Microscope for decapped ICs

[caius]

Hi all, I have some decapped ICs and the chance to do more (actually I don't used nitric acid but a "burning" method) but I'm not able to take magnified HI-RES pictures.I'm not into microscope market so I was wondering if there are some not expensive models suitable for my purpose.Thanks in advance.

[amyk]

It depends what type of ICs and how much detail you want. Individual transistors of early (~10um) process are visible with a magnifying glass, whereas a light microscope won't be able to see much of anything on those made in a sub-micron process as the features are far smaller than the wavelength of light.

[tggzzz]

For amusement, I posted some pictures of late 70s Burr-Brown thick film hybrids. Definitely not sub-micron!
https://entertaininghacks.wordpress.com/2015/07/07/images-of-late-70s-burr-brown-thick-film-hybrid-ics/

[TerraHertz]

I'm not into microscope market so I was wondering if there are some not expensive models suitable for my purpose.Thanks in advance.

Similar to my situation. I found the search term 'metalurgical microscope' is a help. It means microscopes intended for reflected light use, inspecting opaque surfaces.
My Olympus BHM adventures here: http://everist.org/NobLog/20150510_Olympus_bhm.htm
I still haven't got a camera adapted well to the microscope, but that's not an intrinsic problem, just me getting sidetracked as usual.

[EEVblog]

My Olympus BHM adventures here: http://everist.org/NobLog/20150510_Olympus_bhm.htm
I still haven't got a camera adapted well to the microscope, but that's not an intrinsic problem, just me getting sidetracked as usual.

I have an Olympus BHM as well, just need some time to get it going. Unfortunately the light source was broken on it like yours. Thanks for that info on the LED upgrade, will read in detail.

[TerraHertz]

I have an Olympus BHM as well, just need some time to get it going. Unfortunately the light source was broken on it like yours. Thanks for that info on the LED upgrade, will read in detail.

For my next microscope illuminator project, I've come up with a better way of constructing the optical concentrator. Two parts combined:
* Ebay search term: LED reflector collimator     eg item 371258089004  and many others.
 There are sizes to suit the 10W and 50W flat LEDs. You may find that alone will work well enough with the lens included.  Beware that what you end up with is a nice focused image of the LED chips.
I'll be using just the reflector part, to get better matching from the LED to the concentrator cavity, and allow homogenizing of the light.

* I'm planning to use several large LEDs combined for colour control in this one. So still need the optical concentrator method. But I'll be trying making a solid former of the optical cavity shape first (probably of wood), then molding the body structure around it. Either in casting epoxy or wrapped and soldered copper sheet.
Then gluing reflective foil to the inside of the cavities, same as before.

That project is getting posted here as it progresses: http://everist.org/NobLog/20150628_LED_lighting_mk3.htm

The 10W colour LEDs arrived today.

[RogerRowland]

Is it an easy thing to do, decapping an IC?

I'm just asking because I work at a company that manufactures SEMs (scanning electron microscopes) and I'm wondering about using some decapped IC images for testing some of our image stitching software (for large scale microscopy).

So, how does one decap an IC without destroying the contents?

[helius]

It depends on the type. Some large chips like CPUs have metal lids that can be ripped off. For epoxy (plastic) potted chips, they can be embrittled with liquid nitrogen and then crushed, hopefully leaving the die undamaged. Or heated until the epoxy decomposes and crushed...
But to decap the chip while preserving the bond wires so it can operate, the best method is to mill a slot above the die and fill it with red fuming nitric acid.

[RogerRowland]

It depends on the type. Some large chips like CPUs have metal lids that can be ripped off. For epoxy (plastic) potted chips, they can be embrittled with liquid nitrogen and then crushed, hopefully leaving the die undamaged. Or heated until the epoxy decomposes and crushed...
But to decap the chip while preserving the bond wires so it can operate, the best method is to mill a slot above the die and fill it with red fuming nitric acid.

Ahh, so not a simple thing then. Maybe it's easier to dig the window out of an EPROM like the older PIC16C* micros I wonder? I'll have to see what I can do that might give me some mileage.

[helius]

Were you wanting to image the electrical charge in the chip as it operates? I know a SEM can do that.
Quartz-window chips are a ceramic sandwich, the trick there is to separate the ceramic layers without ripping the traces. It might be possible to dremel just above the legs to take off the top layer.

[RogerRowland]

No, not interested in keeping it alive. Too tricky to set that up in a vacuum. I don't want to complicate things but just get the detailed structures for image processing.

I'll have a look for some sacrificial components this evening 

[edavid]

No, not interested in keeping it alive. Too tricky to set that up in a vacuum. I don't want to complicate things but just get the detailed structures for image processing.

I'll have a look for some sacrificial components this evening 

TO-5 packages are really easy, just cut off the lid.

CERDIPs are pretty easy too (don't have to be quartz lid).

[codeboy2k]

For amusement, I posted some pictures of late 70s Burr-Brown thick film hybrids. Definitely not sub-micron!
https://entertaininghacks.wordpress.com/2015/07/07/images-of-late-70s-burr-brown-thick-film-hybrid-ics/

Those are awesome! thanks for scanning and posting.

Most c00l3st ... the bondwire toroid

[tggzzz]

For amusement, I posted some pictures of late 70s Burr-Brown thick film hybrids. Definitely not sub-micron!
https://entertaininghacks.wordpress.com/2015/07/07/images-of-late-70s-burr-brown-thick-film-hybrid-ics/

Those are awesome! thanks for scanning and posting.

Most c00l3st ... the bondwire toroid

You're welcome. They are pretty, aren't they!