Detecting FAKE Chips From China 🔍
ฝัง
- เผยแพร่เมื่อ 3 มิ.ย. 2024
- Support Noel's Retro Lab on Patreon: / noelsretrolab
I ordered 4 chips from Ebay coming all the way from China and they're finally here. I'm going to put them through a battery of tests to find out if they're fake or not. How many fakes will we uncover? Join me in the lab today to find out.
Chapters:
00:00 Intro
00:20 Why did I end up buying 4 VDPs?
01:40 Opening and labeling
05:28 Test 1
07:42 Test 2
09:25 Test 3
10:40 Test 4
15:21 Test 5
25:05 Test 6
27:20 Final results
Arduino-based VDP tester project github.com/pajotsetrapper/TMS...
Discussion about the project www.msx.org/forum/msx-talk/gr...
OSSC (Open Source Scan Converter) videogameperfection.com/produ...
You can also support Noel's Retro Lab on TH-cam by joining the channel:
/ @noelsretrolab
Connect with Noel's Retro Lab:
Discord ➤ / discord
Facebook ➤ / noelsretrolab
Twitter ➤ / noelsretrolab
Instagram ➤ / noelsretrolab
#electronics #conterfeit #china - วิทยาศาสตร์และเทคโนโลยี
![[Full Episode] MasterChef Junior Thailand มาสเตอร์เชฟ จูเนียร์ ประเทศไทย Season 3 Episode 1](http://i.ytimg.com/vi/Qs5zdBqLn_8/mqdefault.jpg)
For those of you interested in seeing how those fake chips are made, here's a fantastic video by @plgDavid that shows some footage of the actual operation: th-cam.com/video/k72SFBOZ_lw/w-d-xo.html
makes perfect sense to sell fake chips, who will bother with refunds? not enough and that is why they are doing it.
does not cost them anything to make the chips, so for those that get a refund the loss is next to nill
I guess many may also think they broke the chip when installing it. Or that something else is wrong with their project.
@@tony359 Very true!
I buy things all the time that I do not immediately use. That's where you're vulnerable.
@@JohnGotts Yeah, definitely. I've had that happen before as well ☹️
9:19 "None of these go away with alcohol"... same for my problems
I think the sellers are being scammed themselves, by wholesalers. The fact that two had the same packaging suggests to me that the items had arrived and been shipped without testing, inspection or verification.
One instance that happened to me, I bought wire strippers from an eBay seller, in my own country. They didn't work, the assembly was so poor they just couldn't work at all, so I contacted the seller with an explanation of how the strippers could never have worked and a photo and a diagram of how they never gripped the wire. All I got back was "I just sell them" and a refund.
They must be "Drop Shipped". The seller never even sees the product.
Was thinking same...drop ship bulk buy
Most likely and I’m willing to bet it’s around 60 percent real and the rest are fake so that the fake ones will seem like “dead” chips while in reality they never worked
True, but a reputable seller would cut that distributor completely once the reports of fake chips started pouring in. I believe these guys continue buying from the same distributors even knowing about the problem with fake chips 😞
Resellers and middlemen can often get good parts when they order for the first few times with a supplier but then the quality seems to drop eventually. Bad batches exist also when buying in bulk. Sometimes there are legitimate excuses but other times not. Buyer beware but having a personally known contact defintiely helps with minimising fakes or broken parts.
I once ordered a bunch of chips from China with OR-gates on them. The number on the chips was correct for OR-gate chips, but when I put them on my breadboard, they turned out to be AND-gates lol.
Go figure! It's crazy that they bother to relabel stuff like that.
Sounds about right.
XD
74 ic can be easily identified by tl866
These things cost less than a bag of cheetos, you should get all possible gates just in case. I don't think they did this on purpose, they have mostly organizational problems and the language is a barrier - they're usually small local businesses and these things do happen. I get it, people want things they buy to work as they should, and they get upset. At least you found it funny. +1 I've bought stuff from China before, you can find really good stuff on a fraction of the price, you just need to shop smart - I've done stupid purchases as well, it's a learning process.
I used to inspect electronics for a distributor in the Atlanta-area, and I saw some pretty strange stuff: Acetone was the primary test that we used to test surface markings. When they'd sand off the surfaces, it would end up creating really sharp lines from the surface to the sides (usually they're curved). Ran into a few where they must have mixed the surface material in with some epoxy or something before they layered it back onto the chip; acetone test didn't remove the surface on those, but other visual clues would show a suspect part: Could see changes in the plastic coloring between surface and sides (or surface and bottom), or inconstancies in the printing, etc. My boss, the QA manager, had some pictures on his office door that showed people crouched over a sterno flame, melting solder from PCBs and then smashing them against rocks to pop the parts out. Then they'd take the components and do their dirty sinful business. It's a pretty crazy sub-industry.
This video presented the most thorough chip test and verification regime I have seen on any channel. Head and shoulders above the rest. Excellent content Noel. Your work is much appreciated.
Thank you so much! Glad you think so!
Unfortunately some sellers use photos of real chips and sell fake ones.
The same problem is with transistors and other components.
The only option is to test as soon as you receive it.
Thanks for the interesting video.
Tha happens, a lot. If you want to swim in the swamp of fake parts have a look at RF power transistors, there are thousands of listings using photos of real parts but they send fakes which have been ground down and re-marked. Which is really worrying because they often use toxic Beryllium oxide and I doubt they clean the dust off properly.
I've had that happen frequently on eBay. Polida in particular . No refund offered. They wanted ME to send them pictures of their fake chips even though they had thousands of those same chips still in their stock.
@@hobbyelectronics6630 Polida banned me for forcing eBay to refund, they sent me a snotty note saying 'we don't sell fakes' and wanted me to smash up a transistor containing beryllium oxide then send them pictures of it. Definitely to be avoided.
I bought a fake low noise transistors. Too bad I only found out when it didn't work at low input voltage (45-60mA) in Joule Thief circuit. But better than someone who flies it into an amplifier, or even an IC. :( They made false positive ratings on the beginning of selling. Only way is to report fake sellers.
Often what happens is that it takes ages to arrive and you may not test it the minute it arrives, so in the meantime time goes by and when you want to open a case is too late. Very useful video! Thank you!
I have bought once a MICRDCHIP! That's right! A letter "D" rather than "O".
