Wondering about gold content on 486 CPU legs vs. the gold on the metal plate and wires?

[NobleMetalsRecovery]

I'm thinking of processing about 6 to 9 pounds of old ceramic CPU's. 386, 486, Cyrix, AMD, IBM, Motorola, Pentium, Pentium Pro, etc.

Soaking them in HCL should remove all the legs. If that does the job, then I would recover my gold from the legs, and leave the gold on the metal plates, and under the metal plates for later.

Does anyone have an idea about the gold content of just the legs minus the metal plates and the gold wires. Would the legs have more than 1/2 of the entire yield?

 

[nickvc]

Before refining your chips check on collector values as that could be way more than the gold value.

 

[kurtak]

Soaking them in HCL should remove all the legs.

No --- unlike the pins on "fiber" CPUs the pins on ceramic CPUs are "brazed" (rather then "soldered") on the CPU - the braze is a gold (alloy) braze so HCl will not dissolve the braze to release the pins

Thats why (generally speaking) we process ceramic CPUs directly in AR

Kurt

 

[NobleMetalsRecovery]

Thank you for addressing the brazing issue. With that being said, that the braze is a gold alloy, I return to the main question, i.e., what percent of the total value in a Ceramic CPU is in the gold legs?

So, nitric acid should be able to remove the legs from the CPU's.

Also, thank you to nickvc for cautioning against processing a CPU with collector value. Out of that entire batch of CPU's I only spotted one that looks to be of collector value. Plan to post a pic of it here.
It's a 486DX2-66. The brand name is " It's ST ". I never seen one of those before in over 20 years of refining electronic scrap.

 

[NobleMetalsRecovery]

Here's some pics, of the CPU and a small circuit board loaded with nice chips.

 

[kurtak]

So, nitric acid should be able to remove the legs from the CPU's.

No - just like HCl alone will not dissolve gold - nitric alone will not dissolve gold

Edit to clarify; - in other words - just like HCl will not dissolve the gold braze to remove the pins/legs nether will nitric

That is why - generally speaking - we process ceramic CPUs by going direct to AR

I am not sure why you want first to try to remove & process just the legs/pins and then run the rest of the CPUs in another process of it own --- when you can simply put the CPUs though the AR process as they are & get all the gold in one process - rather then casing the gold in two different processes

I return to the main question, i.e., what percent of the total value in a Ceramic CPU is in the gold legs?

I have no idea - that's because it makes no sense to process then in two separate processes (pins/legs separate from the rest of the CPU)

You can dissolve the base metal away first - then dissolve the gold in a second step (with AR) but that is the looooong way around to recovering the gold from ceramic CPUs (& other ceramic IC chips for that matter)

If you try to dissolve the base metals away first - you have two different base metals

The CPUs with the "pins" - the pins are made of Kovar - Kovar will dissolve in HCl - BUT - it is VERY difficult to dissolve - you have to run it (the HCl) VERY hot like (simmer) boiling hot - & even then - it takes a LONG time to dissolve the Kovar away - and - because the pins are brazed on the CPU some (if not most) of the foils will still be stuck to the CPU - so you will have to literally have to scrape (at least some of) the foils off the bottom of the CPU to get all the foils

Also - most of the lids are Kover - so in the process of dissolving the Kovar from the pins you will dissolve the Kovar lids --- also - using just HCl to dissolve the Kovar will take 2 - 3 days to dissolve all the Kovar & it will take (about) twice as much HCl to dissolve the Kovar compared to dissolving both the Kovar & gold directly in AR - AND - processing them directly in AR only takes (about) a day --- so both less acid & less time

The nitric in the AR makes the HCl more effective at dissolving the Kovar (so less acid & time) while at the same time dissolving the gold

That is why (though you can dissolve the base metal first) generally we go direct to AR to process ceramic CPUs (as well as other ceramic IC chips

The CPUs with the legs on the edges are likely copper - so can use nitric to dissolve the copper

However - the gold plating is a HEAVY plating - so for the nitric to get under the plating & dissolve the copper you again need to run it HOT (simmer boiling hot) - but - like the Kovar pins - because of the braze - most of the foils will be stuck to the CPU so again you have to "scrape" the foils off the CPU --- that's "a lot" of work just for the foils

Bottom line - you are again better served going direct to AR

The CPUs with plastic lids need to be broken so that the acid gets to the gold inside the cavity

Bottom line - you are best served processing both ceramic CPUs & ceramic IC chips going direct to AR

Kurt

 

[Shark]

AR can be used two ways. One method would be as a recovery step and the other would be as a refining step. Ceramic IC’s would be an example where the first run would be a recovery process. Then the next step would be to use the AR again as a refining method where cleaning up the recovered gold is the target. Learning when to use AR which way can be tedious but rewarding when you enjoy what your doing.

 

[kurtak]

Did a couple edits to my post from yesterday - they are in bold print

Kurt

 

[kurtak]

AR can be used two ways. One method would be as a recovery step and the other would be as a refining step. Ceramic IC’s would be an example where the first run would be a recovery process. Then the next step would be to use the AR again as a refining method where cleaning up the recovered gold is the target. Learning when to use AR which way can be tedious but rewarding when you enjoy what your doing.

Recovery ? refining ?

Although we often talk about these two things as different things - are they really two different things ?

Many years ago - GSP said something that stuck in my head (& though I don't remember the words exactly - it went something like this)

There is no such thing as recovery - anytime you are "upgrading" the content of gold - it's a process of refinement - it's a question of how refined in the end of the process

Kurt

 

[Shark]

I agree Kurt. The problem comes when some people are convinced they can’t use AR as the first step. Often when it is brought up we hear don’t do that and/or trash in, trash out, often confusing the new people. Where as explaining it as a step in an over all process can often clear the confusion up. For me, refining (gold) isn’t a single, clearly defined process so much as it is using a series of steps that can go wrong at any one point. If we know each step clearly it would be much easier to point out the errors and the steps needed to correct them.

 

[NobleMetalsRecovery]

Hi Kurt,

Thanks for the edit on your earlier post.

Edit to clarify; - in other words - just like HCl will not dissolve the gold braze to remove the pins/legs nether will nitric

What got me to thinking that I could remove the gold legs first (for whatever reason) is when I was sorting out the CPU's the other day in preparation for recovery/refining was that I have about 15 or so 486 (or similar) CPU's where lots of the gold legs are gone. They have not be processed before, they were just laying around, perhaps outside for many years.

I've come across this several times before, where I remove a 486 CPU from it's socket, and some of the legs remain in the socket. So that tells me there is something that does cause the gold legs to come loose from a CPU. I'd like to know how to recreate that effect, again, just for my own reasons. Will elaborate more on that later.

Thanks to all that have replied.

 

[ssmontecarlo]

This is good stuff. Still..how much gold is on a IBM 6x86 L pr 150+? A tenth of a gram?