Fake Chips

[granddad]

Apologies if ETO has had similar threads ..But as supplies of 'legacy' ( and newer I guess ) ICs dwindle or non existant the hobbyist is left with unknown suppliers / sources , members may find this video useful ..

 

[Visitor]

I had about 50 boards assembled in China that had two I2C chips – one a digipot and the other a variable frequency oscillator chip. When I received the boards, I tested one of them. It worked nothing like the prototype I had assembled, using exactly the same circuit board.

I wrote some test code that cycled through the frequency range of interest and stepped through the levels. Long story short, the I2C oscillator chip was responding to everything on the I2C bus no matter what address it was being sent to. I replaced the chip with s known-good one, and it worked as expected. The oscillator chips were fake. They looked virtually the same as the real deal.

My assembler had been sold fake chips. I returned the boards (a story in otself), and the assembler changed all the chips. I am certain they were as much a victim as I was.

 

[sagor1]

Really good video. Lots of fakes out there, but difficult to spot unless you can test them out. Even then, recycled chips may have a limited life span.

 

[Beau Schwabe]

I ordered 100 jelly bean LM741's from eBay for a build workshop I was presenting in Las Vegas one year, and NONE of the LM741's worked. For the conference I ended up re-ordering from Digikey and the LM741's from Digikey all worked fine. Good thing I double checked the build ahead of time before the conference.

I held on to the LM741's from eBay and after further testing I discovered they were totally miss-marked. They behaved like LM1458's which is a DUAL Op-Amp instead of the SINGLE LM741 Op-Amp.

 

[gophert]

They behaved like LM1458's which is a DUAL Op-Amp instead of the SINGLE LM741 Op-Amp.

The seller gives you dual op amps for the price of singles and you still complain!

 

[unclejed613]

back in the 1980's, there were a lot of "grey market" chips. technically these weren't counterfeit chips, but they were fab rejects because they failed testing. so they were the real thing, but they were out of spec. some enterprising people went around to the fabs and bought the rejects for less than cost, and unbeknownst to the fab, resold them at full price... it didn't dawn on the fabs that these rejects might show up later on their doorstep(s) as batches of parts rejected by major equipment manufacturers... the lot and date codes matched the codes on the rejects sold to the shady resellers, and at that time a lot of fabs stopped selling reject parts, and began recycling the materials. i forget what magazine the article was in, but it was a far-east industry magazine in 1985 or so...

 

[gophert]

...so now we have tracking and tracing through multiple levels of the supply chain, "authorized distributors" and ISO9000/ISO9001 quality management systems and non-compliance auditing systems.

None of it is perfect but I couldn't imagine launching a medical device into the market (or installing a pacemaker in someone's chest) in the wild west conditions of the 1980s and before.

 

[unclejed613]

maybe not but you still have to watch out for counterfeited audio output transistors and large electrolytic capacitors too... i had been buying parts from a supplier (who at one time in the 1980s and 1990s had been a Toshiba and Sanken authorized distributor), and i don't know what happened but in the early 2000's they must have lost their contracts and began selling parts that no longer were reliable... at that time i was unaware that there was even such a thing as counterfeit parts... i was thinking "grey market... maybe?"... but these failures weren't something slightly out of spec, but catastrophic... transistors spitting flames when the original ones just shorted gracefully with no external marks...