum ULA back in the day and had a look at, nothing interesting on it as far as I recall. I just wanted the Z80 and the RAM (sacrilege I know!) for a shoestring microcontroller build. The last IC die I looked at was a T805 Transputer (of INMOS fame) a ceramic package that I put into a board then hit sideways on with a sharp chisel, the cap popped off to reveal the glory inside 🙂
cool, can't wait to do this
That's so simple i had no idea, i thought everyone was using hot acid, very cool thank guys 👍
The only issue I have with this method is that once the Die is exposed all the heat will cause it to start growing oxide layers. If you do this method too slowly you might get some optical alterations on the die because of this. Which could make your Reverse-engineering efforts a bit of a pain.
Interesting de-capping method. Good remark on the entrapped moisture in the chips. Didn't know that.
As always, very interesting!
This was awesome, love looking at the nano world
This is beautiful!
always amazes me the amount of engineering in these little packages that makes all the cool stuff work, yet are totally anonymous and ignored in the beauty of the design, just disposable items which would otherwise never been seen, who depts of engineers who would have worked on their designs, for some one of those tiny chips they worked on was the epoch of their careers.
Fascinating
It's amazing just how little of the package is actually the chip
and these are from '94... now the miniaturization would make the equivalent of all the PS1 in a 2x2 mm die
wow, I had no idea this could be so simple!
Well done
You guys have the best lab
Amazing! The final zoom in views were reminiscent of Google Earth zooming into a city!
Well that's how the "Matrix" looks like if you look closely :-)
Or the infinite zoom of the Mandelbrot set as new "cores" emerge as one zoom in and in and in.
Koyaanisqatsi used dies like this for that effect.
I just tried with a Zilog Z80180, coincidently using the exact same dessoldering gun as Marc’s. I was easily able to reach the die, but it was covered by a layer of packaging material that just wouldn’t come off. RIP Z80Z180…
This is really neat! All of a sudden decapping has become something I can do!
Oh wow! That is cool
Génial, vite la suite ! :-)
*magnific!*
Это на столько невероятно, что даже начинаешь сомневаться в том, что это создал человек )
I cant wait to try this :-)
The ultimate "Don't turn it on! Take it apart!"
.....yes but dave as well as ave will put it back together and try it afterwards, anyone good with manually bonding wires to a die?
Beautiful! 😍
fascinating
Wow, this method is way easier than using acid, might have to try it out.
So you are Marc, the chef is Antoine and me? Marc-Antoine :D Great vid!
Nice video
Part of me was saying "No not a Playstation". On the other hand, if any one is allowed to decap it's you guys. Considering all the hardware you restore and curate. It's also not like there's shortage of Playstation ones .
Back a few decades we would use a revolving whewt stone in water in water to slowly grind away the casse. It was usefulk for multi-layers. The results were good enough for high power microscopes or electron scanning microscopes
how have I not thought of this. This is so much better than using dangerous chemicals. Guess its time to finally make some cool jewelery out of some old chips
It’s just like breaking a chocolate bar and it’s a very fancy looking gas stove. Antoine is a proper chef de cuisine.
For anyone who likes decapsulation videos, may I suggest ElectronUpdate channel, where the guy gives a bit of a tour of what the different structures and patterns are likely to be - ROM, gate array, etc.
He need his own channel!
Very smart...
2:44 Looks like one of the chips has a premonition of what's about to happen. "Well, I'm out of here!" :-)
Its like a whole universe in there, looks like a city from the future!
These high resolution pictures of dies that Antoine makes, are they available online anywhere?
You can get such pictures from siliconpr0n.org/map/ - It would be nice if Antoine submits his pictures there, too
@@DJChol Is there a chance on the release dates on those chips there? I recently dug in our junk box and found two DIP (lol) atMega8's which were a neat choice for a quick build, as I utterly needed a digital thermometer with a remote sensor. Got all my ASM prefabs close-by so an hour's work, what could go wrong? This: one of the chips, a brand new it seems, just didn't flash, and before you ask I long figured all those tricks like jumpstarting your chips by external clock etc (had to use on the second one, which was a used IC fused for external crystal which I didn't need nor want in this tool). So, looking at how the chip in question was paint-marked, not lased, I guess it was some fake junk I want to find the origins of and do some ass-kicking - though I wanted to X-Ray it first, and compare to the photos (as those things have additional marking on the chip itself too).