A real chame...
🤣🤣🤣
That‘s clever, so they do not have to bother with trademark issues! 😆
That is in fact quite common. Like miloloyo which looks mich like mitotoyo. Or TEXTDOL instead of the brand TEXTOOL. The owner of the brand cannot sue for trade mark violation.
I was looking for a screen protector for my watch, and I came by some products writing "TRCHYMETER" where the original said "TACHYMETER". After a bit of googling I saw many results that made the same mistake!
Yep itz a craying chame...
I've run into counterfeit chips a couple of times. About thirty years ago I came across a batch opto isolators that were flakey. The leads had been trimmed to make them even causing problems for me because the part was being socketed. In recent, I've come across a batch of OP-02 chips that have had a bad failure rate. Examination showed trimmed leads and a top coat. The top of the chip looked nice and clean but the texture was different from the sides and the bottom. Alcohol and acetone didn't cause any of the top to rub off with a Q-tip but it did soften to the point it scratched off with a finger nail.
Its a systematic process. They recycle, sand and remark everything. I made a 45 minute video on this. They recycle even smd ceramic caps.
Yeah adeleparts never sent me anything good.
It's shocking that they have such positive feedback!!
Oh very cool. Watching the video right now!
Wow! Fascinating video. Nothing like seeing those pictures. Mind if I pin a link to that video here? I'm sure other people would love to see it too.
BTW, you said you had 50% success rate ordering from UTSource as well. Is that true? I've been told that UTSource is actually reliable, at least the stuff that comes from them directly.
@@NoelsRetroLab By all means! The more people are aware of those chips the better!
Chip number 1 had identically styled printing to the control chip, implying it was an old chip that had been resold. Older chips have older inks that come off more easily than the newer ones that use better/newer paint/ink. Personally, I think it's an old dead chip that was resold, rather than a fake. Chips 2 and 3 all had newer printing.
What a great service you have rendered those of us who buy a lot of IC's and passives off of eBay! Thanks for your very thorough analysis and clarity of presentation! Top drawer!
I think that chip #1 is actually a legitimate original TI chip but is likely recovered from a used circuit board that was sent for recycling. As you noted at the beginning these chips are particularly sensitive and the recovery process could have damaged that chip or it might have been faulty to begin with. I believe this due to how the markings appear on identical to the original chip, how it passes the acetone test and it does not appear to be sanded. I think 2 and 3 are legitimate new replacement chips from TI. The TI logo in the pit is likely a newer anti-counterfitting measure.
A couple of things come to mind:
1. It might be that they are relying on fact that people won't quibble over a few dollars/euros and won't bother to try returning them. This is the same principal used in rebates...many people buy a product because of a rebate offer, but only 20% actually send in the rebate. Perhaps the law of averages is what's at play here?
2. Maybe there's more nefarious reasons? Perhaps the seller is hoping to damage the electronics of whatever these chips are being used for? I don't know how that would be possible, but if they could make power come out of one of the data pins, you could damage the component on the "other end" of that data line.
The eBay seller probably has no idea that the chips are fake, or even what they are for or how they work. Counterfeit electronics are everywhere, with whole factories dedicated to making inferior or completely bogus parts. It really is an epidemic. Small quantity buyers on eBay and Amazon are not the targets. It's likely that whatever distributor the seller bought them from is not even in on the scam.
I've brought this up with eBay and Amazon several times, and they basically love and support fake sales because it means more sales overall, and whenever someone finds out they've been screwed and complains, they just resolve it thru customer service. It's like you said, most customers aren't going to spend much time trying to get back $4 if they can solve the problem for $5 and move on.
Don't ever buy electronics on eBay or Amazon unless you're absolutely desperate or know exactly what you're buying. There is a darker side to this industry. Children and slaves are used to mine materials like tantalum and when you buy cut-rate parts online you also support that horrible industry. It's not worth someone's life to save five or ten cents on a capacitor.
I've had experiences where a single chip is nearly always fake, but multiple chips most of them work. This leads me to assume a % of people just ordering one chip may think that the fault is elsewhere in their system and never make a claim. Whereas someone ordering 10 of them is bound to claim if all 10 do not work, hence why I think they mix in good ones. I also think that many hobbyists might not realise they have a fake chip and never claim for it, assuming something else is wrong in the system they are looking at. Just recently I've bought 3 different WD33C93 SCSI chips from different sellers, each claiming to be genuine, when they are something completely different rebadged. It's super annoying!!!
Yes! So annoying! Finding reputable sellers is becoming more and more important.
Exactly! I bought a fake chip and thought it was another problem, I never claimed
Great video! Really focusing on the fake chips problem, in a particular type of chip and doing methodical tests, like we got used to on this channel, rather than just mentioning and complaining about it during an unrelated repair. Good job! About buying fakes: I've recently had to go through four iterations, to finally buy a genuine 387sx PLCC FPU for my Amiga A2386sx PC bridgeboard. The first one worked for a few hours, before dying. The second and third plain never worked, and also prevented the computer from booting. The fourth one - bought in Europe from a retro seller not on eBay - finally works well ever since. Also, the lettering on the first three was a lot duller, and it was possible to wash it off/make it almost unreadable with some alcohol. On the working one the lettering was much brighter. Interestingly, checking the bottom of the chips they all say "Made in USA Cyrix 1989" so I guess they're all FPUs, except maybe some of them are broken chips from some leftover stock, or was rated for much lower clock speeds than my machine is running at.
Thanks for the kind words! And that's interesting about de FPU! There's defnitely a problem with chips that they change their speed. Those are harder to detect from the ones you supposedly bought. But if the lettering ever goes away, then that's a pretty clear sign that it's some kind of a fake.
Some months ago, 3 or 4, maybe 5, i'm not sure, Adrian Black released a video about "fake" SRAM chips which turned out to be okay, just not rated for the speed that they were supposed to run at according to the markings, so yeah, that's definitely a "thing" too.
@Hans Otto Kroeger Another good reason to give Amazon a wide berth!
Great video and you make an important point, to check your purchases as soon as you receive them
They may be paying a dime each for those fake chips. Sell them for $3, and you can still afford to pay refunds if someone complains, and still make money.
