Differences

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

--- decap:ceramic_glass [2016/12/03 20:49] – mcmaster
+++ decap:ceramic_glass [2016/12/03 20:50] (current) – removed mcmaster
@@ Line 1: / Line 1: @@
-JM 2016-11 advice:
-  * Use "Top Heat (gentle)" for safest result
-  * Use "Top Heat (quick)" for less safe but cleaner result
-===== Top heat =====
-Gentle procedure:
-  - Place chip into heat spreader block
-  - Place chip + spreader into vice
-  - Torch heat spreader
-    * DIP28: ~10 seconds
-    * DIP40: ~20 seconds
-  - Begin torching top of chip
-    * Use the side of the flame, not the tip, for gentler and more even heating
-  - Every 5-10 seconds try to pry lid up.  If it is starting to budge, immediately do the next few steps
-  - Stop torching
-  - Gently pull lid up and over chip, hinging on the far side
-  - Separate lid from package
-{{:mcmaster:decap:glass_frit:quick_before.jpg?300|}}
-{{:mcmaster:decap:glass_frit:quick_before.jpg?300|}}
-Above: sample quick result
-Quick procedure:
-  - Place chip into heat spreader block
-  - Place chip + spreader into vice
-  - Begin torching top of chip
-    * Side or tip of flame better?
-  - Every 5-10 seconds try to pry lid up.  If it is starting to budge, immediately do the next few steps
-    * Listen for a small popping noise.  This usually indicates the lid has sheared from the frit and can be easily removed, possibly with slightly more heat
-    * If its not the lid, the package has probably cracked
-  - Stop torching
-  - Gently pull lid up and over chip, hinging on the far side
-  - Separate lid from package
-General notes:
-  * Wave torch back and forth to keep heat even
-  * FIXME: torch model I use
-  * DO NOT torch while lid is lifting.  Will quickly destroy bond wires
-  * Heat spreader can be aluminum or steel
-    * Steel won't be melted by propane torch
-    * Aluminum can melt if abused, but will conduct heat better
-  * If the package cracks, finish procedure and then assess damage.  The spreader should keep package integrity until assessment is done.  If the die cracks, the chip is probably a loss
-===== Thermal fracture (old) =====
-Procedure
-  - Place package on a temperature resistant surface
-  - Heat package with a propane torch until hot, maybe 5 or 10 seconds
-  - Gently splash or sprinkle water on the package.  This should create micro-fractures and make it easy to shear
-  - Shear the package open by applying light force.  Unlike a non-fractured package this should be possible using simple pairs of pliers or by hand
-Tips:
-  * Chip does not have to be heated until red or white hot.  Heating this high increases risk to damage the die from glass reflowing as well as general thermal stress
-  * Do not throw the chip into liquid or move it all while it is hot.  This can cause the glass to shift and get onto the die
-  * PDIPs should be supported (ie with a block of metal) as the leadframe will become very weak
-  * If it does not shear open the first time try again.  Consider increasing heat, pressure, or simply do more runs
-Advantages:
-  * Can be very clean and low risk
-  * Tends to be pretty clean and not throw a lot of debris onto the die
-Disadvantages:
-  * Thermal shock can potentially also break die
-  * High temperatures can warp die
-  * The die is exposed but still in the package
-==== Media ====
-Chip before (both identical):
-{{gallery>:decap:ceramic_glass:before.jpg|}}
-One of these was thrown into water and the other had water dripped on it.  Comparison:
-{{gallery>:mcmaster:decap:ceramic_glass:post_both.jpg|}}
-Dunked chip close-up:
-{{gallery>:mcmaster:decap:ceramic_glass:post_dunk.jpg|}}
-Dripped on chip close-up:
-{{gallery>:mcmaster:decap:ceramic_glass:post_drip.jpg|}}
-The motion of throwing the chip into the water caused the glass to flow and stick to the die.  When the top was sheared off some of the die snapped off as a result.  Additionally, it isn't as regular from moving around and so didn't break nearly as cleanly.
-===== Lift-off =====
-Requires some form of torch. Heat up and pry up. Be careful not to let molten glass land on the IC. The best way to avoid this is to lift straight up and let it cool a bit before moving off to the side.
-===== Shearing =====
-The glass tends to be much weaker than the ceramic.  Therefore, its possible to shear a chip open by moving the top and bottom half in different directions.
-Procedure
-  - Place bottom half of package into vice.  It may be desirable to put a metal space in between the pins to keep pressure more even
-  - Firmly hold the top of the chip and twist.  Keep the movement orthogonal to the glass plane
-Ideally this would probably be done using a custom fixture to hold the top and apply even and straight force
-Tips
-  * Plumbing wrenches work well to hold the chip as their width can be adjusted for even pressure while still holding it tightly
-  * Be careful with long CERDIPs that you don't apply too much torque on an end and snap the chip in half instead of shear it
-Advantages
-  * Very readily available materials
-  * Does not use any consumables
-Disadvantages
-  * Throws debris onto die
-  * Can be difficult to effectively grip top half
-  * Can be difficult to tell how much force is required
-  * Unpredictable how it will shear
-===== Grinding =====
-I see no reason why you couldn't grind one open (eg with a Dremel) but it would probably take a long time and would require extensive cleanup after.
-====== Die ======
-This has been a problem area and so the best advice I have is simply don't try to remove the die unless there is a good reason.  If someone knows of a good way to remove glass affixed dies I'd be interested.  All of these assume the package has already been opened.
-===== Hot pull =====
-Procedure
-  * Heat lower package
-  * Gently pickup the die up with metal tweezers or lift up with a sharp knife (which can get between die and substrate easier)
-Surface tension makes simply plucking the die up difficult and so a knife tends to work better.  Unfortunately, metal tweezers / knives chip dies and plastic tweezers melt.  PTFE does a little better but puts off toxic gas (HF) when heated so its not a good idea.
-===== Crack-n-push =====
-Procedure:
-  * Score one side of the ceramic package well
-  * Heat package strongly
-  * Submerge scored side of package in water.  This should cause it to crack roughly along score marks
-  * Re-heat package
-  * Slide die out of open side of well with die
-===== Dissolve =====
-% HNO3 is known to dissolve the frit.  However, it would probably dissolve other things (ie the pins) as well
-It should be possible to use HF to dissolve the ceramic.  Maybe NaOH could work as well?
-===== References =====
-http://www.freepatentsonline.com/4992628.html
-http://www.ti.com/lit/an/snoa280/snoa280.pdf