Years ago I bough some 1F "super capacitors" from China. I put them on a shelf without measuring them. When I did get round to using them they were under 1uF. Lesson learned.
You can buy the entire 14/16 pin CD4000 line from TI in DIP. Notable exception is 4000 “mix and match” gate, 4008 adder, and 14530 majority functions - all no longer made. CD40106BE is right there to buy, thankfully - but for a bit more than a dime per IIRC. Also, sometimes it helps to also check TI’s stock when distributors are out. Many times they had it when distributors didn’t. The only 4000 parts you won’t get in plastic dip today are 24-pin parts. Those are only available in smt or cerdip. I’ve got the entire series at home to play with, and those are fun parts indeed. They make amazing voltage controlled ring oscillators. Going from 3V to 12V is 5x frequency change at least. Also, using CD4066 and delay lines you can generate rather short pulses. Clean 15MHz clocks coming out from this old, old CMOS family are a sight to behold.
Reminds me of my journey to get TDA1453 ICs in a DIP package. After recieving 4 sets of fakes I just salvaged a known legit and working SMD part from somewhere else and adapted it to DIP.
Happened to me too with some rail to rail precision op-amps.. they were more like some LM358 op-amps. Never got my money back, gave some bad feedback and the listing was gone a few days later only to show up as a new listing some days later.
“deeplearnings” lol. Yeah, there was a “deep” learning: buy from TI if you want logic that’s still made, and CD4k is definitely still made and all 14- and 16-pin parts are available in DIP. Yes, in 2022. You could buy them in just the same in the 70s. The family is half a century old now (or maybe more?). I’m a big fan of the 4k family. It’s so easy to have it interface to and from higher voltage analog circuits. And it has essentially zero EMI running at low voltage. At 3V, you’d need awfully long wires to hear those outputs switching. It’s rather neat for teaching the CMOS speed vs voltage relationship - it’s so magnified!
Excellent. In decades of using chips properly, I never thought to wind up vcc above recommendations to determine v failure. But chips were relatively more expensive back then.
I bought 20 pcs Cd4026b all were dead. I realised later when I wanted to use them since I normally buy them randomly for future diy experiments. Nowadays I test components immediately they arrive from AliExpress
You should never use the parameters headlined on the first page of the datasheet when verifying performance. These numbers are average figures only derived from test of large numbers of devices and do not represent the spread in values that can be expected from an individual chip. Experienced engineers take no notice of the headline figures on the first page of a datasheet as they are usually there just to paint a flattering picture of the part performance. If you read the detailed specification on page 7 of the TI datasheet you will find the allowable range of hysteresis is from 1.2V to 3.4V when tested at Vcc=10V and 25C. The example tested was within the valid range as per the manufacturer's datasheet and therefore the hysteresis test is not conclusive as to the part being fake.
Some time ago I bought six "Rockwell 65C02" processors "made in 2014". Two out of six happened to be regular NMOS 6502 CPUs, and four were re-labelled 65C02 parts from 1980s and maybe 1990s. Even the case style of each one was different. If they didn't get into trouble of grinding and repainting them I'd get them anyways, at least for the price. Would even prefer them this way. Shame.
Interesting, though it may be worth a look at the 'old' Harris datasheet as that seems to imply the hysteresis voltage could be between 2.3V and 3.4V at 10V and supply voltage should be no more than 18V (with an absolute max at 20V). No guarantee that this is an old Harris part of course (may still be fake) but Ti may have improved things since the acquisition.
I've bought from "deeplearnings" before. My IC's worked, but I didn't test them out to the extent you did. Thanks for the heads up on this. If you buy a stock pile of parts and don't test them ALL right away, they've just made profit on garbage!
These might be VHC parts. If so, they make great level shifters. Power them at 5v and see at what point an input fails, guessing 11v. You will still have 5 gates to test. On ebay, I only buy bulk from on-shore brick and mortor surplus sellers. Prices are closer to new retail but now are on par with Chinese sellers.
@@Enigma758 flex-direct is my main goto. Free 3-4 day shipping, most sales are $13, and they have this cool 4x40 LCD I want more of but don't have any projects to use them.
Used him to make oscillators for music synths. Either a multi oscillator drone, all those ldr light activated theremin circuits. They're also great for making pwm LED Dimming circuits /or DC motor speed controller.
Chips from China have been pretty good for me. But, I got burnt recently. Tried to buy J113 s from AliExpress ... not a JFET, was BJT of unknown type. Had Fairchild logo, the whole bit. I should have paid close attention to comments, usually do but they were nominal price, didn't expect it!
the remarker probably didnt understand the difference between true 4xxx cmos types and common 74hc types and just went by what the pinouts were ....so assumed they were equivalent ...
Pretty sure I'm the viewer who first told you about having fake 40106s. I was using them in a simple automotive circuit so they never got powered over about 13.8V. I went through several chips before I realized what was happening. Funny thing is that the first chip I used lasted for about a week in the van before frying itself. When I brought it back to the bench the next few I tried all worked fine at 5V but failed within minutes at 12V. I think I still have the remaining bogus chips here if you would like me to send them to you. I don't trust them in actual projects now, even at 5V.
I also have fake chips that have the same problem. they do indeed exist. i use them at 5v and they work fine, but yeah. EDIT: yeah, i should mention its this fake chip, marks as the cd40106. mine reliably die at 12v.