That's only true if the buy pays for shipping. - That's where a lot of money is eaten. It is also why most sellers will not do free shipping (as it isn't actually free, but included in the price).
The only way it may work is if the seller(s) do not want the "bad" chip(s) returned, but will still refund.
@@DanSchallerforPOTUS Doesn't China subsidise all deliveries?
Great video! Another way of comparing/measuring unknown chips is to identify the clamping protection diodes that most designs have on the I/O pins. That makes it relatively easy to separate supply pins from I/O pins and compare that to the expected pinout.
Interesting, I hadn't thought of that! How would you do it? Just making sure there's no current flow in from those pins?
@@NoelsRetroLab Yes, sort of. I/O pins usually have two protection diodes. One towards GND and one towards Vcc. This to protect from transients. You can use your multimeter in diode mode to search for these diodes using the expected pinout as a reference. For example start with one meter lead on the expected GND and go through the I/O pins looking for the diodes. By swapping the meter leads you can determine the orientation of the diodes. Then do the same for Vcc. If things don't add up then after a few rounds with the meter you can usually at least figure what is the supply pins on an unknown or re-marked chip.
@@danieltufvesson1349 Good tip! I'll keep it in mind, thanks!
What I noticed during the alcohol was the suspect chips’ lettering was about invisible when they were wet. In the real ones’ you could see the writing bright and clear.
That's not reliable as I have tested chips that do the same, but are in fact legit. Laser engraving becomes hard to see but some manufacturers use it.
That happens with some bikini's too. .. I like it.
I don't know if it interests any one , the dots come from the ejector pin on the mold, they push the part clear of the mold. Not an expert just work it manufacturing
yeah that was my first thought on that too
I like that they put the TI logo on the ejection pins. A nice little touch that is likely not worth the trouble of the counterfeit chip seller's time.
Texas Instuments also sometimes has elongated directional marking instead of just a half circle.
Another excellent video from Noel. I'm also curious about this strange way of obtaining profit with damaged (or non-working) goods. I've seen that in some cases and similar transaction processes, it is all about money laundering through transactions based on near-to-zero valued goods. We don't have enough information (as simple users), but totally agree with you: this scheme is quite strange and pointless, unless something is going on in the backyard. Love your videos mate: 8bit+electronics+learning+methods for real life+investigation !!!!.
Thank you so much! 👍
money laundering is a huge problem but if the customer refund it before paypal release the funds to the seller, the seller dirty money is never circulated, right?
@@michaelw9852 Yeah, you would think. Otherwise that's a huge loophole there!
An excellent and really interesting video...thanks, Noel!
I repair and restore old pinball machines and this is a constant annoyance. I just assumed they might just hope that the receiver either doesn't bother to ask for a refund for a $5 part or might think they broke it themselves and wanted to avoid any hassle. Most colorful reactions I've gotten from a fake TIP102's, that literally blow up if they get any load. They're a lot of fun to watch when a pinball machine fires up a solenoid and instead emits smoke signals.
Pinball machines...with a hobby including any sort of old electrical stuff...back in 1982 I bought a well used Pinball machine from a guy selling the new ones for arcades...
BALLY 'Space Mission'...exactly zero electronics in there...lots of wires, switches, relays...solenoids and a few motors...switches, a few but thousands of contacts on the 'score' counters on the back-board...
Excellent! Great testing though.
Thank you! I remember you also had problems with fake chips in one of your videos 😖 That's becoming quite a problem.
Woohoo! ADB in the house!
Hi，dear ,i am a chinese，i haven't commented in here before,but i must reply to you in here ,
there are some facts you don't tell in here:
1.TMS9929ANL is a very old chip，this factory stopped to produce for years，this new and original chip is very hard to get ! you may know this condition，and the
web you buy from have not the right picture,you have uesd it and you know the right picture,why you bought the wrong picture web link ?
2.there are A basket of apples,but one apple in this basket gone bad，We cannot say all the apple in the basket are gone bad，what we should do is to pick up the bad apple ,and throw away the bad apple, We cannot throw out all the apple of basket ,you know why ？
the bad apple only is the bad apple ，that's the facts!
you cannot image all the apple are the bad apple.
if you are good man，you should tell us what store it is,you cannot say all the stores of China are bad .
this stores behavior damaged the credit of the business men in China，and you video do the same thing as the store to damaged the credit of us ！
I also sell all kinds of new and original IC parts ,We have a good reputation.i hate those bad behavior ，whatever behavior you want ，you must point out who is the bad person,not say all the peolpe are bad !
Great ideas here! I have some Bucket Brigade delay chips which are very hard to find without ending up with fakes. Looking forward to trying your tests.
Thank you for creating and posting this video. I found it quite useful and entertaining. Subscribed.
Awesome, thank you!
One thing I noticed is that you use a static strap and I am glad to see it. Not many videos show that it is important. Showing an alternate way to test these chips is fantastic. I also had some bad chips that were supposedly out of USA but came from china. I love Paypal, they protect us when the sellers are not responsible companies. Back in the 80s I had gotten hold of manufacture's discards which I was surprised that they weren't destroyed or chopped up. So they are out there. Thank you for your wonderful and smart videos! We get to learn things the right way!
I'm glad you noticed that! You wouldn't believe how many people in the comments "yelled" at me for holding the chips without ESD measures 🤣 Really glad you enjoyed the video!
@@NoelsRetroLab I am glad you told me, I thought I was the only one upset about people not using it!
I remember an electronics/computing magazine article from the late 70s/early 80s, which reported that someone had been dumpster diving (or maybe had inside help) at a chip manufacturer and was selling on test-failed chips (EPROMs, I think) as genuine. They had been discarded before labeling, and the scammers had printed new labeling on them, but used a slightly different typeface. (IIRC, one '7' had a straight upright, and one had a curve.)
Some things never change.
It's interesting how all the various vendors had 99+ % feedback ratings... Evidently there's some way of gaming the system? Good investigative work, and presented very clearly.
Yeah, kind of scary, isn't it? I suspect it's not gaming so much as they sell a mix of good and bad components. And if you assume that only a certain percentage are going to leave bad feedback... there you go.
Thanks for this, most educational. I have been lucky so far and have not got fake chips off ebay (but I've only bought a few). You've given me some techniques to test before I blow up my project. Happy to subscribe also
Great to hear! Always good to at least do a quick check when you get some ICs online.
