Gold Recovery Calculations Way Off - What Happened?

Hello all,

I had bought 10 Pentium Pro cpu's earlier this year before I knew about a .pdf file called "eScrap Refiners Gold Mining Guide". It was told in two separate YouTube videos that there is approximately 3g of gold in each one. So I bought the cpu's I saw on eBay. After refining them, I got way less than what I thought. Afterwards, someone pointed me to that .pdf file and when I opened it up and found the Pentium Pro listed in there, it calculates that I should get approximately 0.3 -0.5g of gold. My recovery was more like 0.27g per cpu. Not a big deal, and the guide was almost accurate on the recovery from what I did recover.

Now, here is where my recovery calculations are way off. I had 300 ceramic cpu's (150 AMD 486 & 150 AMD K6). The AMD 486 should yield around 0.12g each and the AMD K6 should yield around 0.11g each, according to the guide. So, by doing the math, I should have recovered approximately 30g, and that's using just 0.1g from every one of them each. I have nowhere near that amount of powder at all.

My stannous is good (I make my own) and I test it with my gold standard. So, I'm asking myself "Why is my recovery from these cpu's so far off?"

Question 1: Once the pins have been dissolved in the solution, is that all the gold?
Question 2: Is there any gold inside of the ceramics and if there is gold in them, didn't the solution dissolve all of it?
Question 3: Am I wasting time trying to crush the cpu's now that the pins are dissolved?
Question 4: Do I need to evaporate the solution and try to add some fresh AR to it?

I still have all of my solution left over and I still have the cpu's.

Kevin

[EDITED] I meant to say 3g gold instead of 1g. I put the correction in red above.

With material like cpus and provided you did everything correctly what you get is what you get. That's why those type of "guides" are almost useless. I wouldn't even use the guides that try to catalogue items and their gold content as a general guideline. I understand you already bought them but for anyone else looking at this thread I would say when it comes to cpus, "chips" and fingers either toll refine them or get them for very cheap - free.

goldenchild said:

With material like cpus and provided you did everything correctly what you get is what you get. That's why those type of "guides" are almost useless. I wouldn't even use the guides that try to catalogue items and their gold content as a general guideline. I understand you already bought them but for anyone else looking at this thread I would say when it comes to cpus, "chips" and fingers either toll refine them or get them for very cheap - free.

Click to expand...

Good point. The Pentium Pro's I lost $126 or I paid $126 more than what they were worth. With the 300 cpu's, I pretty much broke even, to say the least. But my time and chemicals put me at a loss too. Overall, I got the experience in refining ceramic cpu's, which I never did before.

Thanks Goldenchild for the reply.

Kevin

The guide is throwing you off. As far as I know there are two yield guides for ceramic CPUs. One written by Samual-a, a board member which is considered more accurate and another written by anonymous which should not be relied upon.
Looks like you are using the latter yield guide.
Personally, I have both of them printed out and on my desk and I use them for some chips. I'll also use my own yield data and update the guides. If you pull out both guides and compare yield data you'll see that one or the other is wrong. I tend to rely on Samual's guide more although I still question the stated yields on some CPUs. My returns matched on some and didn't on others.
An AMD K6 only yields 0.035g each which would explain why your return didn't match your expectations.
Once you process a number of ceramics you'll start to learn what to expect from a CPU for yield by looking at it's construction... does it have gold bond wires, does it have gold braze, does it have one or two gold plates?

resabed01 said:

The guide is throwing you off. As far as I know there are two yield guides for ceramic CPUs. One written by Samual-a, a board member which is considered more accurate and another written by anonymous which should not be relied upon.
Looks like you are using the latter yield guide.
Personally, I have both of them printed out and on my desk and I use them for some chips. I'll also use my own yield data and update the guides. If you pull out both guides and compare yield data you'll see that one or the other is wrong. I tend to rely on Samual's guide more although I still question the stated yields on some CPUs. My returns matched on some and didn't on others.
An AMD K6 only yields 0.035g each which would explain why your return didn't match your expectations.
Once you process a number of ceramics you'll start to learn what to expect from a CPU for yield by looking at it's construction... does it have gold bond wires, does it have gold braze, does it have one or two gold plates?