Just last week, I have found that my LM393 are fake, It works very well but it' s inputs consume a few hundred micro-amps instead of a hundred nano-amp. And this can be bad in a few applications. Avoid anything that label TI.
Probably these fake chips are the reject of the scrap of the discard of the wafer outermost part. I bought a stock of twenty 555 chips: burned a quarter on the breadboard trial session and another quarter onto the corresponding definitive pcb 🤦👎
That's so infuriating. Then you check what you did wrong for some time before you discovered that they're fakes. I redid a whole pcb before I realised that the transistors were fakes, thinking it was my mistake.
The actual frustration of using counterfeit parts is that the crunchy numbers ,given in the datasheet, that we require to do our design calculations is not what the parts actually present us with. On top they are genuinely unreliable. Out of a 100 of them, you might encounter one part that actually works, but the main point of getting them cheap per unit gets overlooked that way. I'm a mere hobbyist. So it's somewhat ok for me to get away with them in the low end projects but if I ever intend to do something serious, like doing some mod to my car or motorcycle, I'm ready to spend that extra dollar or cent just for the sake of reliability.
Probably CMOS all round, none of the output drop that TTL has on the scope as the voltage changes, at 5V TTL will never be above 4V unless you have a pull up resistor externally.
They buy in batches, so you can get good stuff until they get a dodgy one slipped in somewhere. Just takes one person who figures they can swap the real with the fake and make a profit quietly along the way.
I'll pay a premium for known NOS parts - humidity, surface corrosion, those are nothing compared to getting fake stuff that you don't know ever worked. As long as the silicon still works, the other issues are usually fixable.
Problem is that ebay is convenient and cheap/free shipping. And people/me are usually lazy and those damn chinese have to ruin everything. Wouldn't mind if they sold reclaimed parts with original marking and some bent leads instead of fake laser marked shit.
At the very least when I look on evil buy, I use the filter to block out China parts listings. Then at least I got a US vendor who may be selling legit parts, or at least not live on the other side of the world and not care about feedback or customer anger.
there was a seller on ebay selling supposedly 4116 ram chips that turned out to be remarked 4164s, definitely not drop in compatible, luckily they tested good as 4164s, but would fail if fitted in place of 4116s as they have a +12v supply where the +5v pin is on the 4164!! i messaged the seller they should alter the listing and state incorrectly marked, but they never changed it ..
A lot of DIY synth builder use 40106's for square wave oscillators. Very popular from that crowd - it seems like the DIY audio/synth crowd is not super-sophisticated when it comes to electronics, but may not use them in a way which pushes the chips anywhere near advertised specs, and therefore may not care about a chip that doesn't meet specs. Doesn't mean its OK to sell fake chips, but the point is that some people may not care/notice and that's why there's a market for them. Then again, it may be that there are a lot of frustrated DIY synth enthusiasts that are being turned off from their hobby because of the frustration of dealing with fake parts and not understanding why things aren't working the way they are supposed to :/
In analog synthesizer and organ technology most of the components doesn't operate anywhere near their limits. Even the slowest gate will usually work fine. However, some synthesizer applications are very sensitive to leak currents and voltage off-sets so it all depends on what type of deviations the fake components are showing.
I believe I may have bought some of these same chips, same marking issues, same hysteresis issues, dies at far lower voltage than expected, etc. They were part of a 4000 series assortment from a Chinese Amazon seller though.
Just got an Amplifier board today from amazon. An yep it too was fake. They didn’t even use the same footprint. What they think we won’t pry that Heatsink up and look. Jeez I wonder how many people don’t even question it. Has to be worth it.
I didn't get a clear understanding of how you are measuring the hysteresis limits. A guess: the positive output of Schmitt is charging the capacitor, up to the rising trigger threshold, which then causes the output to go low, and thus cut the charging. Then the capacitor drains through the resistor, until the lower threshold is reached, at which point the Schmitt output goes high and resumes charging. Is that about right? I looked back at the chip of the day video on this part, but I don't think you showed this particular oscillator circuit.
Funny thing is standard TTL, at least the 1970's era stuff, would happily run on 14V with almost no failure, simply because that 7V rating is conservative, and you really need to be running the part at 125C to make one in 100 fail at 7V. The parts are likely remarked 74C or HC/T parts, mostly because there are almost no fabs left who can deal with those old process steps any more, the silicon used to make the IC would, if they used the original feature sizes and diffusion depths, mean the actual silicon area used would be close to $0.50 just on the wafer, while the modern HC/T uses around $0.01 of silicon, as the die is so much smaller. Not with features that you can see and examine with the naked eye any more, you need a microscope. Incidentally, try your experiment again with old TTL IC's, even though you will have to change oscillator values to get them to work with the high input bias currents, and you will probably find a good number of the TTL IC's will work well past 7V.
At >7V it won’t necessarily fail in breakdown, but it’ll run hot and electromigration becomes a bit of a problem. For short-lived projects of course not a problem at all. I’ve got a few 4K CMOS parts that survived almost 30V. They run super hot when you get them switching fast at that voltage though. Hot enough to feel it at just a couple hundred kHz. At a couple MHz they pretend to be ECL in terms of heat.
