InP Test Chip

[mawyatt]

Here's a tiny InP Test Chip. This image was created with the Mitutoyo 20X and Sigma 200mm LSA as a tube lens. Stack was about 330 images at 2.5 microns.

I had to file the head down of the smallest pin (0.5mm dia) to mount the chip.

In the image note the brown large squares, these are through vias to the chip backside, also the probe scribe marks on the end pads. Another consumption of my time, besides posting here, is indicated in the ULC.

Hi rez version.

http://img.gg/eTr9uBe

Best & Happy Holidays,

Mike

[siliconGary]

That's a great image! The thick top metal layer gold? (Or is it Cu) layer on the surface really jumps out. Do I see your initials in the bottom right corner of the chip? :-)

[mawyatt]

Thanks. Yes it's gold on the top, and yes those are my initials

I probably should edit out the black company coverup in the LLC, it does look bad!

Best,

Mike

[mawyatt]

Here's the black space edited out.



And the hi Rez version.

http://img.gg/v2qM98i

[siliconGary]

As a old process guy I will admit that I used to be a bit jealous of you cool designers getting to put your initials onto the chips (Even though probably only a few dozen people ever saw them normally.)

I think high frequency chips alway look interesting, the symmetry and precise layout is so tidy (Never seen an ugly RF chip). I can see what I think is an LC match on the left, would love to know what the chip does? (I understand if you do not want to say). I am surprised to see that (in 2016) electronic circuits are still being made in on InP. I thought it was only being used to make optical components these days (Tunable lasers and Mach-Zehnder Interferometers etc.)

[mawyatt]

Hi Gary,

InP is actually one of the fastest semiconductor technologies available today. The hetro-junction bipolar transistor (HBT) devices in this process have an unity current gain (ft) of 275GHz and a unity power gain of 600GHz (fmax).

The circuit is fully differential input and output. On the left are two mirror image inductors (small spiral circles) which help tune out the input capacitance.

This is an experimental circuit that emulates a square-law characteristic (Y=K*X^2) from DC to over 30GHz, it's use is intended to detect broadband Electromagnetic Energy over that range.

Getting your initials on the chip can be a good or bad thing, there's no hiding or blaming someone else if the chip doesn't work

One of the first folks to see this image from where is used to work, noted "who put the stuff in the upper left corner", I couldn't hide from that either

Best & Happy Holidays,

Mike

[Smokedaddy]

Sweet image Mike. Out of my league here but would this more than likely be found in a cell phone? Ya never know where to look for 'stuff' like this. <g>

-JW:

[mawyatt]

Thanks JW,

This could be used in cell phones, but the concept wouldn't need to be implemented in Indium Phosphide (InP) since phones work around 2GHz, Silicon would work fine at the cell phone or WiFi (2.4 & 5GHz) frequencies.

Certain types of RF, Microwave and Millimeter Wave (RF/MW/MMW) instrumentation and specialized applications are the intended uses.

BTW your art work is absolutely stunning!!

Best & Happy Holidays,

Mike

[Olympusman]

I just love your oblique images of ICs. Another winner!
Mike

[mawyatt]

Thanks Mike.

I think they show some depth to an otherwise planar subject.

These tiny ones are a real bear though! I have a lot of appreciation for those folks doing the tiny insects....I have enough problems with these chips!!

Best,

Mike

[Smokedaddy]

Thanks JW,

BTW your art work is absolutely stunning!!

Mike

... thanks for the kind words.
-JW:

[mawyatt]

Here's another version of the circuit. This was shot using D850 EFCS and continuous lighting from 5 studio strobes modeling lights (2 LED and 3 incandescent) and 4 Ikea LED lamps.

Processed with Zerene Pmap and Dmap.

Hi Rez version.

http://img.gg/E9He82j

Best,

Mike