Click to expand...

Here is a snapshot of the pdf file I relied on.



I did document the cpu's and the yield data so that way I'll know what I'm dealing with in the future.

Where can I download the other guide?

Kevin

If you can find a post by Samual-a you will find a link to the guide in his signature line.

Guess what? I must be processing the cpu's correctly because according to the yield you just gave me, I have approximately that amount of twice-refined gold powder.

Thanks for the info. It appears the guide that I have been relying on has been a 50/50 on yield accuracy. I'll look for the other guide now.

Thanks for your input.

Kevin

[ADDED INFO]It appears that the answers to my above questions have been answered, especially crushing up those cpu's I still have, and the solution I have. Thanks again for the learning experience being taught..... Oh, and I lost money on the cpu's too. :x I'll take 50% fault on that one because I bought those cpu's after referring to the guide first.

I downloaded the other pdf file and it doesn't show the AMD at all. So, all in all, even with both pdf files, a person should and must take their own yield data and record it and hopefully pass it back to the author(s) of the pdf files so that way they could be revised. The one I originally downloaded has more data to it than the one I just now downloaded. Either way, I'm making my own notes.

Thanks again.

Kevin

Counterfeit chips ! Used to be as big a market as counterfieit money.

testerman said:

Question 1: Once the pins have been dissolved in the solution, is that all the gold?
Question 2: Is there any gold inside of the ceramics and if there is gold in them, didn't the solution dissolve all of it?
Question 3: Am I wasting time trying to crush the cpu's now that the pins are dissolved?

Click to expand...

Hi testerman.
If you have not broken the cpu's into pieces I have some good news for you. That is were your profit is hiding.
Watch this movie by our member Lazersteev and it will become clear.

https://www.youtube.com/watch?v=tQ_R8Gr26kk

lanfear said:

testerman said:

Question 1: Once the pins have been dissolved in the solution, is that all the gold?
Question 2: Is there any gold inside of the ceramics and if there is gold in them, didn't the solution dissolve all of it?
Question 3: Am I wasting time trying to crush the cpu's now that the pins are dissolved?

Click to expand...

Hi testerman.
If you have not broken the cpu's into pieces I have some good news for you. That is were your profit is hiding.
Watch this movie by our member Lazersteev and it will become clear.

https://www.youtube.com/watch?v=tQ_R8Gr26kk

Click to expand...

That video is from Samuel-a, not Lazersteve.

The gold is in plating of the lids and pins, plating inside the chip cavity, bond wires and possibly brazing of the lids and under the chip. If you have removed all visible metal and the chip is loose from the ceramic package then you have gotten all the gold. There is no gold inside the ceramic body... at least not in any ordinary CPU.

Göran

Below is what the cpu's look like, the exact ones.



To lanfear: A person would be foolish to not break the cpu;s into pieces because of two reasons.
Reason no#1. If there is gold inside it would be close to impossible to recover if they weren't broken into pieces.
Reason no#2.. You would be using way more solution (AR in my case) just to cover the cpu's. That would simply be a waste of solution.

I even took some of the broken cpu's and even broke them some more and put them under a loupe and a lighted loupe and I could see small lines running across the length of the cpu's but it look like the wires have been dissolved away. All pieces I break some more all look the same inside. I do, however saved 2 of the cpu's, one from each that I didn't process yet. Maybe I'll break one of them apart to see if the lines inside are gold plated, or maybe my loupe isn't powerful enough to really see the gold wires.

To g_axelsson: I did break the cpu's into pieces but I didn't specifically remove the chips. You're talking about the glass spot in the middle on the top? If so, I can't say I completely removed them, but I can say that I took a hammer to each one by one, and broke them into about 5 pieces or so with one whack with a hammer. Most of my cpu's broke from the middle anyway and broke like a spider web into pieces (about 5 pieces each).

To resabed01: I meant to ask you earlier as to where did you get the yield of 0.035 from? It's not on any of the pdf files I have. I'm thinking those are your yields? If so, that yield amount has been the most accurate I've seen for these cpu's so far.

Unless someone can somewhat show a different yield data for these chips, I'm going to go with the yield of 0.035 for these cpu's I have refined.

On another note, I did read in that one pdf file that I have is that people miss alot of the gold hidden inside of components mainly because what they see on the surface is what they think all there is.

I should take about 1 lb worth of the broken cpu's, crush them up real fine and then process them and see what I can get. That way I will know for certain about if there's more gold in there or not, and if I actually missed most of the gold.

Oh.. one question to ask. Have any of you crushed your ceramic cpu's and recovered more gold than what a data chart said? or was it the same?

Thanks all for the input.

Kevin

What you are showing are NOT an AMD 486 http://www.chipdb.org/img-amd-am486-dx4-100v16bgc-n-in-corner-3177.htm

The other I would guess is a typo for the AMD K5 http://www.chipdb.org/cat-pr120-27.htm

The picture is an AMD K6 and AMD Athlon

Eric

etack said:

What you are showing are NOT an AMD 486 http://www.chipdb.org/img-amd-am486-dx4-100v16bgc-n-in-corner-3177.htm

The other I would guess is a typo for the AMD K5 http://www.chipdb.org/cat-pr120-27.htm

The picture is an AMD K6 and AMD Athlon

Eric

Click to expand...

You're absolutely correct about the AMD Athlon. I was sold what was suppose to be AMD 486, but after I zoomed in on the glass chip on the pdf file, it says Athlon. Thanks for pointing that out to me. But, according to the pdf file, their yields are practically the same, so, with either one I would have, AMD 486 or AMD Athlon, the yields are way off according to the guide I used.

Thanks again for pointing that out to me. I even looked at my saved purchase history and it was listed as an AMD 486. Oh well!

Kevin

Frankly a lot of this is about product knowledge. One only has to goggle "AMD 486" and look at the images as a cross check to see the massive difference between the products. As you rightly mentioned it's also written on the chips clearly so I would have checked on delivery and gone back to the seller.

With regards to the Pentium Pro chips. I've not seen anyone suggesting 3g of gold per chip before but again I would have goggled "gold yield pentium pro" or something similar and done my own homework before buying. Strangely enough it was a similar search that brought me to this forum. Hey ho we live and learn Sir, we all make mistakes.

Jon

testerman said:

To resabed01: I meant to ask you earlier as to where did you get the yield of 0.035 from? It's not on any of the pdf files I have. I'm thinking those are your yields? If so, that yield amount has been the most accurate I've seen for these cpu's so far.

Unless someone can somewhat show a different yield data for these chips, I'm going to go with the yield of 0.035 for these cpu's I have refined.

On another note, I did read in that one pdf file that I have is that people miss alot of the gold hidden inside of components mainly because what they see on the surface is what they think all there is.

I should take about 1 lb worth of the broken cpu's, crush them up real fine and then process them and see what I can get. That way I will know for certain about if there's more gold in there or not, and if I actually missed most of the gold.

Oh.. one question to ask. Have any of you crushed your ceramic cpu's and recovered more gold than what a data chart said? or was it the same?

Thanks all for the input.

Kevin

Click to expand...

Kevin,

As pointed out by the others, the aluminum top AMD K6 is the top CPU in the picture and the lower one is the Athlon. I call them Durons but that's just me.
The yield of 0.035g for the K6 was calculated from my actual return of these chips. I ran several batches of ceramic CPUs this summer and collected my own yield data. I did not crush the AMD chips, I processed them whole after removing the silicon die and any MLCCs from the tops. I know, it's extra work but I wanted to process the MLCCs separately. The AMD Athlon has yield of slightly better than 0.04g each, less than the stated 0.11g but better than the K6. I just finished a batch of Athlon chips so once I finish washing the gold powder and melting a button I will know the exact yield better.
On the list it has the AMD 486 which is a much older chip and different construction. The stated yield of 0.12 sounds about right for this type of CPU.