@@absurdengineering Well, I did run some 54S00 parts at 14MHz as a clock buffer, using 3 cerdip packages stacked on top of each other, and with a decoupling capacitor across the top of the package. It ran hot, with one single gate as inverter, driving the other 11 to boost the clock power ability. Easily gave enough drive to use 50R coax to connect it to the scope with good edges.
Just received 5 LM2917 from China. Worked as f-v converters, but the internal voltage regulators had fallen out of the chips in transit. They need to provide better packaging. An unkind person would suggest that they are LM2907's rebranded. AliExpress refunded my payment. One ended up as a 13 pin. Had to straighten all the pins that were bent because of the apparent foam packaging shortage in China.
@@big0bad0brad The spring contacts/legs are removable in the standard Chinese ones, so if you have broken legs, remove the good ones from the worse socket, and do some transplanting, or remove them from cheaper sockets from China, ie, 8,14 or 16 dil ones, if you don't want to waste your 40's. Hope that you haven't already dumped them.
@@big0bad0brad Another thing that you might not realize about those Chinese sockets. Because the pins are inserted into the body after moulding, they don't need such a big end overlap, so if you can make a 16 pin one by simply soldering in two 8 pin ones, butted together, or a 14 by removing one end, by one pair of pins. With 5 x 8 pins, sawn in half to make sil sockets, you have a 40 pin one, once soldered into the pcb. If I ordered specific sockets today, I'd be fortunate if they arrived within 2 months, and sometimes they never arrive. I keep lots of 8 pin ones, and adapt.
I wouldn’t ever buy any parts from eBay or Ali Express unless they’re no longer produced and can not be substituted with a modern part. The prices from possibly reputable offshore sellers are almost the same as buying from Digikey or direct from TI anyway. I needed to get a small quantity of adjustable LDO regulators recently and direct from TI was less expensive and faster than dealing with distributors anyway.
Never was particularly interested in sorting my coffee can full of logic chips and some of the vintage ones that I got in junkboxes. Now I realize that they may be pure gold when needed, rather than having to buy a bunch of nice looking junk from China and having to do in depth testing before being able to use. Which is impractical. I also question the mentality of making fake logic chips as if they have a clepto compulsion to sell bad stuff for small profit per unit. Another insane thing in modern world. Guess it comes down to who you are dealing with. No ethics at all.
They aren't always made, sometimes this is how electronics get recycled, no joke. For better or worse, these Chinese companies will do anything they get paid for. You want them to manufacture something good? Sure. You want crap quality? They'll do it. You've got a brick of pulled ICs wrapped up in cling wrap like it was a package of drugs? They'll buy it. Because someone else will too.
I got done on fake FTDI chips. They worked fine until they suffered some kind of latch up condition and went dead short and melted off the circuit board
I stopped buying parts from China long ago. Even sellers that look like they're based in the US but aren't I avoid like the plague. I wanted an AD633. Guess what I did ? I bought them from Digikey. Problem solved.
Digikey is selling these for $.53 each now. I believe you when you mention they were $.11 when you filmed this. Crazy to see how prices for everything has gone up.
@@IMSAIGuy I understand how hysteresis works on the schmitt triggers and inputting a triangle ramp to test the range. But is there a special feature or technique on that scope to test the range without monitoring the output ?
@@rcarioca I am not inputting a ramp. the output is fed to the input and the inverter will swap (change) on the high voltage and low voltage of the hysterisis. nothing fancy needed on the scope
I think you may have meant to say, " you really SHOULDNT pay new prices" rather than, "youre Not going to pay new prices- I guarantee many are paying well over this on a daily basis so much so in fact, the cargo ships from Big Rock Candy Mountain are overflowing and threatening to capsize! Im trying to figure out something to do with my 40106 fakes guitar-effect related.
Every CD4000 dip part I bought from TI this year was laser marked. Yeah, I did buy all they made to have a lab stock of known good parts. Some markings were actually hard to read without proper light - they were quite filigree, thin lines. None had screen printing or molded numbers, although perhaps if TI uses more than one packager, they may come in two varieties. There are no “out of spec” parts you’ll buy new that are not fake. I’ve tested many chips and they were usually very close to typical values or better, and the process control seems to be much tighter than it was decades ago.
Some of the Chinese manufacturers will take defective parts, package & mark them, and sell them through back channels, resulting in new, legit, yet out of spec parts. There are some _very_ shady manufacturers out there.
I have bought a lot of ICs , transistors and passive components from China ... are they fake ? I have no idea. I wonder if all those cheap electronic gadgets they sell from China use fake chips ?
Often the "grey market" has parts that are left over, or rejected, from manufacturing runs. You have no idea what the quality of these things are going to be. I learned my lesson when a cheap voltage regulator wrecked about $20 of other components when it failed.
I'm still taking a chance at burning up my stock of amazon 74hc595's on a project. I only had one fail, and I'm not certain it wasn't my fault. Even if all is well, I'm no longer buying parts outside of trusted suppliers. Ya cant even trust 1% resistors on the internet shopping market.
Actually, the one time I bought a pack of assorted 1% resistors on eBay, they were indeed as stated-I tested some of them, and found that: 1.they were resistors(axial lead, film. 1/4 watt). 2.they were actually within 1% Note:I did NOT check if they could actually dissipate 1/4 watt, so they could be deficient that way. So perhaps resistors are a reasonable risk.
@@m.k.8158 most of my requirements are satisfied with 5% tolerance. I would still buy the ebay 1%, but I wouldnt go plugging them right into a 1% critical circuit without measuring them first.
