ZeptoBars - RSS feed http://zeptobars.com/en/ Microelectronics. Die-shots. Artificial intelligence. Lasers. en-us Tue, 10 Jun 2006 04:00:00 GMT Sun, 25 Feb 18 06:59:12 +0300 webmaster@zeptobars.com 120 10 <![CDATA[MOC3061 - Zero-cross optoisolated triac driver : weekend die-shot]]> http://zeptobars.com/en/read/MOC3061-Zero-cross-optoisolated-triac-driver-phototriac Die size 1378x1309 µm.

Sun, 10 Dec 17 04:22:04 +0300
<![CDATA[Noname LED blinker : weekend die-shot]]> http://zeptobars.com/en/read/noname-LED-blinker
Chip internals are more complicated than one would imagine - there are 5 LED channels, only 2 of which are used here. Layout resembles other blinkers we've seen in the past (flicker, RGB). It's fascinating that there is a company somewhere specialized in making LED blinkers of various kind and doing just fine.

Die size 587x479 µm.

Mon, 27 Nov 17 02:44:19 +0300
<![CDATA[UMC UA6538 - Dendy (NES compatible) PPU : weekend die-shot]]> http://zeptobars.com/en/read/UMC-UA6538-Dendy-NES-Nintendo-PPU nesdev.com.

Die size 4578x4069 µm.

Max resolution, if you really want to (288Mb).

After etching metal layers:

Max resolution, if you really want to (200Mb).]]>
Tue, 31 Oct 17 08:05:46 +0300
<![CDATA[Philips NE5532A : weekend die-shot]]> http://zeptobars.com/en/read/Phillips-NE5532A-opamp-audio genuine Ti NE5532 than to fake.

Tue, 03 Oct 17 11:36:12 +0300
<![CDATA[Ti NE5532 - real vs fake : weekend die-shot]]> http://zeptobars.com/en/read/Ti-NE5532-real-vs-fake-opamp
Genuine Ti NE5532 has die size of 2103x2154µm.

Fake Ti NE5532 is much smaller and have much simpler design. Bare minimum to operate as opamp:

SnipTheCat at EEVBlog forum noticed that these have much slower slew rate vs datasheet.

This IC was from one of these "Ti NE5532 100pcs for 4$" ebay lots. On the back side it was marked as "Tdsemic", which probably means it was manufactured/packaged at 拓电半导体(香港)有限公司 (TD SEMICONDUCTOR LIMITED.).
Tue, 03 Oct 17 10:21:50 +0300
<![CDATA[ST TSC1031IDT 70V high-side current sense amplifier : weekend die-shot]]> http://zeptobars.com/en/read/STM-TSC1031IDT-high-voltage-high-side-current-sense-amplifier STMicroelectronics TSC1031 allows to amplify differential signal with high common-mode voltage (up to 70V) while having low VCC (2.7 to 5.5V) which is very useful for high-side current sensing.

Die size 422x437 µm.

Sun, 17 Sep 17 17:22:30 +0300
<![CDATA[Ti TLC5540I 8bit 40MSPS semiflash ADC : weekend die-shot]]> http://zeptobars.com/en/read/Ti-TLC5540I-8bit-40MSPS-ADC-semiflash Ti TLC5540I is a CMOS semi-flash 8bit 40MSPS ADC. Die is much smaller than simple 8bit flash ADC.

Die size 1699x1801 µm, 1µm minimum half-pitch.

Core at higher resolution:

Closer look at resistor matrix. Not much is visible through 2 metals:

Mon, 24 Jul 17 01:25:47 +0300
<![CDATA[Fairchild LM555CMX : weekend die-shot]]> http://zeptobars.com/en/read/Fairchild-LM555CMX-555

Sun, 23 Jul 17 20:31:48 +0300
<![CDATA[Fairchild BCV27 - npn darlington BJT : weekend die-shot]]> http://zeptobars.com/en/read/Fairchild-BCV27-npn-darlington-bjt Fairchild BCV27 - is a tiny SOT-23 Darlington npn BJT with hFE of 20'000.
Die size 612x611 µm.

Sat, 08 Jul 17 06:44:49 +0300
<![CDATA[Ti NE555 - real vs fake : weekend die-shot]]> http://zeptobars.com/en/read/Ti-555-NE555-real-vs-face-china-chinese fake 80$ "audiophile-grade" opamp. But why would someone fake one of the cheapest microchips in the world?

With volumes high enough that's still decent money for a small company. It could also be an economic warfare: each fake chip sold means 7 cents less in US GDP and Ti sales.

Genuine Ti NE555D. This one is directly from distributor. You can see several mentions of Texas Instruments on the die.
I really like designs with pads in the middle of the die, it looks so elegant and efficient (when it's not reducing yield).
Die size 1034x762 µm.

First fake 555 timer marked as Ti NE555 from ebay:

Different fake 555 design marked as Ti NE555 found on ebay:
Die size 747x762 µm.

Update: microcolonel was first to realize that this G1083 die is exactly the same, as the one we've decapped as HA17555. That means our HA17555 was also fake. These fakes are getting convoluted.

Obviously, basic functionality is the same, but parameters limits and their tolerances are unknown.
If you're not reaching for the limits (like max VCC or precision) and have some luck - that might not cause issues.

Looking at these designs one can see that routing in Chinese designs is very "intense": they are squeezing every fraction of a cent in die manufacturing cost.

That not necessary means that Chinese designs are superior : in genuine analog circuits precision, tolerances and noise are much more important than die size. But there is no doubt that these Chinese designs have reached lowest possible manufacturing cost for a 555. If only have they marked it "CN555" with proper data-sheet - it would have been totally legitimate and fair product. ]]>
Mon, 03 Jul 17 05:55:50 +0300