We’re all aware that there are plenty of fake components to be found if you’re prepared to look in the right places, and that perhaps too-good-to-be-true chip offers on auction sites might turn out to have markings which rub off to reveal something completely different underneath. [IMSAI Guy] saw a batch of OP-07 laser-trimmed op-amps at a bargain price, so picked them up for an investigation. You can take a look at the video below the break.
A perfect op-amp has a zero volt output when both of its inputs are at the same voltage, but in practice no real device approaches this level of perfection. It’s referred to as the offset voltage, and for instrumentation work where a low offset voltage is important there are parts such as the OP-07 which have each been adjusted using a laser to trim their components for the lowest offset. This process is expensive, so naturally so are genuine OP-07s.
Identifying real versus fake op-amps in this case is as simple as hooking the chip up as a unity gain non-inverting amplifier and measuring the voltage on the output (we can’t help a tinge of envy at that Keithley 2015 THD multimeter!), from which measurement the fakes should be clearly visible. First up are some 741s with their > 1 mV offsets (though an outlier 741 had a 40μV offset) to show what a cheap op-amp could be expected to do, then we see the OP-07s. Immediately with an offset of > 1.2 mV we can tell that they’re fake, which as he admits for the price is hardly a surprise. Meanwhile we’ll keep an eye out for Korean-made 741s like the outlier low-offset device.
If you’re interested by op-amp internals may we suggest a look at the first IC op-amp, meanwhile this isn’t the first fake chip we’ve seen.
Thanks BaldPower for the tip.
Can’t you just put 2 cheap opamps in series, with the second one with a trimpot to adjust the first ones bias to zero?
That’s assuming that the offset isn’t going to drift by temperature or overtime. Now you have offsets from both parts!
You can add small amount of offsets as a bias to existing circuit without adding a 2nd opamp just by adding a DC summing node.
Temperature drift.. *sigh*🙄
Then let’s just add some small heating?
Back in the old days, this was common practice, kids.
You can read about it here.:
https://en.wikipedia.org/wiki/Crystal_oven
An NTC or PTC can be used to build a regulated heating, also, I suppose.
You are adding extra power, yet another opamp, thermistor, resistors and may be a couple of transistors just to fix a fake opamp…
These day, there are things like TCXO that might have better drift specs better than the OCXO you clobbered together and uses less power.
Look at the datasheets.
Actually both the 741 and the OP07 already have trim pins for adjusting the offset.
The only thing you need to add is a potentiometer.
Every added component incoporates some additional uncertainty. For example, temperature drift. There’s a reason to pay for precision in specific applications.
But I agree, stacking a second op amp in there isn’t necessary a great solution.
741 is geriatric at the least. A very good and affordable op-amp from a decade or less later is the LM308 which has typical input offset current of 500pA. https://www.mit.edu/~6.301/LM108.pdf
Buy them from a known source/supplier instead. Because you do not know which parameters might be different/worse (if any), you need to test them all one by one (including longevity and temperature). And that is just inane, unless you actually need so many that it can pay off
I’m guessing your haven’t tried to order parts recently? Or at least nothing remotely modern, ancient parts are still somewhat available, but microcontrollers are basically unobtainable, save for some BGA devices.
What are you bellyaching about? We’re here to talk about op amps, not microcontrollers.
I bought some genuine TI NE5532 op amps from Mouser recently to replace some definitely-fake ones that came with a kit, in easy-to-handle DIP packaging. It was not a problematic experience.
There also does not appear to be any particular shortage of real Analog Devices OP07 opamps through proper distribution channels. You can have many thousands of them as soon as tomorrow, if you want.
I mean: Sure. There is a shortage of MCUs, but that’s about as relevant as the fact that I have a purple bike.
All my projects: check if there’s stock or backorder before designing. Then come back and see if the parts are still available when completing the design before ordering boards. Then check again when populating the first test assembly. Then check again before placing larger orders. Octopart has been my friend.
Telling signs: Laser engraved markings and residual scanlines from erasing the existing marks.
This guy is clearly having some fun, and that’s ok.
If you’re not willing to invest into some expensive benchtop DMM, that is twice as long as your desk (My neighbors are not going to like it if I hack a hole in the wall to mount it flush with my other measurement equipment) then you can always simply configure the opamp itself to amplify it’s offset voltage 1000x or so, by adding a single extra resistor.
For the measurement probes.
Some like those gold plated things with silli cone wires and other fancy features.
I prefer to solder wires to double (or triple) 0.1″ pitch square single row headers.
This way you can always just stick them in the breadboard, and you have your hands free to fiddle with knobs or other appendages.
I have also made some adapters myself.
Start with high quality (round) female headers for those same 0.1″ pitch single row connectors, then break out one of the bushings and solder it to a needle. Add a bit of hot glue and two layers of shrink wrap to give it some thickness.
Very nice and sharp probes. You can poke through oxide layers, stick them in between stranded wires or in breadboards, and on the back side you just juse standard “dupont” wires. These are very handy and hardly use up any space on your workbench.
You should be happy they still work at that price 😉
Yep, I agree.
A few years ago I bought 500 100nf caps and something wasnt right so I used an octopus and found out they were 100k resistors.
For f*ck sake, what are you trying to save? These Opamps available from Digi-Key right now for less than $3 a piece? Why to support Chinese IC counterfeiting industry?
I face this issue almost every day. People complain that my projects they try to build don’t work, and in 90% of cases it is due to faulty components they purchased from China on eBay. So they end up buying these components from Mouser anyway. I just don’t get it, why not to purchase from official distributors in the first place?
Relax, isn’t that basically the “lesson learned” from the article? Well, that’s what the point seemed to me anyway.
A big clue is that it’s made out of licorice.
“How To Spot A Fake Op-Amp” These days it’s easy. If the part is in stock and you can buy it, it must be fake!
The rule of thumb is that if a product has a much cheaper similar one that can be cheaply relabeled in order to become identical to the 1st one, it *will* be faked and sold where there is no quality control, that is, most far east online vendors. Never ever buy chips from Ebay/Amazon/Ali* sellers and the likes because the risk of being burned is about 100%. If you buy a low noise opamp , say a NE5532 or OPAxxx there, you can bet all your horses you’ll get a relabeled TL072. All 2SC1969, 2n3866, 2N3055 etc. transistors, 2SKxxx low noise jfets etc. sold there are fakes too, as nearly all other costly parts. Just ignore their photos and their prices.
The same principle applies to batteries, USB flash disks, processors etc. and passives too, especially electrolytic capacitors which are easy to relabel too. Only get those parts from reputable sources.
As for now, Chinese resistors and ceramic/poly caps are still decent; probably because they’re already cheap to manufacture and there’s no gain in faking them.
I purchase components for myself as well as for my work. Work get the parts ordered from local distributors, but it’s not too often that I have any critical needs in my hobby electronics, so I can safely order them from overseas. I’ve only once gotten parts that don’t work, and they came, oddly, from a local seller who was reputable. I see no need to save a few dollars on parts that NEED to work, but I can risk it for my own stuff…