I don't get it🤔...why is it fake if you're driving the chip voltage pass the recommended specification to prove it will fail? You need to return the seller chips😇
@@IMSAIGuyYes but that doesn't mean you should drive it to destruction...a bad component in my opinion...SCR that would overheat with the correct bias voltage or a logic chip that leak between pins...etc
I assume you have never designed with 4000 series or 74C series logic before. I try to educate people. please read the data sheets. these are not 5V parts.
For the hobby people low price stuff and learning a new design circuit is fun. However, chasing your tail when everything is OK and just bad new parts. 😏 Grrr. This is where experience helps as in knowing how to run these tests to enhance the learning curve. 😎 Thanks a lot for something to try out. PS. These factory people are very creative even if the components only sort of work when making chips. Something over my head.
There was 555 likes. I just had to make it a double instead of single by clicking the like. Not it has been changed to 556. :D In addition to relabeling/fakes, there is also a risk of getting factory rejects.
Comparing any parameter with just a typical datasheet figure is nothing to make any reliable conclusion. Certain manufacturing spread is a normal thing, so you should always compare with the specified limits (min-max), as those are the really bounding values. The thing is, the parameters and their limits of these standard logic families are not set for the individual manufacturer or even part, but standardized for the type family. And are set so, they are not that difficult to achieve by practically anyone designing these devices. That does not only mean various makers, but also various processes used for the same part number by the single maker, the device from 1980 (or from the early years of the CMOS4000) will have the same datasheet as the one made today, but the exact process had changed so this component had to be redesigned at least 5x over that time, leading to 5 different real mean parameter values. So you will find similar differences between a part 40 years old and a new one, even so both will really be genuine from the same maker. However this does not mean your parts are genuine good parts or so. Because there is another way how substandard parts come to market: They are genuinely made by the company (here TI), but the end-of-line test reject units stolen from the testing floor (normally these bad ones are supposed to be destroyed) and then sold on such shady places. Normally the production yields certain percentage of parts that are bad in some way. And apart from really dead devices, the most frequent defects cause the breakdown voltages to become lower than specified. As a side effect, in CMOS it may cause seemingly faster speed, but the too low breakdown was the reason why they were rejected. For you vs seller relation this does not mean anything different, the devices are counterfeit either way, regardless if they are remarked other part or other maker, or whether they arerejects that should have never left the manufacturers premises but were stolen, smuggled out and then sold...
I really don't even mind buying used parts, but to re-label one part (or even an equivalent) is just straight and outright deceit & manipulation to me. I almost never receive OEM parts when buying from AliExpress and often get unexpected results from circuits I build using these components. At the end of the day - its just not worth buying from the Chinese because they're dishonest.
@@IMSAIGuy Your misunderstanding is partly my fault. Sometimes I don't always express myself clearly. One time I got in a long argument on line several years ago because I never made myself clearly. My apologies for not being clear.
Our thoughts go out to the brave parts that gave their lives to prove a point.
Rest in piece, and may the holy smoke be with you.
F
Ah yes, the magic smoke of doom, and the unmistakable smell of time trouble and expense, not to mention a few swear words!.
Years ago I bough some 1F "super capacitors" from China. I put them on a shelf without measuring them. When I did get round to using them they were under 1uF. Lesson learned.
lmao
You can buy the entire 14/16
pin CD4000 line from TI in DIP. Notable exception is 4000 “mix and match” gate, 4008 adder, and 14530 majority functions - all no longer made. CD40106BE is right there to buy, thankfully - but for a bit more than a dime per IIRC. Also, sometimes it helps to also check TI’s stock when distributors are out. Many times they had it when distributors didn’t. The only 4000 parts you won’t get in plastic dip today are 24-pin parts. Those are only available in smt or cerdip.
I’ve got the entire series at home to play with, and those are fun parts indeed. They make amazing voltage controlled ring oscillators. Going from 3V to 12V is 5x frequency change at least. Also, using CD4066 and delay lines you can generate rather short pulses. Clean 15MHz clocks coming out from this old, old CMOS family are a sight to behold.
when the shipping to your country is 150/200$ from official sellers, you don't have a choice but to buy from china.
@@mahadotube I agree. That’s unfortunate.
"So I went on E-bay"....well there's your problem....!
Reminds me of my journey to get TDA1453 ICs in a DIP package. After recieving 4 sets of fakes I just salvaged a known legit and working SMD part from somewhere else and adapted it to DIP.
Ive stopped buying semiconductors from china years ago. It's a mess, and it isn't worth the time.
Where do you buy from instead? Digikey?
@@harrysvensson2610 Digikey, arrow, mouser, farnell, TME and even LCSC, which is based in china but all in all pretty reliable.
almost everything is made in china 🤣
@@artecno882 unfortunately
Is anyone going to tell him that didn't pick this stuff from China?
Happened to me too with some rail to rail precision op-amps.. they were more like some LM358 op-amps. Never got my money back, gave some bad feedback and the listing was gone a few days later only to show up as a new listing some days later.
Oh boy, don't even try to enter the MOSFET Chinese parts market since almost everyone fails at the RDS on specs.
