shared reference silicon wafer for comparisons

rjlittlefield · Post by **rjlittlefield** » Sun Feb 16, 2020 4:14 pm

Smokedaddy wrote:What was the final story on this slide (EPI illumination)? I know a few purchased it here as I did.

https://www.bw-optik.de/restposten/test ... tpraeparat

I don't know as there ever was a "final story". There was a long discussion at https://www.photomacrography.net/forum/ ... 881#204881 and in the surrounding thread.

--Rik

Smokedaddy · Post by **Smokedaddy** » Sun Feb 16, 2020 4:44 pm

This one says the holes are 170nm.

https://www.bw-optik.de/restposten/test ... skop?c=347

-JW:

Macro_Cosmos · Post by **Macro_Cosmos** » Wed Feb 19, 2020 10:43 pm

https://live.staticflickr.com/65535/495 ... f32e_o.jpg

Here's the second chip and full resolution.
Shooting info: https://flic.kr/p/2ivqxwg

Lots of pixel-level detail. Seems like the lens is struggling to resolve them.

mawyatt · Post by **mawyatt** » Thu Feb 20, 2020 6:22 am

Really nice chip image, and beautiful chip subject with great colors!! Most of the wafers/chips I've purchased are kinda mundane compared to this chip! Are these specific chips still available?

The chip size, looks to be ~2.6mm by 2.3mm if those are 100um bond pads.

Ever consider doing a stack of this chip at an angle, around 25~40 degrees usually renders a nice 3D perspective with a 1~1.5mm depth. This should produce a stunning result with this chip subject I would think.

Best,

Lou Jost · Post by **Lou Jost** » Thu Feb 20, 2020 6:45 am

Yes, beautiful picture. But not a good test wafer. Too small for low magnifications, and too colorful. No test pattern either? Test patterns would be really useful.

Macro_Cosmos · Post by **Macro_Cosmos** » Thu Feb 20, 2020 7:55 am

mawyatt wrote:Really nice chip image, and beautiful chip subject with great colors!! Most of the wafers/chips I've purchased are kinda mundane compared to this chip! Are these specific chips still available?

The chip size, looks to be ~2.6mm by 2.3mm if those are 100um bond pads.

Ever consider doing a stack of this chip at an angle, around 25~40 degrees usually renders a nice 3D perspective with a 1~1.5mm depth. This should produce a stunning result with this chip subject I would think.

Best,

I have 2 trays of these, a total of about 200 cut chips. I'm currently doing a mundane looking one at 45deg, which as you've stated, looks quite dimensional.

Lou Jost wrote:Yes, beautiful picture. But not a good test wafer. Too small for low magnifications, and too colorful. No test pattern either? Test patterns would be really useful.

Yeah it won't work for low mag. I could have done a 20x stack even. I don't think it's possible to have a chip that covers say 1x all the way to 20x or 50x unless we have something custom made.

mawyatt · Post by **mawyatt** » Thu Feb 20, 2020 8:15 am

Lou,

The test patterns are usually not on the chip but in the channels between chips on the wafer. These are used during wafer fab to verify the process parameters and sometimes even have a tiny circuit to measure performance (ring oscillator for example). So diced single chips would not be a good candidate for a test subject, but full or partial wafers would.

Awhile back I actually severed a wafer at Beatsy request and we exchanged sections from that severing. This might be a good option for the various lens testing folks to pick a specific wafer and sever it and send out the sections. This somewhat assures a common test subject. There are firms that will dice up wafers for a fee, and probably custom sever a wafter.

A few on-line videos show how to sever wafers with just a very sharp tool (I used a new Xacto type blade) and a towel. The crystal orientation of the wafer is usually best to sever along and this usually aligns with the chip lanes. Sometimes a thin sliver will crack off as this image from my 1st attempt, the other section came out nicely tho. You can see the various test patterns in the lanes.

Best,

Large image, be sure to download. Note the lanes are full of various features and text which might prove useful as a test pattern.

http://img.gg/10oj9Jm

chris_ma · Post by **chris_ma** » Thu Feb 20, 2020 8:23 am

mawyatt wrote:Really nice chip image, and beautiful chip subject with great colors!! Most of the wafers/chips I've purchased are kinda mundane compared to this chip! Are these specific chips still available?

The chip size, looks to be ~2.6mm by 2.3mm if those are 100um bond pads.

i think it's this chip:
https://www.ebay.de/itm/100pcs-Toshiba- ... 3743175642

description says 2.5x2.5m, so pretty much spot on with your estimate.
free shipping world wide.

what's more, I think I have some of those... somewhere - gotta go digging :)

Lou Jost · Post by **Lou Jost** » Thu Feb 20, 2020 9:03 am

Macro_cosmos, I also have a 40-50 of those, or something similar. They are useful but much better would be a complete wafer with chips all over.

