DACs - die pictures

[Miyuki]

I have a new DAC for you. It´s a DAC 32 built by KWH (Keramische Werke Hermsdorf).




....

Another gorgeous piece of tech 

And intercomunist cooperation on chip supply. What a time back then

[Noopy]

Thanks to some nice guys I now can tell you that the current sink circuits are B724X manufactured by KWH (Keramische Werke Hermsdorf) and HFO (Halbleiterwerke Frankfurt Oder):

https://richis-lab.de/DAC03.htm#Update

Instead of designing and manufacturing nearly perfect transistors like in the ICL8018 the B724X contains circuits to compensate the weaknesses. The current sink transistors are darlington types with two inputs. Between the current sinks there are differential amplifiers that compare the neighbouring channels to compensate Vbe-drift.

[Noopy]

Update DAC709:

https://richis-lab.de/DAC01.htm#Schaltplan

I got an IEEE article which describes the DAC709! Now I understand the connection of the current sinks and I was able to mark the different parts on the die.





 

[Noopy]

Hi all!


Today I have a newer DAC and some pin compatible relatives: AD7226, MX7226 and TLC7226

They provide four 8Bit-DACs and the structures are still big enough to study the blocks:




It´s a R2R-design without current sinks:




You can also take a look at the opamps:




The three relatives are quite similar but although different.
...is that proper english? 


More pictures here:

https://www.richis-lab.de/DAC04.htm


 

[mawyatt]

Beautiful chip images. Great detail and lighting, well done

What camera setup and lens are you using? I ask because I starting doing images of the chips we've developed since 2004 and evolved to a microscope lens (usually Mitutoyo) with a tube lens assembly with a DSLR camera (Nikon) on a commercial focus rail (now evolved to fully custom rail and controllers).

Best,

Mike

[Noopy]

Beautiful chip images. Great detail and lighting, well done

What camera setup and lens are you using? I ask because I starting doing images of the chips we've developed since 2004 and evolved to a microscope lens (usually Mitutoyo) with a tube lens assembly with a DSLR camera (Nikon) on a commercial focus rail (now evolved to fully custom rail and controllers).

Best,

Mike

Thanks!
I have posted a HowTo here:
https://richis-lab.de/Howto.htm
I use a Canon 60D with a twisted 10-22mm and distance rings between camera and lens.
The trick is to place the light behind the die so it gets reflected by the lens and hits the die like it does in the better microscopes.
And of course I have spent a huge amount of time and Gigabytes until I got this kind of quality.

You take pictures with a combination of a microscope lens and a DSLR? Interesting... How does that look like? Do you get better pictures than I do?

Best regards,

Richard

[mawyatt]

Your images are excellent, beautifully done and illuminated. I use a different technique and equipment to achieve a more 3D look. Illumination is from multiple flashes highly diffused.

You can see some older work that I can show here on Nikon Image Space, download for the higher resolutions.

http://img.gg/taIZ99M

I use a lens configuration based upon a microscope objective, infinity corrected, with a 200mm FL tube lens which attaches to a DSLR (Nikon D500, D800E or D850, and soon a Z7). This mounts to a custom developed focus stacking rig I just posted a listed here on the development of this setup. This allows extreme macro levels of resolution, but coverage of a larger field of view with a stacking and stitching technique. Images of 29,000 by 22,000 pixels have been achieved with this setup, soon this will surpass 40,000 by 30,000 pixels which maybe I'll be able to show in a few years.

This was all started way back in ~2004 to image the chips we had and were developing, many of the images I can show are from older designs that I was involved with and many are or contain patented circuits.

I've attached some old images of the lenses and one of the vertical setups, we use both V and H setups and also an lab grade Thorlabs setup (last image). Things are more evolved now, but were the images I have on hand.

BTW we've now developed techniques to allow sub-micron stacking, and lately utilizing Piezo Electric Device Stages for approaching under 50 nanometer levels.

Best

[tggzzz]

Other Burr Brown devices from a late 70s calendar can be seen here:
https://entertaininghacks.wordpress.com/2015/07/07/images-of-late-70s-burr-brown-thick-film-hybrid-ics/

[Noopy]

Your images are excellent, beautifully done and illuminated. I use a different technique and equipment to achieve a more 3D look. Illumination is from multiple flashes highly diffused.

Thanks!
Your pictures are also very nice! 

I use a lens configuration based upon a microscope objective, infinity corrected, with a 200mm FL tube lens which attaches to a DSLR (Nikon D500, D800E or D850, and soon a Z7). This mounts to a custom developed focus stacking rig I just posted a listed here on the development of this setup. This allows extreme macro levels of resolution, but coverage of a larger field of view with a stacking and stitching technique. Images of 29,000 by 22,000 pixels have been achieved with this setup, soon this will surpass 40,000 by 30,000 pixels which maybe I'll be able to show in a few years.

Unfortunatelly I´m not so familiar with optical systems. I´m just a hobby photographer who invested a lot of time in trial and error. 
Until you showed me your equipment I thought with a DSLR you can´t get much more details than i got.
Did you calculate that stackup before you built it or was there also some trial and error in this system?

This was all started way back in ~2004 to image the chips we had and were developing, many of the images I can show are from older designs that I was involved with and many are or contain patented circuits.

I would love to develope integrated circuits. 

I've attached some old images of the lenses and one of the vertical setups, we use both V and H setups and also an lab grade Thorlabs setup (last image). Things are more evolved now, but were the images I have on hand.

I still use a third hand... 
You setup is way more professional! Really cool! 


Thanks for all your information!


Best regards,

Richard

[mawyatt]

Other Burr Brown devices from a late 70s calendar can be seen here:
https://entertaininghacks.wordpress.com/2015/07/07/images-of-late-70s-burr-brown-thick-film-hybrid-ics/

This BB images are great, they were one of the best analog houses back in those days. Those hybrids they made were beautiful as you've shown and they performed beautifully also, used many of their parts back then.

What type of setup did you use?

Best,

Mike

[mawyatt]

Your images are excellent, beautifully done and illuminated. I use a different technique and equipment to achieve a more 3D look. Illumination is from multiple flashes highly diffused.

Thanks!
Your pictures are also very nice! 

I use a lens configuration based upon a microscope objective, infinity corrected, with a 200mm FL tube lens which attaches to a DSLR (Nikon D500, D800E or D850, and soon a Z7). This mounts to a custom developed focus stacking rig I just posted a listed here on the development of this setup. This allows extreme macro levels of resolution, but coverage of a larger field of view with a stacking and stitching technique. Images of 29,000 by 22,000 pixels have been achieved with this setup, soon this will surpass 40,000 by 30,000 pixels which maybe I'll be able to show in a few years.

Unfortunatelly I´m not so familiar with optical systems. I´m just a hobby photographer who invested a lot of time in trial and error. 
Until you showed me your equipment I thought with a DSLR you can´t get much more details than i got.
Did you calculate that stackup before you built it or was there also some trial and error in this system?

Over at Photomacrography.net you can find all you need to build these lenses, it's really not difficult. Many folks there are doing 10, 20 & 50X routinely, using microscope objectives and DSLR or mirrorless cameras.

This was all started way back in ~2004 to image the chips we had and were developing, many of the images I can show are from older designs that I was involved with and many are or contain patented circuits.

I would love to develope integrated circuits. 

I must admit now that I'm retired, I would have worked for almost free. Chip design, especially analog, is like art and very addicting

I've attached some old images of the lenses and one of the vertical setups, we use both V and H setups and also an lab grade Thorlabs setup (last image). Things are more evolved now, but were the images I have on hand.

I still use a third hand... 
You setup is way more professional! Really cool! 