“deeplearnings” lol. Yeah, there was a “deep” learning: buy from TI if you want logic that’s still made, and CD4k is definitely still made and all 14- and 16-pin parts are available in DIP. Yes, in 2022. You could buy them in just the same in the 70s. The family is half a century old now (or maybe more?). I’m a big fan of the 4k family. It’s so easy to have it interface to and from higher voltage analog circuits. And it has essentially zero EMI running at low voltage. At 3V, you’d need awfully long wires to hear those outputs switching.
It’s rather neat for teaching the CMOS speed vs voltage relationship - it’s so magnified!
Excellent.
In decades of using chips properly, I never thought to wind up vcc above recommendations to determine v failure.
But chips were relatively more expensive back then.
I bought 20 pcs Cd4026b all were dead. I realised later when I wanted to use them since I normally buy them randomly for future diy experiments. Nowadays I test components immediately they arrive from AliExpress
You should never use the parameters headlined on the first page of the datasheet when verifying performance. These numbers are average figures only derived from test of large numbers of devices and do not represent the spread in values that can be expected from an individual chip. Experienced engineers take no notice of the headline figures on the first page of a datasheet as they are usually there just to paint a flattering picture of the part performance.
If you read the detailed specification on page 7 of the TI datasheet you will find the allowable range of hysteresis is from 1.2V to 3.4V when tested at Vcc=10V and 25C. The example tested was within the valid range as per the manufacturer's datasheet and therefore the hysteresis test is not conclusive as to the part being fake.
Some time ago I bought six "Rockwell 65C02" processors "made in 2014". Two out of six happened to be regular NMOS 6502 CPUs, and four were re-labelled 65C02 parts from 1980s and maybe 1990s. Even the case style of each one was different. If they didn't get into trouble of grinding and repainting them I'd get them anyways, at least for the price. Would even prefer them this way. Shame.
Interesting, though it may be worth a look at the 'old' Harris datasheet as that seems to imply the hysteresis voltage could be between 2.3V and 3.4V at 10V and supply voltage should be no more than 18V (with an absolute max at 20V). No guarantee that this is an old Harris part of course (may still be fake) but Ti may have improved things since the acquisition.
I've bought from "deeplearnings" before. My IC's worked, but I didn't test them out to the extent you did. Thanks for the heads up on this. If you buy a stock pile of parts and don't test them ALL right away, they've just made profit on garbage!
Would be interesting to see the IC die, there's usually some kind of a vendor marking. And also compare it with genuine TI part and the 74HC series.
Do these guys ever bust one open and see what the die looks like? I have seen them do it with processors, I think but not regular ICs
These might be VHC parts. If so, they make great level shifters. Power them at 5v and see at what point an input fails, guessing 11v. You will still have 5 gates to test. On ebay, I only buy bulk from on-shore brick and mortor surplus sellers. Prices are closer to new retail but now are on par with Chinese sellers.
Hi, would you be willing to name a few of those vendors?
@@Enigma758 flex-direct is my main goto. Free 3-4 day shipping, most sales are $13, and they have this cool 4x40 LCD I want more of but don't have any projects to use them.
Used him to make oscillators for music synths. Either a multi oscillator drone, all those ldr light activated theremin circuits. They're also great for making pwm LED Dimming circuits /or DC motor speed controller.
Fake op amps are the worst!!! I bought some eBay rail to rail op amps that were definitely not and would only trust them as voltage comparators.
Chips from China have been pretty good for me. But, I got burnt recently. Tried to buy J113 s from AliExpress ... not a JFET, was BJT of unknown type. Had Fairchild logo, the whole bit. I should have paid close attention to comments, usually do but they were nominal price, didn't expect it!
the remarker probably didnt understand the difference between true 4xxx cmos types and common 74hc types and just went by what the pinouts were ....so assumed they were equivalent ...
Pretty sure I'm the viewer who first told you about having fake 40106s. I was using them in a simple automotive circuit so they never got powered over about 13.8V. I went through several chips before I realized what was happening. Funny thing is that the first chip I used lasted for about a week in the van before frying itself. When I brought it back to the bench the next few I tried all worked fine at 5V but failed within minutes at 12V. I think I still have the remaining bogus chips here if you would like me to send them to you. I don't trust them in actual projects now, even at 5V.
I also have fake chips that have the same problem. they do indeed exist. i use them at 5v and they work fine, but yeah.
EDIT: yeah, i should mention its this fake chip, marks as the cd40106. mine reliably die at 12v.
Just last week, I have found that my LM393 are fake, It works very well but it' s inputs consume a few hundred micro-amps instead of a hundred nano-amp. And this can be bad in a few applications. Avoid anything that label TI.
They are probably Suzhou HYC Technology CO., LTD with a crooked TI logo engraved next to it
with many Chinese companies, why did you assume that one?
@@IMSAIGuy It's marked "HYC", which should not be there on TI part, pretty sure
make a giant ring oscillator with them
Probably these fake chips are the reject of the scrap of the discard of the wafer outermost part. I bought a stock of twenty 555 chips: burned a quarter on the breadboard trial session and another quarter onto the corresponding definitive pcb 🤦👎
That's so infuriating. Then you check what you did wrong for some time before you discovered that they're fakes. I redid a whole pcb before I realised that the transistors were fakes, thinking it was my mistake.