There is also a large collection of die photos taken by an individual to be found somewhere on commons.wikimedia.org, but I can't find the right keywords to find it.
@@winstonsmith478 I don't know if this the one you are talking about, but there is this user: commons.wikimedia.org/w/index.php?title=Special:ListFiles/Epop&ilshowall=1
11:20 that sounds *magical!*
I'd really like to know the name of the song or where to get it from. Unfortunately it's not the one Marc put into the video description. :-(
Edit: @Q-Set Oldskool mentioned it in another comment: "CXNXR - EON". :-)
A friend found out it is:
*CXNXR - Eon*
I wonder if tile cutting nipper pliers would split the casing perfectly along the seam so the top pops off. They have 2 chisel type jaws that could cut precisely into the seam.
lol.... I guess I have been watching too many delid videos if I can spot a ROM as soon as I see a bare die. #electronupdate All the older lithography is easier to see too though... Always a cool video, thanks Marc!
don´t worry, when you`re out your wife calls her girls to have a barbecue in your lab
Very interesting and seemingly simple method! Does Antoine have a public access site to see his collection of die photographs?
great trick, going to decap my chips now...
antoine is a hidden gem! i understand the man in the labcoat is hopping from one foot to the other?
Looking at those dies is like looking at a city from miles above... pretty cool stuff. BTW, how hot is the air coming out of that hot air desolderer?
I remember doing this a few times in my teen years, but I used a burner instead of a heat gun as I didn't have one back then. This generated tons of black toxic fumes, so I stopped doing it. I still have a drawer full of random chips that I wanted to decap. Somehow it never occurred to me that I can just use a heat gun!
...turn down the heat? 😝 I used to do in my teens too, heat up a board on the hob, extract ICs, found this die splitting technique from being bored and wondering what was inside, it wasn't all that interesting, but then I didn't know then what I know now 😎
@@davidhunt240 I’m a current teen doing it in by basement lol
@@BGTech1 yup, we were all young once 😝 I often wonder how I'm still alive after those years 😂
Antoine - quickest die draw in the world. Not even autofocus can keep up 😊
Wow. I may have to try this.
The Cell has a big beautiful die, but it's a flip chip so I don't know if that technique works as well.
@@SteelSkin667 Will not work, and the beauty will be hidden by an intricate top metal layer on the die that needs to be removed.
That seems a lot safer than when I was just sticking chips in a fire as a kid.
What do you do if the die has a metallization layer on it?
somewhere in a small sheet of note paper i have stored an ic from soviet elektronika wristwatch... it would be fun to know die of that, likely western-cloned-stolen-pirated ic, would look like
i hope i still have it, i was only teen then, it was decades ago, someone's watch broke and i took it apart, eventually discarded everything but the die, which i'm not sure how i managed to decap then but it remained intact... i still don't have any microscope yet to look at it
Watching the dies under the microscope makes me want to go play factorio
1:35 spotted a real Chef ;-)
That's not "Chef"..... It's "Shirriff" :D :D :D :D :D
@@edgeeffect yea I spotted master Ken ofc :D
0:52 : "No judge, I didn't realize that doing this in the kitchen would cause Marc's wife to kill him. It was totally unforeseeable."
Also, I hope you dig around for a 555 die. Curious how small those are.
It's already been done. www.righto.com/2016/02/555-timer-teardown-inside-worlds-most.html
Is there a way to reuse such a decapped chip by connecting wires to it directly? Or they don't work anymore after this procedure?
I have an old motherboard from a vintage computer broken in half and the chip packages are cracked, but your video made me realize that the chips inside might still be intact, so if I could take them out from the plastic, maybe there's still a way to revive them somehow? :J
Please tell me what glue he uses for glueing the dies on the microscope slides.
In an alternative universe Wolf would advertise the versatility of their stoves like this. The image of Gordon Ramsey and your wife giving a big thumbs-up to it would be surreal...
Mark, What do I do if some of the Resin packages does not separate from the top of the die?
Then you can't use this simple technique. You have to thin it down then use some nasty acid to attack the remaining plastic.
does antoine upload photos from this to anywhere?
Cool, is there a way to get those hires photos of the dies? Would be a great poster :-D
Think I'll leave mine in the Packages. They're much easier to solder. You guys are saving the Gold Wires, eaaaa?
"What have you been cooking?"
No simple answer for that
Yeah there is.