Thanks, actually it's not as expensive as you might think if you assemble some of this yourself. The small vertical setup shown, this is based upon a Wemacro Vertical Stand which is ~$130 US. You can build custom lenses for under $100. Wemacro also sells an excellent focus rail with controller for $250 US, so for under $500 US you can build a system for vertical and horizontal use that will yield excellent results in hands as capable as you've shown with your excellent images.

Of course this chip imaging is addicting, and you will likely spend more over time as you want/need/develop new requirements. Obviously that's happened to me


Thanks for all your information!

You are quite welcome, and thanks for showing your wonderful chip images!! It's nice to find folks doing this with chips


Best regards,

Richard

Just noted that you have the LM399 reference image and details, I've developed a reference based upon this details over on LTspice@groups.io,

https://groups.io/g/LTspice/topic/73959871?p=Created,,,50,1,0,0::recentpostdate/sticky,,,50,2,0,73959871

DP Review.

https://www.dpreview.com/forums/thread/4486053


Best,

Mike

[Noopy]

I think I will have to do some shopping... 

Thanks again!

More coming soon... 

[tggzzz]

Other Burr Brown devices from a late 70s calendar can be seen here:
https://entertaininghacks.wordpress.com/2015/07/07/images-of-late-70s-burr-brown-thick-film-hybrid-ics/

This BB images are great, they were one of the best analog houses back in those days. Those hybrids they made were beautiful as you've shown and they performed beautifully also, used many of their parts back then.

What type of setup did you use?

I used my digital camera to take pictures of the calendar that I had placed on my floor

[Noopy]

I used my digital camera to take pictures of the calendar that I had placed on my floor

 

But definitely some nice hybrids. 

[magic]

Until you showed me your equipment I thought with a DSLR you can´t get much more details than i got.

One thing you could still try with standard equipment is stacking lenses. You only need an adapter ring which joins two infinity-focused lenses by their filter threads. The first (reversed) lens projects an image of the object towards infinity, the other lens is normally mounted on the camera and focuses this image from infinity to the sensor. Since both lenses are focused at infinity, this configuration could (maybe) offer better quality than simply reversing a single lens and forcing it to focus closer than it normally does.

IIRC, the diffraction limit of a reversed f/2 lens is be a bit over 1µm. That's good enough for common analog tech, if optical aberrations don't limit performance. In terms of microscope numeric aperture ratings, f/2 is about 0.25NA.

I have successfully employed this technique using a point and shoot camera and reversed f/2 webcam lens. My biggest problem was noise from the shitty compact sensor, flare from the shitty webcam lens and poor alignment of my shitty duct tape job   Better results should be possible with better gear.

[Noopy]

One thing you could still try with standard equipment is stacking lenses. You only need an adapter ring which joins two infinity-focused lenses by their filter threads. The first (reversed) lens projects an image of the object towards infinity, the other lens is normally mounted on the camera and focuses this image from infinity to the sensor. Since both lenses are focused at infinity, this configuration could (maybe) offer better quality than simply reversing a single lens and forcing it to focus closer than it normally does.

IIRC, the diffraction limit of a reversed f/2 lens is be a bit over 1µm. That's good enough for common analog tech, if optical aberrations don't limit performance. In terms of microscope numeric aperture ratings, f/2 is about 0.25NA.

I have successfully employed this technique using a point and shoot camera and reversed f/2 webcam lens. My biggest problem was noise from the shitty compact sensor, flare from the shitty webcam lens and poor alignment of my shitty duct tape job   Better results should be possible with better gear.

Two lenses? An interesting stackup. Sounds reasonable.
Meanwhile I try take pictures only with at least some distance rings. The 10-22mm can focus down to 0,24m that is achievable.

In the first place I will try to mount a microscope lens on my camera. I have to try this. 

[magic]

Beware that the lens must have some minimum glass diameter and minimum working distance to produce good lighting. You don't want the whole lens to end up in the die's shadow. That's another problem I had with my webcam lens setup - the lens is so small it's only good for tiny ICs, up to 1mm.

Objectives up to 10x magnification typically have 1cm or more working distance and relatively large front glass, but 20x or more may be problematic.

[Noopy]

I completely agree with you. 

[mawyatt]

Stacked lenses have an advantage that the effective aperture scales as the Magnification, whereas a single lens on extension scales as 1 + Magnification. At higher magnification diffraction becomes a limiting factor and this simple difference in effective aperture becomes very valuable. Infinite corrected microscope objectives (like the Mitutoyo) are designed to work with a special focal length (200mm) "tube lens" and can be considered as a stacked arrangement. These dedicated "tube lenses" are normally expensive, but folks over at Photomacrography.net have discovered that the Raynox 150 magnification lens works equally well and only cost ~$60, also the old Nikon 200mm F4 "Q" lenses work well as "tube lens", these can be found on eBay for under $50. Sometimes you can use microscope objectives with "tube lenses" that are not the designed focal length, for example folks have used 125 and 135mm lenses like the Raynox 250, or Nikon 135mm F3.5 (all cost under $60) with 10X 0.28NA objectives and produced stunningly sharp images at ~6X magnification.

But infinity corrected lens are more complex and costly than simple lenses. Many times simple lens can produce acceptable results, one simple lens that's a bargain is the AmScope 4X 0.1 for $18.

https://www.closeuphotography.com/seventeen-dollar-plan-4x-objective

or the old Soviet Lomo 3.7X NA 0.11 for under $60.

https://www.closeuphotography.com/lomo-3-7x-objective

There's loads of information and discussion on all these techniques as such over at photomacrography.net, and as you can see things can be quite affordable if you know what to look for and how to assemble setups.

Caution, it' becomes quickly addictive

Best,

[magic]

On AliExpress there are vendors selling adapters from various filter threads to RMS or from 52mm filter thread to M42 mount, which could be further converted to RMS using widely available adapters.

With those, any standard photographic lens could be used as a tube lens for any microscope objective. Or in other words, any scope objective could become a strong diopter filter / closeup lens for any photographic lens.

I have recently snagged a cheap vintage 10x objective from an auction site and I'm planning to play with it. Early results with just holding the setup by hand are somewhat encouraging, I wonder how it will improve with precision alignment.

edit
Is it known if those AmScope objectives are equivalent to any random unbranded Chinese objective or if the AmScope vendors on AliExpress are legit? I don't think there is any official distributor of AmScope here.

[Noopy]

Caution, it' becomes quickly addictive

It is! 


In my view there are still some questions left:

- What is better: Tube lens and infinitiy corrected microscope objective or a simple 160mm microscope lens?

- What magnification do I need. Too much magnification could be problematic but I already have something between x15 and x20 with a resolution of 1-5µm so I would need 30x or 40x to get more details?
Would it be possible to put light on the die with 30x or 40x?

[magic]

- What magnification do I need. Too much magnification could be problematic but I already have something between x15 and x20 with a resolution of 1-5µm so I would need 30x or 40x to get more details?

1µm magnified 20x is 20µm on the senor. Find out how many pixels that is and you will have an idea if you are limited by the camera or by the optics. Probably optics, because sensors and image processing algorithms have no trouble resolving lines spaced only a few pixels apart as you can confirm yourself by photographing any random thing with a known good lens.

In a two lens / infinity system, magnification can be adjusted independently of the primary lens by changing focal length of the second lens. Of course, aperture and quality of the primary lens still limit available detail and resolution. Vignetting may occur if you try too low magnification.

Would it be possible to put light on the die with 30x or 40x?

With a single 160mm 40x objective applying your lighting method will be rather difficult. Those objectives have dimensions similar to my webcam lens - take any die bigger than 1mm and the center will be dark.

[Noopy]

1µm magnified 20x is 20µm on the senor. Find out how many pixels that is and you will have an idea if you are limited by the camera or by the optics. Probably optics, because sensors and image processing algorithms have no trouble resolving lines spaced only a few pixels apart as you can confirm yourself by photographing any random thing with a known good lens.