Very interesting and thorough test! Was super interesting, thank you!
open up the chip CuriousMarc's method, maybe someone recognize the die, or maybe has some markings.
I would like to know it thiers anything snuck inside these chips.
That would be awesome!
There's probably no actual die in there - just an empty chip package. Sometimes they put a little slug of metal in there too, sometimes it's just the metal that normally is there part of the lead frame that the actual chip would sit on if there was a chip.
@@gorak9000 just a lead frame lol. how do they sleep at night.
We do recurring training at work to guard against counterfeit parts like this.
Great video. As someone who has fallen victim to this practice more times than I care to admit to, this gave some really great tips on how to check and test.
Just Subbed, Noel!! Very Well Done and informative video! Thank You!
Awesome, thank you!
Nice work identifying the counterfeit ones! This is really becoming a problem especially with the 68060 chips. It might be a cool follow-up to try to resolve what those counterfeit chips are in reality.
Yeah, it's definitely becoming a problem in general. For any non-trivial chip, I'm going to start ordering only from trusted sources.
Not sure how I would go about finding out what those chips are since the label was erased. Any ideas?
Unfortunately, no ideas that wouldn’t turn out to be tedious. It would be safe to assume that the counterfeit chips are not faulty TMS9929 because those wouldn’t need re-labelling. I am not sure whether they could be TMS9918. One would think the pinout would be the same between the PAL and the NTSC versions. I could be wrong, though.
@@terosaarela4555 They're definitely not TMS9918 because I accidentally got one of those first and the clock oscillator works and they output an NTSC signal, which can kind of be displayed on a PAL device. So they're probably something else completely different. I should rub with acetone some more. I saw in the video in the pinned comment that if you keep rubbing sometimes you see the label underneath.
Noel, In your video It looks like all working chips have a somewhat larger hole between the middle connecting pins on both sides. So that may be an easy way to recognize real VDP chips from a photo..
Do they? The righmost two (numbers 3 and 4) have... something there. But the working one from the TI seems totally even.
@@NoelsRetroLab Have a look @ 5:32 it has the same cutout between the middle legs.. (pin 10-11 & 30-31). Also take a look at the photo's on this wiki page: en.wikipedia.org/wiki/Texas_Instruments_TMS9918 all chips have the same notches between the middle legs..
@@L0wcash You are right! On the genuine chips there is a tiny rectangle of metal cut out from the top part of the middle pins....take a look, Noel! Well spotted, Retroperator!
I used to work for a company that did some counterfeit chip testing. You would be amazed at what we saw. Chips with die inside but no wirebonds. Chips with no die at all. Chips with a date code years after the last time buy notice. Chips with one company's logo externally, but a different company logo etched in the die pattern. As long as the package is the same, and demand exists for the obsolete chip, counterfeits will be made.
Better fakes use laser markings because the solvent (acetone) test was so effective at detecting poor ink marking.
Using x-ray and comparing to a known good sample of similar vintage and from the same fab is the best method of detection, second only to electrical testing.
The vendors are very fly by night and operate as a number of companies, changing identity frequently enough to evade detection. Even with an escrow agreement, they will just try to stall long enough to get the funds released before the counterfeits are identified. Or, they send a handful of good chips for advance testing, get the funds, then ship a load of fakes.
Another scam is to salt a lot of mostly good chips with some quantity of fakes, especially with higher cost chips. The hope is that it takes some time to identify the fraud, either to clear escrow or to give them time to change the company identity.
Buying from "different vendors" does nothing to protect the buyer because if they are selling the same chip they can easily be the same fraud company. Best to avoid buying obsolete chips from China unless you want to buy a handful and hope at least one works.
Wow! That's even worse than I thought. I suspect that one that wasn't giving any kind of resistance between pins might have had no wirebonds at all. Any other checks you can recommend that we can do at home for counterfeit chips?
Nice video. I think you should do a followup on this video and create your own simple PCB with a few LEDs that just check for basic resistance when a battery is attached. I know it would not probably show if it is fully functional, but it should show if it fake or not to those who do not have much skill to test like you do.
Hi Noel,
Never watched your channel before but I must say I'm impressed. I like how you establish a baseline and follow up to the next step.
Was glad to see the amount of logical choices in this video that make sense, especially regarding not just testing the possibly fake chips on a working board. So many retro channels just seem to do whatever and may cause more damage just for a video :)
Safe to say I'm a fan, subscribed!
Thank you so much! Really appreciate the comments and feedback. 😃👍
Three things to say:
Firstly I came to TH-cam, I am an electronics enthusiast, but I like repairs, simple objects are my curiosity and motivation to enjoy repairing.
Second, a very good video, on how to detect fraud, would obviously go on trial and error, and you explained very well how to detect a chip fraud, without damaging the equipment in question since these older ones have chips and components that do not manufacture anymore and it’s dangerous to fix it on the market.
Third I liked the shirt, I had to look for the periodic table to know what chemical element it was.
Congratulations.
Três coisas a dizer:
Primeiramente vim por indicação do youtube, sou um entusiasta na área de eletrônica, mas gosto de reparos, objetos simples é minha curiosidade e motivação em gostar de consertar.
Segundo, um vídeo muito bom, sobre como detectar fraudes, obviamente iria na tentativa e erro, e você explicou muito bem como detectar uma fraude de chips, sem danificar o equipamento em questão visto que esses mais antigos possuem chips e componentes que não fabricam mais e fica perigoso consertar com que tem no mercado.
Terceiro gostei da camisa, tive que procurar a tabela períodica para saber que elemento qumímico era.
Parabéns.
Great thing about that Arduino Uno is that if you fry the IC you can just pop it out and put another in. Of course you'd have to plug in an ICSP adapter to flash the new chip, but I did so to my own Uno when I blew out a pin.
I have seen another video showing a bit similar comparison. However, there was one additional observation. The bottom side markings typically contain additional information like the manufacturing plant and batch numbering or similar. People doing relabeling ("remanufacturing") may not bother sanding and redoing that info. I am not sure what exact info you could dig out from the bottom side markings in this case, but in any case,, there could be some useful info before or after alcohol (or acetone) wiping.
