Electronics > Projects, Designs, and Technical Stuff

Identifying a decapped chip marked NE5534 - fake?

(1/2) > >>

ckocagil:
Over a year ago I had, in my infinite wisdom, bought 10 pieces of DIP chips from Aliexpress that boldly claim to be TI NE5534 all with the same date/batch code. So I did what any curious and sensible person would do: measured the slew rate boiled it in acid and extracted the die.

I took a terrible picture using my phone's camera and a salvaged lens from my parts box. Unfortunately I was unable to find a die image of a real 5534. Zeptobars has an NE5532 image, but that doesn't help me much, does it?

So far I have 3 hints that point to this chip being a fake:
- Same date/batch code on all chips
- The die is absolutely tiny. Longest dimension is at most a milimeter. For comparison, Zeptobars says the TI NE5532 has a die size over 2mm x 2mm.
- The silicon is a bit too simplistic..? I would expect a more complicated circuit, but take this with a grain of salt, I have very little clue.

Any ideas?

Wimberleytech:
Why dont you send one to Noopy and have him do a proper teardown.

Looking at the datasheet, I see a 100pf, 12pf, 7pf, and 40pf capacitors.  Those will be large and noticeable.  I dont see good candidates on your die photo.  Moreover, I found this on the interweb.  Of course, I cannot vouch for the accuracy.

magic:

--- Quote from: ckocagil on March 01, 2020, 07:48:26 pm ---So far I have 3 hints that point to this chip being a fake:
- Same date/batch code on all chips

--- End quote ---
:-DD
This normally is not a problem, but you are right. When it comes to too-cheap-to-be-true jellybean chips from AliExpress, the only semi-legit ones are batches of mixed-date, mixed-manufacturer recycled ICs.


--- Quote from: ckocagil on March 01, 2020, 07:48:26 pm ---- The die is absolutely tiny. Longest dimension is at most a milimeter. For comparison, Zeptobars says the TI NE5532 has a die size over 2mm x 2mm.
- The silicon is a bit too simplistic..? I would expect a more complicated circuit, but take this with a grain of salt, I have very little clue.

--- End quote ---
Yep, I have opened some legit 5534s from Philips, TI and JRC and the die size is about 1x2mm in each case. The layout is quite different than half of 5532, by the way, but all the same components are there if you look for them. Minor differences in capacitor area and biasing resistors appear to be responsible for unity gain stability of the 5532.

Finally, an immediate giveaway that this is fake is the fact that it's a dual opamp die with only one half connected to the package.

edit
I can confirm that the above picture is a 5534. The four transistors surrounded by capacitors are the 1st and 2nd stage pairs and this block is quite characteristic for 5534/5532. The whole thing looks like 5534.

ckocagil:
Beautiful. With that reference image, the lack of capacitors, and your observation that this is in fact a dual op-amp (I thought it was a bit asymmetric, but that was just bad lighting) we can confirm that this is a fake chip.

Thank you folks.

magic:
I may add that every legit 5532/5534 has diodes between IN+ and IN- which can be detected with a DMM. This chip probably is a random Chinese LM358 or RC4558 with custom package and a bit of creative wiring to make it work as a single and doesn't have such diodes.

Also, those dual-opamp-in-a-single-package fakes have no connection whatsoever to pins 1,5,8. Another easy test with a DMM.

Navigation

[0] Message Index

[#] Next page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod