Powers of 10- doxxed!
Moderators: rjlittlefield, ChrisR, Chris S., Pau
-
- Posts: 40
- Joined: Tue Apr 25, 2023 4:38 pm
- Location: Cleveland, OH
Powers of 10- doxxed!
This series examines a small chip from National Semiconductor, a LM3402, fabricated in 2004. I tried a few new techniques with this series. Here's the 1:1 reproduction:
The chip is barely visible. One new thing I tried is the mounting method- taking a cue from the diatomists, I glued the chip to a coverslip with nail polish rather than mount it to the slide. This helps ensure the face of the chip is flat and perpendicular to the optical axis of the microscope. My first attempt was a little messy, but it worked well, so thanks, guys!
The trade-off is that most of my (Zeiss) objectives can't be used as they are designed to be used w/o a coverslip. So, the other new thing was to try and get reasonable images using Nikon and Olympus objectives on my Zeiss microscope. I did notice a decrease in overall contrast due to the coverslip, but I can live with it.
Here's the next image using an Olympus SPlan 4x/0.16 objective:
And then using a Nikon 20x/0.75 Fluar: If you look closely, there a small line of text beneath the yellow squares, and that's what I am going to keep magnifying.
And then using an Olympus SPlan 40X/0.95 w/ correction collar: Huh... it's a name! Let's zoom in a bit more, using the Leica 100X/1.47 lens: This is a real person, raising interesting questions about privacy- I blocked out the last name because this is a public forum.... When I found this, I did an online search for the person, and he is still alive and living in the US. I believe he is retired. I haven't tried to contact him and wonder if I should.
Does he wonder if anyone discovered this? Was he in fact responsible for putting his own name on the design? Who else knows about this? So many questions!
The chip is barely visible. One new thing I tried is the mounting method- taking a cue from the diatomists, I glued the chip to a coverslip with nail polish rather than mount it to the slide. This helps ensure the face of the chip is flat and perpendicular to the optical axis of the microscope. My first attempt was a little messy, but it worked well, so thanks, guys!
The trade-off is that most of my (Zeiss) objectives can't be used as they are designed to be used w/o a coverslip. So, the other new thing was to try and get reasonable images using Nikon and Olympus objectives on my Zeiss microscope. I did notice a decrease in overall contrast due to the coverslip, but I can live with it.
Here's the next image using an Olympus SPlan 4x/0.16 objective:
And then using a Nikon 20x/0.75 Fluar: If you look closely, there a small line of text beneath the yellow squares, and that's what I am going to keep magnifying.
And then using an Olympus SPlan 40X/0.95 w/ correction collar: Huh... it's a name! Let's zoom in a bit more, using the Leica 100X/1.47 lens: This is a real person, raising interesting questions about privacy- I blocked out the last name because this is a public forum.... When I found this, I did an online search for the person, and he is still alive and living in the US. I believe he is retired. I haven't tried to contact him and wonder if I should.
Does he wonder if anyone discovered this? Was he in fact responsible for putting his own name on the design? Who else knows about this? So many questions!
Last edited by Andy Resnick on Tue May 30, 2023 8:02 pm, edited 1 time in total.
Re: Powers of 10- doxxed!
Fascinating series. I don't think there can be any expectation of privacy since Perry has deliberately signed his work and offered it to the public.
- rjlittlefield
- Site Admin
- Posts: 23563
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Re: Powers of 10- doxxed!
Nice imaging!
I'm confused about one aspect of the device. I looked up the datasheet for LM3402 and found that it's a "constant current buck LED driver", whose pinout has 8 connections including power and ground. But on the face of this chip I'm seeing only 5 obvious contact pads.
Do you know how to resolve the discrepancy?
--Rik
I'm confused about one aspect of the device. I looked up the datasheet for LM3402 and found that it's a "constant current buck LED driver", whose pinout has 8 connections including power and ground. But on the face of this chip I'm seeing only 5 obvious contact pads.