Another thing relates to the sourcing of the chips. We once got a batch of certain chips from an authorized distributor. These chips failed in short order in our product burn-in test. When we started looking closer, the part marking was on a wrong side. Apparently the printer of the re-labeling operation could not print on the correct side for one reason or other. The part was a power MOS FET in TO-227 package. Inside there was a pair of small chips, while the original ones used just oner large die. The distributor claimed they got these parts from the official source. The conclusion might be that there was some foul play somewhere in the supply chain -- somebody with access to the parts did a nice swap, replacing the good official parts with cheap fake parts and then got good parts to sell at a nice profit elsewhere.
This is really useful -- thank you! Not only some good techniques for detecting fakes, but I'm building a video card around a TMS9918A right now and that Arduino tester circuit will be something I can adapt to ensure that I'm using working parts!
Awesome! Glad that helped!
As an aside, while the TMS9928A/9929A outputs are labelled as Y,R-Y,B-Y they are in fact YUV signals. I'm working with a small team investigating this and the palette associated with these VDPs. We've taken vectorscope readings and these have confirmed our suspicion that the output is YUV. When our work is complete we'll document and publish our findings.
That'd be great. Let me know when you do! It certainly confirms what I'm seeing that I can't feed the signals from the VDP into a TV accepting component input. I just tried with directly with a TI99/4A (same VDP) and got the same results. I want to go a bit more in depth about it in a followup video, so it would be great if I can include your findings too.
@@NoelsRetroLab I'll let you know.
@Ikrananka Yes, most interesting and probably true ! Today I connected the Y,R-Y,B-Y output pins from my TMS9129NL to a TV with component input (YPbPr) and it just showed nothing ! But when connected to the YPbPr inputs of a GBS-8200 converter board, it detects 'something' and it is converted to a VGA image.
@@fa497007 Interesting - thanks for the info. Apart from our vectorscope readings showing 100% certain that the signal is YUV, another strong piece of evidence is from the obscure Tatung Einstein computer. This short lived computer was aimed at business users and had a separate, optional, YUV video output connection. The YUV connection was literally taken directly from the TMS9929A, only passing through voltage follower op amps to buffer the VDP from potential damage.
Wow ... this is exciting to watch. I was routing for each and every chip during the review XD
I work with plastics and we use a silver colored Sharpie to mark our materials. Much easier to see than black marker on black plastic.
Yup. I've switched to those since then and they're way better!
Hey!! I was wondering about the same problem with all chips I buy on AliExpress and Ebay. I just bought 74HC logic chips, 555, Z80, SRAMs memories, CPLDs and sound chips. I have to test all chips and I think a 50% of them are counterfeit or failing. Thanks, thanks for doing this video. Interesting, entertaining and outstanding complete study. Please, make more videos like this!!
Thank you! I haven't found this a problem with simple chips like 74XX ones, although someone said they got LS chips instead of HC chips. So that's already happening apparently 😞
Buy logic chips from DigiKey.com, MUCH MUCH cheaper than eBay and AliExpress, and is 100% working. Guaranteed! Same for SRAM PDIP format.
@@unebonnevie I've used DigiKey a few times and the initial price isn't cheaper. It might be if you take into account the time wasted reordering chips to Ali/Ebay though. Also, shipping from EU is a big deal as I believe you need to order over 50€ to get free shipping (or pay a ton). So that's kind of hard when you just want a couple of chips to finish a project. They also don't seem to stock up old chips like this one, so that's a bummer. I'd love a reliable source of old chips.
Well, it's so expensive to order from Digikey here in Latin America because of the shipping costs. Here in Chile we have RS Chile for some chips like those. But AliExpress and Ebay are way more more convenient for their superior catalog of circuits.
@@xyz2112zyx Yeah, totally. Especially for small orders.
As that chip with slab heatsink is toast, you could use a heat gun or put it in the oven to get the heatsink off. You could reuse it in one of the projects for the sake of originality or just keep it "just in case."
Cool. I wasn't sure if it was some weird glue/epoxy that wouldn't go away with heat. I may give that a try, thanks.
Great informative video.
Thanks Noel this wonderful piece of information !!!
My pleasure!
Very interesting. I thought the last 2 chips were promising when I saw the TI logo in the round dip. There are quite a few fake AY sound chips around at the moment as well. I think they hope to make a bit of money off people who can't be bothered to send them back, or leave it too late to find out they are fake.
Yeah, that's true. It's happened to me a couple of times actually.
They will also keep communicating with you "trying to solve the problem" until it is too late to apply for a refund under buyer protection.
@@hobbyelectronics6630 True
So: are we talking fake as in something like the actual chip, but with maybe a different speed or temperature rating, maybe untested, but at least there's *some* hope of the chip actually doing the job, or are they fake as in random 40-pin packages that have no relationship to the real one, and have no hope whatsoever of working?
This is what I want to know... kind of an important question to be resolved.
Not really, it could be a chip that was recycled from a larger board. But the chip might have been damaged. So, the chip might be sold at a discount. Than a distributor might try to rebrand the chip as new. IEEE had an article on this subject. A few years ago, a Florida distributor made the news when they were caught selling rebrand chips to us government and military contractors.
This encompases everything. So maybe counterfeit is a better name. But I suspect those two failing ones are just some random DIP 40 ICs that were relabeled.
Even sometimes it is the original chip, buuuut used and gotten the worst treatment (like having a bath in molten tin) imaginable. So you often get those selled as new.
@@NoelsRetroLab I'm with the other guys! Look into decapping it!
Great test, very correct approach.
Thanks for the exciting thriller, Mr. Noel.
Subscribed :)
Thanks! Welcome aboard!
I like your Subterfuge shirt.
Gosh, I haven't played that game in such a long time. I think I'll download it again and start again.
Oh no! What have I done! 😜
@@NoelsRetroLab Wait a minute... 🤔
Are you _that_ Noel? Didn't even occur to me until just now, almost a week later. 🤦
@@greenbanana311 Yes, that one! Small world! 😃
Interresting video :)
Another thing to look for when doing the visual inspection is the date codes on the chips.
The original working one says it were manufactured in the 10th week of 1983 which fits both with the chip type and the device it is used in.
