Merged Topics on Ceramic CPU Recovery with Yields

[lazersteve]

Steve, is there the posibility that the first PP's models had more gold & as the technology advanced, less gold was needed to make them?

Thanks!
Phil

I've given the idea some thought today and here's what I've come up with so far:

• differences in scales (calibration, battery level, accuracy, level surface, etc.)
• differences in cpu runs (ie: date or country of manufacture )
• differences in resulting gold purity (contaminated verse pure gold produced)
• surface elemental gold on the cpu that is not fully removed (pin holes, under cores, caught in braze on top lid perimeter)
• gold that has deposited into the tungsten top plate lattice after the copper dissolves out. (Think copper cementation on the top lid while in AR)
• surface gold solution on the cpu that has adhered to the ceramic or other crevices of the cpu housing.

I'm sure you guys can come up with some more possible losses/errors.

Steve

 

[rotorpowa]

Ok I think this was the site I got my info from. As I said before I might have missread or not remembered correctly and I did. Here is a link to the site lots of you should find interesting http://www.ozcopper.com/computer-cpu-gold-yields/

Oh and please excuse my spelling not one of my strong points

 

[cnbarr]

Samuel please take no offense, I'm not trying to hijack your thread, just thought it would be good to keep pertinent information all in one place. Since we are on the subject of processing ceramic cpu's, I thought I would post some yields from recent test batches of various sorted ceramic cpu's. All were processed in the same manner as Samuel's tutorial, except I used nitric instead of sodium nitrate, along with a second refining for purity. I DID NOT extract 100% of the values, there was still some plating left under the silicon chips. I feel this may prove a good starting point for some newbies, as to what they can expect to recover, but also knowing that these yields can be increased with experience.

 

[lazersteve]

This is the new merged topic of ceramic cpu related posts and threads.

I have merged together many old and new threads on the subject to make this thread.

Steve

 

[cnbarr]

Hi all,
This is a great call Steve, organizing it all into one place, it should definitely make it easier to search for.

Thank you.

 

[philddreamer]

CN, the yields you've posted are a good clue for anyone processing similar ceramic cpu's.
Now, you probably just broke them into big pieces, that's why the yields could be a bit higher. I say that, because on the PP's you had the same 1.8g yield I had on my 6 that I just simply broke into 6 or 7 pieces. On my second batch of 6, I made sure I broke them smaller & that ALL the silicon waffers were off & crushed. That batch yielded 2.4g, thus .4g each, closer to my .5g of a previous test, which I happened to crush down to about 1/4".

I went thru my left over ceramics from previous batches that I only broke & also left the silicon waffers on, & the pic's show the results of why the yields were lower than expected.

I also took the Corning were the ceramics have been sitting for over 6 mos. after the final rinse; there was a dry yellow sediment to which I added 1/4 cup of HCl & a few drops of nitric, then applied some heat & the stannous test was + for gold.

So, after doing 3 tests with PP's & acquiring .3g each by just breaking; .4g each by breaking into smaller pieces plus crushing the silicon; & .5g each after breaking into very small pieces, I believe these older ceramic cpu's should be crushed & leached twice in order to achieve a higher yield.

I can't wait to learn of the results from tests of Steve's samples.

Phil

 

[rotorpowa]

wow Phill that is a good pic to show just how much is hidding behind those little suckers. Maybe one of the mods might even turn this into a sticky

 

[Barren Realms 007]

Phil,

I see a lot of those chips still have the copper tungsten heat spreader on them. You might think about soaking these in nitric to seperate the heat spreader and you might find that the solder used to attach the heat spreader to the ceramic chip could be gold based. They can also cement out some of your gold in solution and you might miss valvues there.

 

[samuel-a]

cnbarr

I have posted this just for someone like you to post what you did 8) All cool.
------

Steve, you done a great work collecting all the scatterd posts into one place, bet it took a while, Thanks.
Waiting for your results on the re-processing.

 

[lazersteve]

Phil,

The yellow is most likely tungsten oxide and may trap some values in the salt structure as well.

The actual yields on the PP may actually be closer to 0.75g as Sue once alluded.

I still think the gold is all on the surface and not in invisible wires inside. Perhaps breaking them up exposes the gold on the feed thru connections for the legs to the internal wiring and some from the top lid (cemented and/or braze). The most likely canidate for trapping values would be the copper in the tungsten copper alloy top plate. You've processed pins and seen how copper cements out gold when the reaction is not driven to completion.

Perhaps you can get it all as Barren has stated, by dissolving down until the lid falls off.

