Differences

This shows you the differences between two versions of the page.

--- decap:epoxy [2018/02/27 20:49] – [Sandblast] mcmaster
+++ decap:epoxy [2025/08/04 21:23] (current) – external edit 127.0.0.1
@@ Line 1: / Line 1: @@
-[[live_analysis|Live analysis considerations]]
+Also known as "plastic" packages
-====== Background ======
-[[carrier:epoxy|Find out what an "epoxy package" is here]]
+See also
+  * [[:recap|Re-encapsulation]]
+  * [[carrier:epoxy|Find out what an "epoxy package" is here]]
+  * [[live_analysis|Live analysis considerations]]
 ====== Getting started ======
@@ Line 133: / Line 133: @@
 One problem from this lack of circulation is that it may create pockets of low concentration acid, leading to corrosion.  The above dies were washed before that picture was taken to remove copper nitrate residue that had built up.  It probably came from a paddle but may have come from the PCB.  After rinsing with water, the bottom of the die was CA glued down.  Then the back of the die was gently cotton swapped with acetone.  Finally, the whole thing was soaked in acetone to release the die.
+===== McMaster live 2019-11-10 =====
-====== Live analysis considerations ======
+{{:mcmaster:decap:epoxy:cob_silicone.jpg?400|}}
-Commercial "automatic" machines use a variety of techniques to etch the IC appropriately. One such technique is to seal off the area under attack using a chemically inert gaskit. See nitric acid gasket compatibility chart[9].
+Experiments trying to decap COB while keeping PCB intact
-Sulphuric acid may be added to nitric acid to help preserve copper, if present (see the acid page for details).
-====== See also ======
+TLDR best result
+  - Cut a polypropylene tube (ie from a microfuge tube) roughly the size you want to expose to acid
+  - Heat tube up to soften it
+  - Warm PCB
+  - Attach hot tube to PCB, letting plastic deform into place
+  - Let cool a little
+  - Seal outside of tube with silicone
+  - Let silicone cure
+  - Decap with gentle heat on hotplate (maybe 160F max) using 2 WFNA : 1 H2SO4
-[[:recap|Re-encapsulation]]
+Above is a compromise between PP providing the best chemical resistance and silicone providing the best seal
+{{:mcmaster:decap:epoxy:silicone_surface.jpg?400|}}
-====== Laser ======
+Above: hardened silicone bubbles on WFNA surface
-General results
+Silicone only experiment
-  * Some settings work decent
+  * Made silicon dam by layering up
-  * Risk of melting bond wires and/or die
+  * Silicone surface cured, but internally didn't always
-    * Use lower power and/or pulsed?
+  * Uncured silicone dissolved in WFNA and floated to top. This then cured and blocked the surface
-  * Practice on samples before trying something you care about
+    * Interesting, but not what I was looking for
-  * Puts off nasty fumes. Make sure to have good ventilation
+  * Acid may eventually undermine the PCB, but eating the silicone itself came first
+  * Consider using a lot more silicone than you need to help mitigate acid splashes
+    * Maybe cover board surface with kapton?
-https://www.jamiecraig.com/de-encapsulating-ics-with-a-laser-cutter/
+Someone also suggested I should try PTFE gasket. In the past I had tried kapton which didn't work well (acid tended to wick into kapton). Maybe PTFE would repel acid better
-  * Seemed to work okay
-  * What would it look like if touched up with WFNA?
-{{:decap:epoxy:img_20180227_101636.jpg?400|}}
+Reference: https://twitter.com/johndmcmaster/status/1193807272171827200
-Above: poor results. Maybe from CO2 laser
+====== Live analysis considerations ======
-http://siliconexposed.blogspot.com/2013/08/laser-ic-decapsulation-experiments.html
+Commercial "automatic" machines use a variety of techniques to etch the IC appropriately. One such technique is to seal off the area under attack using a chemically inert gaskit. See nitric acid gasket compatibility chart[9].
-  * Poor results
+Sulphuric acid may be added to nitric acid to help preserve copper, if present (see the acid page for details).
-  * Cranberry glass
-https://www.youtube.com/watch?v=-x4lyX1C4q0
-  * Gold backing block
-  * Good leadframe exposure
-  * Die unclear how well exposed
-  * Scanned entire chip (including die area) up until the end
-https://www.youtube.com/watch?v=2HMocYiwMYc
+====== Laser ======
-  * Cavity into pic
-  * Gives a close up at the end
-https://www.youtube.com/watch?v=SvWgSs2_f00
-  * Laser Decapsulation FALIT System Software | Image Templates
-  * Pro system demo
-https://www.youtube.com/watch?v=sSRl3Vqbos8
+[[decap:laser]]
-  * They do the whole chip but avoid die area
-  * Light Ray - IC Decapsulation
@@ Line 259: / Line 254: @@
-====== A. Zonenberg etch rate measurement ======
-FIXME: move this somewhere better
-Copper etch - 70 ml 3% H2O2 + 10 ml conc. HCl. Heat to 70C, agitate with magnetic stirrer.
-Test sample was Altera Stratix IV. Silver or tin plated copper. Plating was gone after <1 min in etchant, metallurgy of the plating was not studied. 20 minutes of etching = 0.13mm removal so 6.5 um/min. This number is likely +/- 30% or so because the temperature of the solution slowly increased over the etch period.
 ====== References ======