All the other chips have either nonsensical date codes or datecodes that seems wrong. The two new working chips, no 2 and 3 do not follow Texas Instruments convention date code convention of YYWW (year, week) and if read the other way, it says they were made in the 32nd week of 1970, which makes little sense since this TMS chip were first introduced in 1979.
Despite them working, they might still be "fake" in the sense that they are reused from discarded electronics and then rebranded to seem new.
Good point. I'm not an expert at codes, so I don't know if they simply changed their codes at some point or they're fake. Although you'd think that if someone goes to the trouble of rebadging a chip, they're most likely copying an existing one, but who knows!
I actually doubt the last two are "fake" or "clones" or anything like that given some of the other comments (this is not a VDP that has been replicated and manufactured recently from what I was told).
@@NoelsRetroLab The datecode on those chips is the one after "DDU"
The datecode has not been changed for the DIP chips as far as I'm aware, nor according to TI's "Device Marking Conventions" document.
Table 6 on page 5 here: pdf4pro.com/cdn/device-marking-conventions-rev-c-texas-f8a2.pdf
The code is of the 4 digit format with year first, week second (YYWW) so the datecode on chip 2 & 3 would read 32nd week of 1970 which does not make sense since the TMS chip were first introduced in 1979. Which is why I suspect those two chips have been relabeled.
They are most likely genuine chips, but old, used ones, which have been relabeled to make them look brand new instead of used/refurbished.
This is not uncommon for the counterfeiters to do, despite it seems like a lot of work. Nor is it uncommon they just put random numbers on them to make it seem legit.
@@Gin-toki Very interesting! I wonder how they can relabel those and make them look so good and even. And all that work for very little payoff, because it's not even like they show that picture in the Ebay auction. So I would have been fine getting a beat-up, scratched-up chip as long as it worked the same way. Very puzzling that whole thing. Thanks for that link. I'm definitely going to read more about it.
I am new to the channel. Intro made me smile (in good way), content made me subscribe.
I am in Australia and have come across people buying vintage synthesizers, replacing the original chips with fake ones and selling them back on as original condition. These original chips willl be worth heaps in the future.
Kind of an aside, but still worth noting: I've seen a lot of Chinese power supplies lately that have fake UL listings.
Yikes! One more thing to watch out for!
My thoughts.
Chip 0 is clearly fake.
Chip 1 is more interesting. The markings look genuine but the readings are no good. My guess is this chip may be genuine but is faulty.
Chips 2 & 3 genuine.
I thought for sure Chip 1 was real. Like you stated maybe just a lemon.
Notice every other chip has printed letters, but chip 1 is laser etched. Clearly the faker knows what a legit chip looks like and wanted to copy that.
I love the randomness of ordering electronic components and such on eBay. I have received things that should be in antistatic bags on regular ziplock, and things that did not need ESD protection at all in ESD bags. Including ESD wrist straps.
And a soldering iron tip, but at least that one was a resealable bag. The wrist straps were not.
TI/994A was my first computer! Concerning your IC's I had 2 fake lm338's from a local store, 1.2-32v/5A voltage regulators. I was told that they were packaging lm317/1.25-37v/1.5A* into them in china. Not the same thing at all. It's annoying, but I'm glad you at least got 2 out of 4. Good video.
-They buy old chips in bulk.
-They sort the chips by what socket they fit.
-So you place your order
-They look up what socket the chip fits.
-They grab a chip that matches the socket.
-They sand off the number and print the number you asked for.
-Wrap it in a zip lock bag and ship.
So they are selling thousands of chips under a hundred different numbers.
Since most people would assume they misdiagnosed the problem or there are additional problems they dont bother asking for a refund.
I've actually had a chip arrive for an old stereo and it was obvious that they had cut down a longer chip so it would have the right number of pins, but the printing looked right.
The reason for doing this with cheap items is that a high percentage of people will not complain or ask for their money back ( equals profit ), if the item was 50 euros or dollars then everyone would want their refund and there would be no purpose to the whole process
That’s what I was thinking too. I’ve done this multiple times actually where I will get a wrong item or something and if it’s cheap enough just take the loss rather than spend my time trying to return it. It sucks but it makes sense that’s why they would do this. For $3 if I had to actually return the item for a refund I would probably just eat the cost instead :(. Guess I’m part of the problem
@@dandoyle I always ask for a refund, multiple times, and have never been asked to return the product.
There is a free option with eBay ,you need to sign up for it, where eBay will refund any returning costs should you need to pay,say for larger more expensive items.
Right. Same here. Instant refund without sending anything back.
eBay has been getting worse recently: non-delivery, late-delivery, wrong item, broken item... I've not had a problem raising a refund request via eBay - that seems to push the vendor into action. It's just a few clicks, and most often you don't have to pay an extortionate amount to return it as they just tell you to keep it. But it takes so much time, and you have to reorder from somewhere else. Some of the delivery times reported from China are now into next year.
Enjoyed the video! Just got my Arduino project kit in the mail today. I am definitely going to try and make that test circuit.
Awesome! Yeah, it was very straightforward. Let me know if you run into any problems. I'm still wondering if the video output I got was the intended one, and it was just my LCD TV filtering things a bit weird, or if I miswired something. I expected more vertical bar patterns. Let me know what you get.
@@NoelsRetroLab Hi Noel, try playing with the timings in the loop section of the sketch. Changing the delay, can have impact on your display output. The 'setRegister(7,...' function is responsible for showing the backdrop colors, one by one. You can change them more quickly if you lower the delay value. Also to see the bars, you will have to put the 'writeByteToVRAM' section into the loop, and maybe try with a slightly higher delay value. Also note that using these jumper wires is not the ideal situation for video signals, sometimes wiggling them can improve image ;-)
@@fa497007 Thanks! I took down the circuit, but I'll keep that in mind whenever I make it again to test some more VDPs. Maybe the next step in the project should be to create a simple Gerber file for a pcb. Especially if we use a BluePill, then all it would take would be to put in the microcontroller, the VDP and power everything up, without a single cable. Bonus points for having RCA jacks already on the board. Actually that sounds like a fun project! 😀
@@NoelsRetroLab Indeed a PCB would be fun to build. Especially when you need to test more frequently. And that's maybe a hint for the Chinese, use it before selling those IC's !
@@fa497007 That's assuming they don't know what they're really selling or that they care! ☹️
Excellent video.