@@xminusone1 😓
The actual frustration of using counterfeit parts is that the crunchy numbers ,given in the datasheet, that we require to do our design calculations is not what the parts actually present us with. On top they are genuinely unreliable. Out of a 100 of them, you might encounter one part that actually works, but the main point of getting them cheap per unit gets overlooked that way. I'm a mere hobbyist. So it's somewhat ok for me to get away with them in the low end projects but if I ever intend to do something serious, like doing some mod to my car or motorcycle, I'm ready to spend that extra dollar or cent just for the sake of reliability.
What if we measure quiescent power consumption to check if they are bi-polar or CMOS devices ?
Probably CMOS all round, none of the output drop that TTL has on the scope as the voltage changes, at 5V TTL will never be above 4V unless you have a pull up resistor externally.
why do they bother ? In China they make pirated knock offs of already pirated knock offs.
Needs a trigger warning (pun intended) - "Warning: Innocent Logic Chip died in the making of this video". LOL.
I've bought from that same vendor, considered them trustworthy -- I guess they all source from the same shady places.
They buy in batches, so you can get good stuff until they get a dodgy one slipped in somewhere. Just takes one person who figures they can swap the real with the fake and make a profit quietly along the way.
I'll pay a premium for known NOS parts - humidity, surface corrosion, those are nothing compared to getting fake stuff that you don't know ever worked. As long as the silicon still works, the other issues are usually fixable.
25x CD40106 are less than $12 on Digikey, genuine from Ti. 58 cents a piece. Honestly buying from China in many ways is a false economy.
Problem is that ebay is convenient and cheap/free shipping.
And people/me are usually lazy and those damn chinese have to ruin everything.
Wouldn't mind if they sold reclaimed parts with original marking and some bent leads instead of fake laser marked shit.
At the very least when I look on evil buy, I use the filter to block out China parts listings. Then at least I got a US vendor who may be selling legit parts, or at least not live on the other side of the world and not care about feedback or customer anger.
there was a seller on ebay selling supposedly 4116 ram chips that turned out to be remarked 4164s, definitely not drop in compatible, luckily they tested good as 4164s, but would fail if fitted in place of 4116s as they have a +12v supply where the +5v pin is on the 4164!! i messaged the seller they should alter the listing and state incorrectly marked, but they never changed it ..
Laughs in 1/4 inventory of fake CD40106 🤣
Buy stuff from Digikey/Mouser. Doesn't worth the time trouble shooting faulty/fake parts.
A lot of DIY synth builder use 40106's for square wave oscillators. Very popular from that crowd - it seems like the DIY audio/synth crowd is not super-sophisticated when it comes to electronics, but may not use them in a way which pushes the chips anywhere near advertised specs, and therefore may not care about a chip that doesn't meet specs. Doesn't mean its OK to sell fake chips, but the point is that some people may not care/notice and that's why there's a market for them.
Then again, it may be that there are a lot of frustrated DIY synth enthusiasts that are being turned off from their hobby because of the frustration of dealing with fake parts and not understanding why things aren't working the way they are supposed to :/
In analog synthesizer and organ technology most of the components doesn't operate anywhere near their limits. Even the slowest gate will usually work fine. However, some synthesizer applications are very sensitive to leak currents and voltage off-sets so it all depends on what type of deviations the fake components are showing.
Send some to CuriousMarc they have looked at dies in the past.
I suspect they are floor sweep /rejects packed and hawked on eBay
I believe I may have bought some of these same chips, same marking issues, same hysteresis issues, dies at far lower voltage than expected, etc. They were part of a 4000 series assortment from a Chinese Amazon seller though.
Serious manufacturers can't use too old parts event when they are in their own stock.
The fake chip problem is outrageous coming out of China. Most of the time I'm just paying for digikey parts for peace of mind.
Now I want to check mine. I bought 30 on amazon they work like these. But haven’t tested the maximum thresholds.
Just got an Amplifier board today from amazon. An yep it too was fake. They didn’t even use the same footprint. What they think we won’t pry that Heatsink up and look. Jeez I wonder how many people don’t even question it. Has to be worth it.
I didn't get a clear understanding of how you are measuring the hysteresis limits. A guess: the positive output of Schmitt is charging the capacitor, up to the rising trigger threshold, which then causes the output to go low, and thus cut the charging. Then the capacitor drains through the resistor, until the lower threshold is reached, at which point the Schmitt output goes high and resumes charging. Is that about right? I looked back at the chip of the day video on this part, but I don't think you showed this particular oscillator circuit.
Funny thing is standard TTL, at least the 1970's era stuff, would happily run on 14V with almost no failure, simply because that 7V rating is conservative, and you really need to be running the part at 125C to make one in 100 fail at 7V. The parts are likely remarked 74C or HC/T parts, mostly because there are almost no fabs left who can deal with those old process steps any more, the silicon used to make the IC would, if they used the original feature sizes and diffusion depths, mean the actual silicon area used would be close to $0.50 just on the wafer, while the modern HC/T uses around $0.01 of silicon, as the die is so much smaller. Not with features that you can see and examine with the naked eye any more, you need a microscope.
Incidentally, try your experiment again with old TTL IC's, even though you will have to change oscillator values to get them to work with the high input bias currents, and you will probably find a good number of the TTL IC's will work well past 7V.
At >7V it won’t necessarily fail in breakdown, but it’ll run hot and electromigration becomes a bit of a problem. For short-lived projects of course not a problem at all. I’ve got a few 4K CMOS parts that survived almost 30V. They run super hot when you get them switching fast at that voltage though. Hot enough to feel it at just a couple hundred kHz. At a couple MHz they pretend to be ECL in terms of heat.
