Non-Chemical Removal of top lid from ceramic CPUs?

[Anonymous]

Hello everybody!
Has anyone ever tried to remove the gold plated top lids which are found on some ceramic CPUs?
The purpose of these top lids is obviously to make a good thermal contact to the cooling device of the CPU. On some of the CPUs they are gold plated - for example on the Pentium PROs. They are made of some kind of base metal (probably some kind of steel?) and are really stuck on the CPU. These top lids increase quiet a lot the amount of acid needed to dissolve all the metals from a CPU with the AR method. I guess the gold plating on these lids isn´t worth processing them together with the rest. Seems that the amount of acid needed to dissolve them passes a lot the value of their gold plating. So I decided to remove this plating by reverse electroplating and than to remove the "naked" lids from the CPUs and go to the further processing. But I couldn´t remove it!!!!!! I tried a gas torch (same as to remove the bottom lids) without any result! The chisel - hammer method just broke the CPU including the lid. Finally I tried an angle grinder. Well, this worked, but this took a lot of time and electric energy. Has anyone probably found a better method to remove these top lids? Or does anyone know what type of "glue" was used to get them stuck in such a way? Following the logic of the good thermal contact, it should be some kind of solder. But wow, what kind of solder is this?

thanks in advance, donbrasil

 

[Smitty]

A heat gun running at 1200 degrees should be able to soften the solder enough to remove the lid. Interesting that the torch did not do the job.

 

[lazersteve]

You can remove the Pentium Pro and similar lids with an oxy/acytlene rig with no trouble.

Bring the lid to red heat and it slides right off. It's brazed on the cpu. It really messes up the gold plating and usually damages some of the legs, but it works. Heating in a furnace or oven may also work and be easier on the remainder of the cpu.

The Pentium Pro lid is made of tungsten (~27 g) alloyed with copper (~3 g) and plated with a small amount of gold (<0.05 g).

Steve

 

[damezbullion]

i have not a torch what can i do? i just smacked one with a hammer over a bowl and the lid shattered along with the ceramic, its cast alloy i think, theres no gold underneath, it can i just process them like this, the solder also has a layer of gold on it?

 

[patnor1011]

Yes. Lids are soldered with gold braze and thus not ideal for reverse electroplating. I just use hammer to broke them (CPU) to small pieces prior AR to let acid penetrate under (I meant under where lid was). Black epoxy in center - I remove it from chip (from black fibre CPU's), break in mortar using pestle and incinerate.

*edited - added little explanation what I meant, added things are colored

 

[damezbullion]

thats a spot on answer, thanks partnor, i dont think i can incinerate this lot i will just have to do my best crushing them with a hammer, but i will probably do this the second time i process them

 

[samuel-a]

You should distinguish heat spreader from lid.
Heat spreader is of sintered W/Cu "alloy" but needs no special treatment, just toss em in with the rest of the cpu body.

Lids are made of Kovar or Invar most likely and can be treated with cold dilute nitric for few days to remove the plating.

 

[damezbullion]

your all brilliant thanks samuel, thanks partnor

 

[lazersteve]

I'm of a different school of thought than Sam. I leave the cpus with spreaders (ones with the gold tops) whole and remove only the copper to collect the tungsten plates free of damage.

I sort all of my cpus and process all those with tungsten plates separately. This keeps pesky tungsten out of my AR reactions and out of my silver chloride. If you mix ceramic cpu types and run them all through AR together the tungsten will partially dissolve in the highly acidic solution (nice red brown solution if you treat pure tungsten with AR). Dilute this solution with water and the tungsten crashes out as what I believe is an oxide (pretty yellow powder) and it is all mixed in with your silver chloride. Tungsten will be evident in your gold only when you melt the brown powder. You'll see a lot of tiny bright dots all in the gold and the gold is stubborn to fuse to a molten mass if some tungsten sneaks in.

Here's another oldie, but goldie about tungsten oxides with more photos.

Tungsten Oxides with Pictures

Another example of GIGO (Garbage In Garbage Out) in action.

Steve