Same thing happened to me also with fakes LT1084, LM 338, uA7815 and other IC's I bought off eBay China and AliExpress.
Another video that goes way beyond most others, Thank you Noel.
My pleasure! 👍
5:20 - This is purely a guess... but if I had to put a finger on the process it'd be something like government subsidization. They sell a trinket and a packet leaves the country, their government gives them maybe a euro of credit or something, maybe some other kind of power within their infrastructure. Government backing plays a HUGE part in the Chinese economy. You've got to remember, the country is communist, they're selling to a free market but are themselves not.
My dear colleague! As they say in Hungarian, I am an "old fox" in electronics. I came across counterfeit transistors 35 years ago !!! (and more and more since then, MOSFETs, etc.) in 1985 (the Iron Curtain still stood between Eastern and Western Europe.) Hungary belonged to the Eastern Bloc. I asked my friend (who has „western passport”) buy me 100 pieces of 2N3773 transistors In Western Germany. He bought, (it was not cheap) and all 100 pieces had a breakdown voltage of less than 100 volts (according to the data sheet were 140) it was “RCA made.” He was soon able to return it to the dealer who returned the price without dispute. They started counterfeiting back then, and I've been testing everything before soldering. I wish you much more success in your work and good health!
This one is a great one!!! thank you again Noel, a couple of cuestions: 1. Why did you use the VCC and clock signals to check? 2. On the breadboard you reproduce the scheme showed by yourself, well.. How do you protect the circuit (Arduino) ni case it has a short circuit? Again thank you very much
Thanks! 1) I chose those just on instinct because they're pairs of somewhat related pins. I could have probably chosen any bus pin vs. ground, and really, pretty much any pair as long as it has some kind of connectivity internally. 2) Good point (one I know you're well aware of 😃). That circuit doesn't do any kind of protection! It assumes you have a real VDP and it would probably be rare to have a destructive short. With fake chips, you can have anything though! So you're absolutely right that ideally it should have some kind of protection, at least on the output pins.
14:334 i mark my bad chips / auto parts with "NFG"
.
for "no fuc@ing good"
"When even the real one, comes from China..."
if is come from China is not neccessary false. But if is a false , it probably come from China..
I had a similar issue with some RCA 1802 microprocessors I bought from China. I built a tester using an Aruduino Mega that let you run a machine code program to test the uC. It turned out that 8 out of 10 of them were good. The other two sort of worked but were not 100%. It pays to functionally test all such parts as soon as you get them.
Yeah, for sure. It's scary that they "sort of worked". At least in this case they might have been plastic chunks, but that means you got faulty ones, or slightly different models. Scary stuff.
I always assumed the ripoffs were just any defective chip in the same package where they just remove the silk screen and have blank bins. Then as orders come in they are wiped with whatever numbers the customers want from existing stencils. So yea, your obscure chip silk screen might only get thirty wipes a month, but they probably have thousands of silkscreens that match blanks on hand so they fill those orders.
Other issue I had was the chips that didn’t quite make spec at factory were just rebranded and sold at the same price.
I was expecting some kind of impostor doritos or something
send them to Ken Shirriff for de-caping !
Usually when I need an I.C. for a repair I need it yesterday, so I order from U.S. sources that can get it tomorrow or the following day. But working on 60's thru early 80's legacy receivers and amplifiers I need matched pairs of very popular NPN/PNP audio power transistors. I found a deal on matched pairs in to-3 cases that was too good to pass up. I ordered 25 sets, and the first time I swapped a pair the PNP was much hotter than the NPN. You can imagine what that does to your signal in a push-pull circuit. In addition, these transistors were rated for 60 watts properly heat sinked, but they both died quickly at less than 20 watts. So I opened the top of the can and found my original transistors were about 1 cm square, thermally bonded to the base of the case, with pretty beefy gold leads from the wafer to the 2 pins. When I cut open the new ones, the wafer was about 2-3 mm square with much smaller wires, and the wafer was floating, held in place by the wires. It was obvious that I was sold fakes, and since this was before PayPal, I don't think I ever got my money back. Since then, I only buy from reputable, established distributors, unless it's a discontinued, legacy component with no equivalent replacement,
Great video. Liked how you put the which seller's were the decent one's.
Keep up the good work. 😁
Thanks! 😀
These won't be the only items that these Chinese sellers deal in. You'll find that they sell all kinds of stuff.
Also, the 2 chips with identical packaging from 2 different sellers were probably from the same people. They set up multiple store fronts and peddle the same stuff. That's meant to be against Ebay's rules but Ebay doesn't actually care as long as the sales commissions keep coming in.
More like same supplier. Many sellers have multiple accounts, that's true. However, a much simpler explanation is that they got the chips from the same supplier.
@@chedwick Yes, except the packaging (envelope, carrier, etc.) is identical, as well.
@@DavidSprings lots of places don't have the goods themselves but get them from "drop shipping" so they just say send this item to this person... and it is out of their hands and they get a "fee" for doing the sale
The space between the 2 most central pins on each side have a distinct look on the chips that work. Interesting.
Yes, some people pointed that out. Same with the original chip, although it's less noticeable.
T O P C O M M E N T
Thank you for your helpful information. This is a real problem.
I purchased some new eprom chips that were also relabelled. After removing the paint you could see the original chip was the same but from different batches so they made the entire lot look like it was from the same batch. To me it would not be an issue if they are not the same batch so I am not sure why go through so much effort for this.
That’s some good info! Especially since you are forced to order from China as there’s often not available.
I try to avoid ordering from China as much as possible. I find their regime deplorable and they also have grown by merely learning from our IP that we send there to mass produce. And now they are an economic threat to the western world.
not all buyers start a dispute as i think so their profit is from those peoples.
That makes no sense..
Wouldn't the wasted cost on creating and sending the fakes that do get refunded outweigh those profits really fast?
@@Zebra_M no for sure it is all done by hand and almost cost free in China
@@DougieBarclay I guess what he means is if somebody buys a chip speculatively, as most would, hoping to repair a machine but not sure if they could, they might assume the chip works but something else is wrong.
@@obadiahnormal8070 Why would anybody go and buy a chip without testing the whole machine? Nobody in this hobby can be so stupid. Also, none of these comments make sense from OP.