Chips
You are cooking microchips in the kitchen... Yes, your wife is a SAINT. Take her out for dinner, she deserves it.
I've done this to a few broken ICs to test a cheap microscope I bought. Just cracked them open with a wire cutter.
i've done this, fire works but if you take it too far, the die is discolored, still fully readable but clearly burnt :D
also, try clamping the entire thing down and just hit the edge with a utlity knife once it is heated, you can usually pop the entire top off in 1 go, it is a bad idea to bend it in the middle as silicon is so fragile, a pair of side cutters off to the side and even diagnally in form the corner tends to snap it right at the egde of the lead frame and the die so you dont break the die as easily
FWIW a friend of mine uses old-fashioned methylene chloride Nitromors paint stripper to dissolve the epoxy packaging. Not tried it myself however, so your mileage may vary 😅
I tried this recipe and I failed perfectly: each time, the chip ended in several pieces, still solidly attached to the epoxy. What did I do wrong ? (the carrier plates came out quite nicely, however)
in the video, those are older chips. modern chips are typically thinner, so you have to be very very careful.
Antoine tells me it is due to the particular epoxy used in the package. On some packages the epoxy sticks to the die and he observes the same failure as you do. He is developing other promising methods to get around this right now.
Surprised to see you here. I am currently working with this technique with ic chips with lots of success
Unfortunately I’ve had this failure a few times with more complex chips and it can be frustrating
Is it possible to decap Desfire Ev1 chip and read data?
What's the technique when you have a die with bond wires, BUT the die and bond wires were encased and covered in ceramic cement? Some of the Intel BGA chip (mid 90s) were like this, where the die was on the underside (but not a flip chip die). They would cover the die and bond wires over with black/brown ceramic. Is chemical the only option to dissolve the die away without ruining its top layers?
i don't have an hot air gun, just an hair dryer and a blow torch, but i'm curious if it would be possible to hook some needles on the dies and see if they still respond, of course big digital ones may be a bit out of hand
Yes, you should be able to make IC dies work by probing them. I don't have good probes in the lab for that, but have been trying to acquire some reasonably priced used ones for quite a while now. Not easy.
If you want to do this, buy a paint removal hot air gun at the hardware store: hair dryer is not hot enough, while torch is too hot and leaves much damages on the die
@@CuriousMarc AH... i simply turned on a 1W white LED without the phosphor (so the bonding wires were off) to let the blue light come out, i used a pair of regular needles and a 7$ USB microscope
I had no idea that the die inside is typically that small compared to the whole chip!
Awesome! I will try this in my lab with BGA soldering air gun. However I missed the temperature you guys used, 400°?
We used 500˚, but can't remember if it's C or F...
I use a heat gun for paint removal at medium setting, Marc did not had one. ~500°C should do.
@@adberco Thank you both! I tested this method (500°C) at lab today on approx 20 chips. About 60% success. But that smell is horrible, dont recommend to to this at home. My lab smells really bead even at night :D Anyway its good but not perfect, in 40% of attempts I got nice and clead die with blob of epoxy in the middle of structure without able to get if off. It seems that various epoxy are used because RAM chips can reach 100% success...
@@marekhelebrant7497 Fume hood or outside is the way, but it should not smell too much. For the 40% where epoxy is stuck there's not much that can be done, but they can be processed using chemical ways (and with a lot less material to remove). Epoxy composition or die surface is likely at play: some manufacturers like National Semiconductor give me 100% failure.
@@adberco Thanks Antoine! Good to know that National Semiconductor is known failure. My tests were with mostly "no name" china chips, maybe these smells worse than others :)
The close-ups look like pictures off of a future google earth......post borg invasion, of course :)
Google Earth of the Google Earth
Try this with Apple T2 chip ;)
I put them on a firebrick and hit them with a propane torch till they glow red hot. The epoxy can be flaked off easily as a chunky ash residue. The die is usually unharmed and I get results similar to this technique. I wish I had a microscope that could actually get some good pics, though :(
I used to do that with a gas torch, but it does damage the dies significantly at a microscopic level
@@adberco I can't refute that, I honestly don't have the tools to tell the difference. It'd be neat to see the difference under a microscope.
Wow, truly geek stuff. I wish I had a microscope some day...
Does Antoine post is die shots somewhere on the internet? I would love to see them.
He posts them mainly on Twitter. He is @Siliconinsid