The Canon 60D has a pixel size of 4,3µm. So with 20x that would allow me to distinguish lines separated 200nm. OK 200nm is more than we need but the dslr sensor is not perfect and the lenses are not perfect too. So 20x could be a good factor.

In a two lens / infinity system, magnification can be adjusted independently of the primary lens by changing focal length of the second lens. Of course, aperture and quality of the primary lens still limit available detail and resolution. Vignetting may occur if you try too low magnification.

I´m a bit afraid of too much glass. Every part could add degradation.

With a single 160mm 40x objective applying your lighting method will be rather difficult. Those objectives have dimensions similar to my webcam lens - take any die bigger than 1mm and the center will be dark.

I agree with that. Perhaps reflecting the light around the lens would also be possible... 

[mawyatt]

With those, any standard photographic lens could be used as a tube lens for any microscope objective. Or in other words, any scope objective could become a strong diopter filter / closeup lens for any photographic lens.

This is not exactly true, some photographic lens work well, others don't. A good example is the superbly sharp Rokinon 135mm F2, stunning sharp lens from corner to corner, but a very poor "tube lens" with the Mitutoyo infinity corrected objectives, whereas the old Zeiss 135mm F3.5, or Nikon 135mm F3.5 work well. Another old 135mm, the Vivitar has 2 versions, one works well the other does not, and this has to do with the particular "build". Also watch for vignetting when using photographic lens, it can be a problem on full frame sensors.

edit
Is it known if those AmScope objectives are equivalent to any random unbranded Chinese objective or if the AmScope vendors on AliExpress are legit? I don't think there is any official distributor of AmScope here.

These were discovered by Robert O and reports have shown that some of the similar looking Chinese objectives are not the same optics producing inferior images. I have the AmScope version I purchased direly from them a few years ago, so don't know about the AliExpress vendors. For $18 it's worth a gamble IMO. The mentioned old Lomo 3.7 NA 0.11 is very good for stacking, I got mine about 8 years ago for under $50, now the "word" is out the price has jumped up. However even at $100 this is a very good lens for stacking and both the AmScope and Lomo are finite objectives, so no tube lens required and magnification is set by extension, a bellows is ideal with these lenses. Robert has shown that the already good Lomo, gets even better when used with a tube lens!!
 

Here's a low resolution image captured with the AmScope lens on a bellows. This is test chip in a QFN package of one flavor of a patented circuit (7903016 & 7939857) developed over 12 years ago, called Direct Digital to Antenna, or DD2A.

Best,

[mawyatt]

Caution, it' becomes quickly addictive

It is! 


In my view there are still some questions left:

- What is better: Tube lens and infinitiy corrected microscope objective or a simple 160mm microscope lens
The infinity corrected are preferred for overall IQ, the Mitutoyo is considered the "Benchmark". Nikon has some good inf. corrected objectives as well.

- What magnification do I need. Too much magnification could be problematic but I already have something between x15 and x20 with a resolution of 1-5µm so I would need 30x or 40x to get more details?
We have successfully illuminated die at 20X & 50X using Mitutoyo objectives which are noted for their long working distance that really helps with lighting. We use multiple diffusions to reduce specular artifacts.

Would it be possible to put light on the die with 30x or 40x?
Yes, however higher magnification create all sorts of problems, best to start out at 1~2X and work your way up.

Check over at mentioned Photomacrography site for more details on all these questions you have, and more that are likely to come up, this is where the experts in high magnification and focus stacking reside. This quickly becomes a very complex and difficult task as you increase magnification and demand higher IQ, exponentially more difficult, and why I've taken to design and customize my own lenses and systems including the controllers & focus rails.

Here's a tiny ~1mm Indium Phosphide chip designed a few years ago, this is a low resolution version and used a Mitutoyo 20X inf. corrected with a Raynox 150 tube lenses on a custom focus rail setup. Lighting was from highly diffused multiple strobes, subject inside a white styrofoam cup I recall. Note my hobby in the upper left corner 

Best,