Very informative video as always. Like you mentioned, these fake chips (depending what they actually are) could cause serious damage to our retro machines, so it's vital that we actually know what we're getting. I have to admit I've been careless in the past and trusted sellers to be honest, and so far I've been lucky. I'm definitely going to make more of an effort to try to verify any chips that I receive in future and, as you say, do it when they arrive rather than a few months down the line.
Thanks! Yes, it's always good to check, even if it's just visually and some acetone. That will catch most of the really bad problems.
Great stuff Noel thanks for sharing
37 bad ebay sellers has thump down this video.
I suspect they are getting their money from selling your buyer's info to a database. Either in the form of seeing what is being bought, as in buyers' activities database or worst your mailing info, I have been suspecting that from most China-based eBay vendors for years. EDIT: chip 0, I would be curious to find out if it's not just some other chip they had in big volume and they just sand down and re-label it as other chips, maybe look up what was commonly used with the same package and try finding out what chip it matches lol Could be time-consuming endeavour mind you.
Right, chip 0 is definitely something else. Too many things in DIP 40 form to know reliably though. And besides... I hate to admit I threw it away 😞 I didn't want it mixed in with my box of faulty (but real) chips, and I didn't think of doing further investigative work. Oh well. Next fake chip I'll look at it more closely.
Thanks for the kick ass video. You rule. Question: I did this test with some 8-pin opamps and my multi-meter functions the same as yours, but I'm not sure how to correlate fake vs real based on genuine specifications, meaning, if I was looking at a Mouser white page with specifications, what exactly would I look for to correlate fake from genuine (in regards to ohms or megaohms) ? Thanks again man.
No manufacturer provides that type of pin to pin measurement information that I know of.
The only way to do this in the field would be like Noel is doing it, with respect to a known good chip.
the round dots are from the injection molding process of the plastic. the dots are made when pins push the piece out of the mold.
Most of these comments are:
"I came here for a good argument!"
"No you didn't, you came here for an argument."
For those who missed it ~ Courtesy of Monty Python
It's very easy to spot a Chinese fake with no electrical testing whatsoever, at least if you have in your hands an original IC or access to a quality picture of the writing on top of it. There is one thing the Chinese will never ever _ever_ get right: Properly rendering English fonts on the chip's surface! It's a known fact. A Chinese fake with an authentic rendition of all of the characters on the chip goes against the laws of physics, breaks the second law of thermodynamics, allows superluminal travel, and causes all its pins to give birth to unicorns, at the same time. The font on a fake will __allways__ without exception look different than on an original. Just compare letter by letter between what you have as a possible fake and a known authentic chip or picture of the top of one. I'll admit, there were times when the differences were subtle, but 99% of the time, you might as well be comparing how nickel looks compared to a dollar bill, that how jarring the differences are.
If you only have a single chip, look for asymmetry in letters that should be symmetrical (e.g., U, I, S [most of the time], N [kitty-cat symmetry], etc...). Look for bad left/right symmetry as well as bad top/bottom symmetry in letter and number forms. One other thing the Chinese mess up is making upper case letters the same height, this can be another dead giveaway if any of the letters in a part number or identifier are of a slightly different height.
For our example, you can also search for TI 99/4a motherboard image that seems authentic (i.e., from a taken apart TI 99/4a). This will give you clues about the font TI used and the layout of the symbol, lettering and numbering on related chips. For example: www.nightfallcrew.com/22/06/2013/texas-instruments-ti-994a-diagnosing-and-fixing-motherboard-faults/
And, boom, samples: #0, #2, #3, from the video at 5:48 immediately disqualified as authentic (reason: five digit code does not precede place of manufacture {PHILLIPINES})
Sample #1 disqualified (reason: no space between DDU and 1901, no space between A and NL in the part number, TI Logo {just wrong!}, now let's talk letter, the "E" in PHILLIPINES, look at its central horizontal segment {it's as wide as the rest - should be shorter}, look at the two P 's, notice how the curve at the right of the letter has horizontal segments connecting it to the trunk, unlike the P on the authentic chip which has a nice curve, look at the U in DDU {shorter than the two D's), look at the uppercase S in PHILLIPINES {the two halves are the same size, unlike in the original where the upper part is slightly smaller than the lower, as it should be}, how about all the digit 1's {compare the top of them to the authentic one, you can see a difference}, etc...
Conclusion at 5:48, all of the chips on the right are fake! And, at 7:05, yes, nice counterfeit work, but the dot on the "i" in the TI logo should not be hollow!! Compare the logo shape and characteristics with: upload.wikimedia.org/wikipedia/commons/b/ba/TexasInstruments-Logo.svg So, #2 and #3 work, but they are still fakes, just working fakes, or, in the best case, rebadged old chips, perhaps, that were pulled out and restamped.
Not quite that simple. If it's a chip which has been in production for a long time, you will be seeing difference in packaging and text rendering, even if it is an authentic part. There can also be multiple factories producing the same chip and production can be restarted due to unforeseen demand. This varies from chip designer to another. This is especially the case where the chip production is first inhouse and later outsourced, e.g. AMD and Intel.
Right! As a matter of fact, when I posted this video on Facebook/Twitter, very few people were able to spot the fakes correctly just from the picture. I think you'd have to be an expert of that particular chip type to be able to spot them accurately.
@@JamieDonovanD Although you are right in principle, with regard to an exact comparison with a known good chip (i.e., there may be multiple authentic variants for some chips), the rest of what I said still applies (i.e., things you can look out for without access to an authentic chip or chip image).
@@NoelsRetroLab This is because most people don't have experience doing this. If there was more education of the form in my reply (both for those with access to an authentic chip and those without [or for the case of multiple runs by potentially different factories]) there would be fewer users suckered in by Chinese fake silicon, at least with regard to ICs with writing on them.
very interesting and informative video.. thanks Noel :)
My pleasure!
Nice! Good to know for the future!
Checking this chips is like a COVID-19 testing. Haha.
Crewmate
There is 1 impostor among us
Can you tell me how to check 74HC595 8 bit shift register with digital multi meter to find out whether it is in working condition or damaged?
Fascinating video Noel, very thorough investigation. First time on your channel, I'm very impressed with your knowledge and low key approach. Subscribed.
Thanks so much! Glad you enjoyed it and welcome to the Lab! 😃