Not sure if anyone will be able to answer my Q: maybe @Longman, @Valentin and a few professionals hiding amongst the tinkerers like myself know the answer. We all know about counterfeit CDs pressed in China and Mama-Russia, fake 'promo boxes' of Japanese mini-LP CDs, $1k AudioQuest interconnects offered for under $50 on eBay, and those who care they know how to avoid the scam. I am asking about headphone op-amp which may be used as an 'upgrade' for quite a few portables: OPA2107AU Datasheet says a few words about its better specs, laser-trimmed resistors: Sooo, I am willing to try it on some of my hi-end Discmans BUT: there are some choices here. $29.06 from Digi Key, $12 off US seller on eBay, $1.50 (+$4.00 s/h) from China. $12 from "US seller" whose English is even worse than mine had never arrived, and below is the photo of one bought from DigiKey (bottom) + two from China (on top) OPA2107: Is there any way to check (except for actual listening which I am going to do!) that $1.5 Chinese chips are exact copies of the 'real thing'?
It is a big problem, especially for commercial companies. Would you want a fake I.C. controlling the airbags in your car? I would hope that Digi Key would have people more experienced than any of us checking the provenance of what they sell. I suppose for a "non critical" application like HiFi the main questions are Does the part work? Does it sound O.K. My own experience with fakes is buying a couple of memory ICs from eBay for a synthesizer. Supposedly made by Siemens they have Japan moulded into the bottom of the packages. One day I will actually try them in the hope that they are genuine devices from someone like Toshiba that were pulled from a board. The incentive for remarking them would be if Siemens had the best reputation and fastest speed. Of course they might be any random 16 pin device re-marked as memory. Looking through eBay listings there were a suspicious number of listings for exactly the same part with the same date-code. That should raise some suspicions. I would imagine that with a part like an op-amp it might be even more difficult to spot a fake. If they remarked a $2 CA3140 are you sure you would be able to measure or hear the difference ?
In the picture you posted, all 3 parts seem to look a little bit different, but the 2 on the top seem more similar than the bottom one. Of course, simple judgement will tell you the DigiKey one is the most likely to be the genuine, but that's also the most expensive (too expensive in my opinion). What I would do is buy a sample from DigiKey and compare it to the one sold on ebay. In terms of font used, how the capsule looks, if there is any color difference, markings on the back. Keep in mind different batches may have small differences, but nothing too significant. The Chinese one I would exclude becuase the price is just too low to be true. That one is fake for sure in my opinion. Chinese even manufacture counterfeit versions of the Sony CX20084, which is very requested (a genuine one costs about 50-70euros, Chinese sells them 20$/5pcs). Note that the price on TI's website is $11.16/pcs (in 1000pcs+ quantity), but it is out of stock. So the ebay seller does actually pretty good in terms of price, if the ICs are genuine. There are 2 aspects that I would worry about in regard to a potential counterfeit chip: 1. The specs are not as good as the original, including maximum ratings; 2. The longevity of the part itself; What guarantee do you have that the IC will still work 2 years from now ? My personal opinion in regardard to this upgrade is this: I would search for a different part that is newer and has similar specs. TI website can help with that. At least that is what I would do personally, you may have a different perspective on that. Things to consider if searching for a different part are: - input noise (must be specified over the same bandwitdth, otherwise comparisons won't be relevant); - gain bandwidth product (it should be same as original part or higher); - THD; prefferably to be as constant as possible over the audio bandwitdth, as at the specified 1kHz is will always be the lowest; - slew rate (if it's meant to drive headphones directly, this parameter is important, otherwise can be ignored); - power supply rejection ratio (should be same as original or higher); In terms of sound, the input noise, THD and slew rate are the only ones that will influence how it sounds. The rest (bandwidth and gain) will be determined by the feedback network present in the particular device.
Even from packages it can be difficult to tell. At work we had some Maxim parts that wouldn't work in one of our designs. The package and writing were noticeably different from the ones we had used in the prototypes. I sent some to Maxim fully expecting them to be fakes. To my surprise Maxim said they had tested them and couldn't find any problems with them ! I suspect they were made by Maxim but in a different fab using a different process. At least the difference in markings meant we could tell which parts we needed to fit to make working boards.
frankly, I did not expect to see any differences between $1.50 and $30 chips: it is easier to fake the shell than 'laser-trimmed input circuitry'. Did not expect to see any rust either. I am pretty sure they work as op-amps, but I am not going to waste any more time on them Exactly! I have no idea how to measure, and in this case even I wouldn't trust my own hearing