Do you know how to resolve the discrepancy?
--Rik
-
- Posts: 40
- Joined: Tue Apr 25, 2023 4:38 pm
- Location: Cleveland, OH
Re: Powers of 10- doxxed!
Ah... nope. Could this chip be an LM3492? That numeral (0 or 9) is difficult to distinguish on the marking.rjlittlefield wrote: ↑Thu May 25, 2023 1:41 pmNice imaging!
I'm confused about one aspect of the device. I looked up the datasheet for LM3402 and found that it's a "constant current buck LED driver", whose pinout has 8 connections including power and ground. But on the face of this chip I'm seeing only 5 obvious contact pads.
Do you know how to resolve the discrepancy?
--Rik
-
- Posts: 40
- Joined: Tue Apr 25, 2023 4:38 pm
- Location: Cleveland, OH
Re: Powers of 10- doxxed!
Yes it is! I bet this person would be thrilled to know that someone has noticed his handiwork.
- rjlittlefield
- Site Admin
- Posts: 23563
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Re: Powers of 10- doxxed!
Hhmm... I see tiny lettering in the upper right corner, but I can't make it out either.Andy Resnick wrote: ↑Fri May 26, 2023 11:22 amAh... nope. Could this chip be an LM3492? That numeral (0 or 9) is difficult to distinguish on the marking.
LM3492 does not work. That has even more pins than LM3402. Likewise for all other numbers that I checked.
Does the lettering become more clear if viewed at highest magnification?
--Rik
Re: Powers of 10- doxxed!
If image below is the number in question I read it as: LM34822
.
.
- Attachments
-
Zeiss Standard WL & Wild M8
Olympus E-p2 (Micro Four Thirds Camera)
Olympus E-p2 (Micro Four Thirds Camera)
- MarkSturtevant
- Posts: 1946
- Joined: Sat Nov 21, 2015 6:52 pm
- Location: Michigan, U.S.A.
- Contact:
Re: Powers of 10- doxxed!
Apparently leaving Easter Eggs on computer chips is a thing that is done: https://techunwrapped.com/easter-eggs-o ... rises-too/
So from now on, folks who image these devices could maybe look for more and share them with us. It would be fun.
So from now on, folks who image these devices could maybe look for more and share them with us. It would be fun.
Mark Sturtevant
Dept. of Still Waters
Dept. of Still Waters
- rjlittlefield
- Site Admin
- Posts: 23563
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Re: Powers of 10- doxxed!
Yes, that is one of the possibilities that I tried. But search on LM34822 does not find that part number. Likewise for LM3482. There exist LM348 and LM34922, but again both of those have more contacts than I can see on the chip pictured in this thread.
--Rik
Re: Powers of 10- doxxed!
Perhaps LM94022 which is a temperature sensor in a 5-pin package?
- rjlittlefield
- Site Admin
- Posts: 23563
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Re: Powers of 10- doxxed!
That looks promising! Date is about right also. https://www.ti.com/product/LM94022 lists the white paper at https://www.ti.com/lit/pdf/snia011 as being May 2005.
The fellow whose name is on the chip would probably be amused to see how much we're thrashing around with it.
--Rik
-
- Posts: 40
- Joined: Tue Apr 25, 2023 4:38 pm
- Location: Cleveland, OH
Re: Powers of 10- doxxed!
Just read through, here's a more zoomed-in crop (the best I have on hand right now), hope it's good enough!
I'll remember to include images of the part number in the future - rjlittlefield
- Site Admin
- Posts: 23563
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Re: Powers of 10- doxxed!
LM94022 works for me. For the "9", I see a full top loop and a rounded end on the tail, and for the "0" I see two straight sides.
So then that opens the question of what the various parts of this analog chip do. Those rectangular arrays of block shapes are now extra intriguing.
--Rik
So then that opens the question of what the various parts of this analog chip do. Those rectangular arrays of block shapes are now extra intriguing.
--Rik