@@absurdengineering Well, I did run some 54S00 parts at 14MHz as a clock buffer, using 3 cerdip packages stacked on top of each other, and with a decoupling capacitor across the top of the package. It ran hot, with one single gate as inverter, driving the other 11 to boost the clock power ability. Easily gave enough drive to use 50R coax to connect it to the scope with good edges.
Just received 5 LM2917 from China. Worked as f-v converters, but the internal voltage regulators had fallen out of the chips in transit. They need to provide better packaging. An unkind person would suggest that they are LM2907's rebranded. AliExpress refunded my payment. One ended up as a 13 pin. Had to straighten all the pins that were bent because of the apparent foam packaging shortage in China.
I nearly lost my mind when I got a package of DIP-40 sockets with no foam or anything in a plastic bag. I was able to salvage like maybe 15% of them.
@@big0bad0brad The spring contacts/legs are removable in the standard Chinese ones, so if you have broken legs, remove the good ones from the worse socket, and do some transplanting, or remove them from cheaper sockets from China, ie, 8,14 or 16 dil ones, if you don't want to waste your 40's. Hope that you haven't already dumped them.
@@peterrhodes5663 Y'know, I should have thought of that myself, and I do still have them. Thanks for the tip!
@@big0bad0brad Another thing that you might not realize about those Chinese sockets. Because the pins are inserted into the body after moulding, they don't need such a big end overlap, so if you can make a 16 pin one by simply soldering in two 8 pin ones, butted together, or a 14 by removing one end, by one pair of pins. With 5 x 8 pins, sawn in half to make sil sockets, you have a 40 pin one, once soldered into the pcb. If I ordered specific sockets today, I'd be fortunate if they arrived within 2 months, and sometimes they never arrive. I keep lots of 8 pin ones, and adapt.
Maybe there's special counterfeit versions of semiconductor datasheet published for devices made in China
I wouldn’t ever buy any parts from eBay or Ali Express unless they’re no longer produced and can not be substituted with a modern part. The prices from possibly reputable offshore sellers are almost the same as buying from Digikey or direct from TI anyway. I needed to get a small quantity of adjustable LDO regulators recently and direct from TI was less expensive and faster than dealing with distributors anyway.
Never was particularly interested in sorting my coffee can full of logic chips and some of the vintage ones that I got in junkboxes. Now I realize that they may be pure gold when needed, rather than having to buy a bunch of nice looking junk from China and having to do in depth testing before being able to use. Which is impractical. I also question the mentality of making fake logic chips as if they have a clepto compulsion to sell bad stuff for small profit per unit. Another insane thing in modern world. Guess it comes down to who you are dealing with. No ethics at all.
They aren't always made, sometimes this is how electronics get recycled, no joke. For better or worse, these Chinese companies will do anything they get paid for. You want them to manufacture something good? Sure. You want crap quality? They'll do it. You've got a brick of pulled ICs wrapped up in cling wrap like it was a package of drugs? They'll buy it. Because someone else will too.
I got done on fake FTDI chips. They worked fine until they suffered some kind of latch up condition and went dead short and melted off the circuit board
I stopped buying parts from China long ago. Even sellers that look like they're based in the US but aren't I avoid like the plague. I wanted an AD633. Guess what I did ? I bought them from Digikey. Problem solved.
Digikey is selling these for $.53 each now. I believe you when you mention they were $.11 when you filmed this. Crazy to see how prices for everything has gone up.
I don't know too much about advanced features of that scope. Not sure how you measure hysteresis by only looking at input waveform at start of video
ua-cam.com/video/o31_fnTqUOw/v-deo.html
@@IMSAIGuy I understand how hysteresis works on the schmitt triggers and inputting a triangle ramp to test the range. But is there a special feature or technique on that scope to test the range without monitoring the output ?
@@rcarioca I am not inputting a ramp. the output is fed to the input and the inverter will swap (change) on the high voltage and low voltage of the hysterisis. nothing fancy needed on the scope
@@IMSAIGuy ok thanks.. Now I fully understand. I never thought of checking the hysteresis that way before.. Perfect.
I think you may have meant to say, " you really SHOULDNT pay new prices" rather than, "youre Not going to pay new prices- I guarantee many are paying well over this on a daily basis so much so in fact, the cargo ships from Big Rock Candy Mountain are overflowing and threatening to capsize! Im trying to figure out something to do with my 40106 fakes guitar-effect related.
Every CD4000 dip part I bought from TI this year was laser marked. Yeah, I did buy all they made to have a lab stock of known good parts. Some markings were actually hard to read without proper light - they were quite filigree, thin lines. None had screen printing or molded numbers, although perhaps if TI uses more than one packager, they may come in two varieties.
There are no “out of spec” parts you’ll buy new that are not fake. I’ve tested many chips and they were usually very close to typical values or better, and the process control seems to be much tighter than it was decades ago.
Some of the Chinese manufacturers will take defective parts, package & mark them, and sell them through back channels, resulting in new, legit, yet out of spec parts. There are some _very_ shady manufacturers out there.
Thanks for the reminders. Chinese have poor QC and fraud is their game. We should support quality American companies.
I have bought a lot of ICs , transistors and passive components from China ... are they fake ? I have no idea.