The top lids can be removed with an oxy/acetylene torch in a few minutes at glowing red heat, but the gold turns dark/rainbowed as it alloys/absorbs into the top plate structure. The top plate weighs around one ounce when done this way.

Steve

 

[lazersteve]

As a side note, I too re-ran the ten cpus that had been processed and got a positive gold reaction.

I didn't break then up any either time.

I could probably get another yield from repeating the process again without breaking.

I'm still working on some pictures.

Steve

 

[philddreamer]

"The actual yields on the PP may actually be closer to 0.75g as Sue once alluded."

WOW! I like to hear that, since I have broken ceramics from about 80 PP's that the yield was around .3g a piece; that could mean a few more grams!

I also believe that the top plate's solder holds some gold. I only heated with mapp to try to get the tops off, but no success. I should have tried with oxi/acte. We live and never finish learning.
I'm just glad we started "kicking" this around! 8)

Thank you gentlemen!

Phil

 

[lazersteve]

Giving the subject more thought, perhaps breaking the cpus up simply allows the acid to reach areas on the surface of the cpu that would normally be trapped or protected by the cpus piling on top of one another.

By breaking up the cpus you increase the exposed surface area to the acid as the irregular broken pieces do not stack perfectly on each other like larger whole cpus and pieces do.

Steve

 

[Barren Realms 007]

"The actual yields on the PP may actually be closer to 0.75g as Sue once alluded."

WOW! I like to hear that, since I have broken ceramics from about 80 PP's that the yield was around .3g a piece; that could mean a few more grams!

I also believe that the top plate's solder holds some gold. I only heated with mapp to try to get the tops off, but no success. I should have tried with oxi/acte. We live and never finish learning.
I'm just glad we started "kicking" this around! 8)

Thank you gentlemen!

Phil

Phil,

Mapp gas will get the chip hot enough to reove the top plate, you might have to break the chip in two for it to work.

 

[philddreamer]

Hi Barren!
I tried the mapp & got them hot and some shattered in 2 or 3 pieces, but the top plate never came loose. I even used some pliers to try & "help" them come off, but nothing. The solder never liquified. Next time I'll use the oxi/acte.

"... perhaps breaking the cpus up simply allows the acid to reach areas on the surface of the cpu that would normally be trapped or protected by the cpus piling on top of one another."

Steve, my same thoughts. A couple of days ago, when I inspected the left over ceramics I noticed the areas where the acid never did much digesting, because the pieces had laid over each other, preventing the AR to attack the surface. The big pieces were also hard to stir around.

"By breaking up the cpus you increase the exposed surface area to the acid as the irregular broken pieces do not stack perfectly on each other like larger whole cpus and pieces do.
Steve"

I agree!

Thank you gentlemen!

Phil

 

[maynman1751]

Phil,

The top lids can be removed with an oxy/acetylene torch in a few minutes at glowing red heat, but the gold turns dark/rainbowed as it alloys/absorbs into the top plate structure. The top plate weighs around one ounce when done this way.

Steve

Steve: This in no way 'damages' the gold or makes it more difficult to recover??????

 

[cnbarr]

Now with this thread, I wish I would not have dumped my left over ceramics all into one bucket. Something told me I should have kept them separate for testing purposes such as this current discussion.

I guess I can still crush them and post some results.

Thank you everyone, 8)
Chris

 

[lazersteve]

If your gold is already being trapped in the tungsten/copper alloy of the lid, you already have a problem to solve to get the gold.

Removing the lid serves several purposes in my eyes:

• Exposes the braze to the the acid.
• Allows the top lid to be processed separately to determine if the yields are affected primarily by the lid removal.
• Removes the bulk of the tungsten from the remainder of the cpu to determine where the 'extra' gold really is coming from.
• Typically breaks up the cpu package a little to increase surface area and acid exposure.

All in all no gold is lost if you do your due diligence and process until no more values are achieved.

The real purpose of this exercise is to determine where the gold in the cpu is really located and whether you must partially or completely pulverize the cpus housing to get all of the gold out in the minimum number of extraction cycles.

I removed a lid this morning to show how they look removed:

Unfortunately this one broke when heated. It's a lot of work so I would not jump in and do a bunch of these until we determine where the gold really lives in these things.
Steve

 

[Barren Realms 007]

Steve,

If the solder used is a gold based solder would it not be a good idea to run these in nitric before the test is done to remove as much of the solder from the CPU as possible?

 

[lazersteve]

The solder melts silver colored, but at a fairly high temperature so it may contain some gold.

Nitric is a good idea as a preliminary treatment of the lid as the nitric will remove the copper from the tungsten also.

Steve