Mike, yes, that makes sense. I am not thinking to use single chips but wafers, which will work perfectly with both high and low magnifications, and which could have the test patterns.

I imagine cutting wafers would be easy work for a cheap glass cutting tool. But we wouldn't want to cut a wafer too finely, because we might want to use it at low m too.

Macro_Cosmos · Post by **Macro_Cosmos** » Thu Feb 20, 2020 9:17 am

chris_ma wrote:
mawyatt wrote:Really nice chip image, and beautiful chip subject with great colors!! Most of the wafers/chips I've purchased are kinda mundane compared to this chip! Are these specific chips still available?

The chip size, looks to be ~2.6mm by 2.3mm if those are 100um bond pads.
i think it's this chip:
https://www.ebay.de/itm/100pcs-Toshiba- ... 3743175642

description says 2.5x2.5m, so pretty much spot on with your estimate.
free shipping world wide.

what's more, I think I have some of those... somewhere - gotta go digging

Yeah, exactly the same, the one you linked has the information on the tray, thanks for that! I paid $1 each for my trays, these are so cheap on Chinese websites.

Lou Jost wrote:Macro_cosmos, I also have a 40-50 of those, or something similar. They are useful but much better would be a complete wafer with chips all over.

My initial idea was to 3D print a tray tailored to fullframe sensors and embed these chips in the tray, allowing me to place them in the centre, border (APS-C) then corners. I ditched the idea after getting photomasks.

It's totally possible for all of us to come together and design a mask that has all the information we want. I know someone who can make these masks down to a resolution of 0.1um. 1um would be a lot cheaper.

Some stuff include
- dotted pattern and grid for distortion
- ruler for magnification
- Slanted edge for MTF
- Various diffraction gratings for resolution/sharpness

I'm just not sure how much it will cost. It depends on how many are interested. With these things, it's always quantity. More units means considerably cheaper and faster lead time.

Lou Jost · Post by **Lou Jost** » Thu Feb 20, 2020 9:30 am

That's an interesting thought, though I can't imagine it would be cheap at the small quantities we are talking about. However, it might be possible for us to generate our own wafers, using the low-res cheap masks and our photolithography lenses....

mawyatt · Post by **mawyatt** » Thu Feb 20, 2020 9:51 am

Lou Jost wrote:Macro_cosmos, I also have a 40-50 of those, or something similar. They are useful but much better would be a complete wafer with chips all over.

Mike, yes, that makes sense. I am not thinking to use single chips but wafers, which will work perfectly with both high and low magnifications, and which could have the test patterns.

I imagine cutting wafers would be easy work for a cheap glass cutting tool. But we wouldn't want to cut a wafer too finely, because we might want to use it at low m too.

I would think that a 50mm square "wafer-let" would be fine, enough to cover a 35mm FF sensor at 1X.

Think the on-line stuff about breaking up the wafers showed just a single sharp tool pressed, not scribed, on the wafer positioned on a soft towel. The towel allows the wafer to bend slightly and eventually cracking along the crystal orientation hopefully.

Believe it or not silicon wafers when thinned are very flexible, during a demo back in ~1980 one was wrapped around a Coke can and you could see thru it I recall

Best,

mawyatt · Post by **mawyatt** » Thu Feb 20, 2020 10:08 am

chris_ma wrote:
mawyatt wrote:Really nice chip image, and beautiful chip subject with great colors!! Most of the wafers/chips I've purchased are kinda mundane compared to this chip! Are these specific chips still available?

The chip size, looks to be ~2.6mm by 2.3mm if those are 100um bond pads.
i think it's this chip:
https://www.ebay.de/itm/100pcs-Toshiba- ... 3743175642

description says 2.5x2.5m, so pretty much spot on with your estimate.
free shipping world wide.

what's more, I think I have some of those... somewhere - gotta go digging

Chris,

Thanks.

Best,

Adalbert · Post by **Adalbert** » Thu Feb 20, 2020 11:02 am

Hello everybody,
I have already taken and shown some photographs of the chip from BW-Optik:
2x:

5x:

10x:

20x:

50x:

80x:

BR, ADi

chris_ma · Post by **chris_ma** » Thu Feb 20, 2020 11:47 am

Adalbert wrote:Hello everybody,
I have already taken and shown some photographs of the chip from BW-Optik

hi,
that looks like a good target indeed, didn't realize this when looking at the product page, but your samples look great, and I like that with 75x25mm it covers full frame at 1x.

going to order one of those as well :)
https://www.bw-optik.de/restposten/test ... tpraeparat

thanks for the repeated hint
chris