I wonder if all those cheap electronic gadgets they sell from China use fake chips ?
yes and yes
Good tests!
Don't forget the propagation delay of cables between!
Often the "grey market" has parts that are left over, or rejected, from manufacturing runs. You have no idea what the quality of these things are going to be. I learned my lesson when a cheap voltage regulator wrecked about $20 of other components when it failed.
What do you expect from china.
Perhaps they are CMOS 4584?
On AliExpress I don't have problem if you search you find lot of module fake and orignal so always check specifications they write complete details
Very unsurprising!
De-Cap some and compare the silicon
I'm still taking a chance at burning up my stock of amazon 74hc595's on a project. I only had one fail, and I'm not certain it wasn't my fault. Even if all is well, I'm no longer buying parts outside of trusted suppliers. Ya cant even trust 1% resistors on the internet shopping market.
Actually, the one time I bought a pack of assorted 1% resistors on eBay, they were indeed as stated-I tested some of them, and found that:
1.they were resistors(axial lead, film. 1/4 watt).
2.they were actually within 1%
Note:I did NOT check if they could actually dissipate 1/4 watt, so they could be deficient that way.
So perhaps resistors are a reasonable risk.
@@m.k.8158 most of my requirements are satisfied with 5% tolerance. I would still buy the ebay 1%, but I wouldnt go plugging them right into a 1% critical circuit without measuring them first.
@@chetronics I agree of course, just as long as you trust your meter enough.
I don't get it🤔...why is it fake if you're driving the chip voltage pass the recommended specification to prove it will fail? You need to return the seller chips😇
the real parts are good to 20 volts. read the datasheet: www.ti.com/lit/ds/symlink/cd40106b.pdf
@@IMSAIGuyYes but that doesn't mean you should drive it to destruction...a bad component in my opinion...SCR that would overheat with the correct bias voltage or a logic chip that leak between pins...etc
@@basimpsn the datasheet says 20V if you put 20V on Vcc it burns up. I ran the part as it is to be used. I did not try to make it fail.
I assume you have never designed with 4000 series or 74C series logic before. I try to educate people. please read the data sheets. these are not 5V parts.
@@IMSAIGuy Ok just downloaded the full datasheet you're correct my apology 👌
Who makes this fakes
For the hobby people low price stuff and learning a new design circuit is fun. However, chasing your tail when everything is OK and just bad new parts. 😏 Grrr.
This is where experience helps as in knowing how to run these tests to enhance the learning curve. 😎 Thanks a lot for something to try out.
PS. These factory people are very creative even if the components only sort of work when making chips. Something over my head.
IT seems insane to bother to manufacture low value chips as fakes rather than just do it right. Their minds work strangely.
@@tomsherwood4650 Yes, I agree.
There was 555 likes. I just had to make it a double instead of single by clicking the like. Not it has been changed to 556. :D
In addition to relabeling/fakes, there is also a risk of getting factory rejects.
Always TERRIB[LY] InTeReStiNg... Thanks so much for sharing your gray-matter observations (otherwise known as GMO)
Comparing any parameter with just a typical datasheet figure is nothing to make any reliable conclusion. Certain manufacturing spread is a normal thing, so you should always compare with the specified limits (min-max), as those are the really bounding values. The thing is, the parameters and their limits of these standard logic families are not set for the individual manufacturer or even part, but standardized for the type family. And are set so, they are not that difficult to achieve by practically anyone designing these devices. That does not only mean various makers, but also various processes used for the same part number by the single maker, the device from 1980 (or from the early years of the CMOS4000) will have the same datasheet as the one made today, but the exact process had changed so this component had to be redesigned at least 5x over that time, leading to 5 different real mean parameter values. So you will find similar differences between a part 40 years old and a new one, even so both will really be genuine from the same maker.
However this does not mean your parts are genuine good parts or so.
Because there is another way how substandard parts come to market: They are genuinely made by the company (here TI), but the end-of-line test reject units stolen from the testing floor (normally these bad ones are supposed to be destroyed) and then sold on such shady places. Normally the production yields certain percentage of parts that are bad in some way. And apart from really dead devices, the most frequent defects cause the breakdown voltages to become lower than specified. As a side effect, in CMOS it may cause seemingly faster speed, but the too low breakdown was the reason why they were rejected.
For you vs seller relation this does not mean anything different, the devices are counterfeit either way, regardless if they are remarked other part or other maker, or whether they arerejects that should have never left the manufacturers premises but were stolen, smuggled out and then sold...
I really don't even mind buying used parts, but to re-label one part (or even an equivalent) is just straight and outright deceit & manipulation to me. I almost never receive OEM parts when buying from AliExpress and often get unexpected results from circuits I build using these components. At the end of the day - its just not worth buying from the Chinese because they're dishonest.
Why burn a good chip?
you did not watch the entire video. these are not 'good' chips.
@@IMSAIGuy you burned the bad and bjrned a couple good ones! And, yes, I watched the entire video!
ok, I remember. I did kill some 7414 parts.
@@IMSAIGuy Your misunderstanding is partly my fault. Sometimes I don't always express myself clearly. One time I got in a long argument on line several years ago because I never made myself clearly. My apologies for not being clear.
@@charlesmangum2100 no problem. I record videos 3 weeks in advance and so don't remember what I did
More likely factory rejects, which are very common.