From my batches I keep all the broken CPUs except the thin K6/Athlon types for later crushing and more leaching in AR. Hopefully,this will pull out a bit more gold from the bodies. I've read, for example the Pentium Pro, has about 0.1g locked up in the ceramic substrate that can only be reached by crushing the CPU body. Next year I'll try a second AR leach on these to see what I can pull out.

resabed01 said:

testerman said:

To resabed01: I meant to ask you earlier as to where did you get the yield of 0.035 from? It's not on any of the pdf files I have. I'm thinking those are your yields? If so, that yield amount has been the most accurate I've seen for these cpu's so far.

Unless someone can somewhat show a different yield data for these chips, I'm going to go with the yield of 0.035 for these cpu's I have refined.

On another note, I did read in that one pdf file that I have is that people miss alot of the gold hidden inside of components mainly because what they see on the surface is what they think all there is.

I should take about 1 lb worth of the broken cpu's, crush them up real fine and then process them and see what I can get. That way I will know for certain about if there's more gold in there or not, and if I actually missed most of the gold.

Oh.. one question to ask. Have any of you crushed your ceramic cpu's and recovered more gold than what a data chart said? or was it the same?

Thanks all for the input.

Kevin

Click to expand...

Kevin,

As pointed out by the others, the aluminum top AMD K6 is the top CPU in the picture and the lower one is the Athlon. I call them Durons but that's just me.
The yield of 0.035g for the K6 was calculated from my actual return of these chips. I ran several batches of ceramic CPUs this summer and collected my own yield data. I did not crush the AMD chips, I processed them whole after removing the silicon die and any MLCCs from the tops. I know, it's extra work but I wanted to process the MLCCs separately. The AMD Athlon has yield of slightly better than 0.04g each, less than the stated 0.11g but better than the K6. I just finished a batch of Athlon chips so once I finish washing the gold powder and melting a button I will know the exact yield better.
On the list it has the AMD 486 which is a much older chip and different construction. The stated yield of 0.12 sounds about right for this type of CPU.

From my batches I keep all the broken CPUs except the thin K6/Athlon types for later crushing and more leaching in AR. Hopefully,this will pull out a bit more gold from the bodies. I've read, for example the Pentium Pro, has about 0.1g locked up in the ceramic substrate that can only be reached by crushing the CPU body. Next year I'll try a second AR leach on these to see what I can pull out.

Click to expand...

Nice information resabed01. From what you stated so far, I have more precious metals in my solution because I never took off the MLCC's nor the glass center cap.

How do you remove them? I have a heat gun and I tried it to remove the pins before, and even with a torch and that just isn't happening, at least on an easy scale. Would a heat gun remove the silicon glass top and the MLCC's? Or do I have to use a chisel to bang them off?

I still have the Pentium Pro's and I'll crush them up and see what more I can get out of them too. Actually, the other cpu's I still have, many of the MLCC's are still in tact on some of them and some of them fell off the cpu's from the solution and I still have them too.

I'll keep my notes updated and I'll update them so that way I'll know from my own experience of the yields on the chips. I would love to get the approximate 0.6g extra from those Pentium Pro's. That would be an extra 6g gold if I did.

Thanks again for the information. The good thing is that I'm not confused about this at all, but I learned some things I was wondering about.

Kevin

I can only guess... but i believe that at least several hundreds of people have downloaded my list (which i have invested hours compiling) - and not even once i received any yield information from anyone.

Just an FYI... :|

testerman said:

It was told in two separate YouTube videos that there is approximately 3g of gold in each one.

Click to expand...

wow, it's went from the 1g lie to a 3g lie ? it used to be a big lie now its a huge lie (stupid is what stupid does I guess)

ilovegrf said:

testerman said:

It was told in two separate YouTube videos that there is approximately 3g of gold in each one.

Click to expand...

wow, it's went from the 1g lie to a 3g lie ? it used to be a big lie now its a huge lie (stupid is what stupid does I guess)

Click to expand...

Just like myself, probably never refined them and have no idea on the yields.

Also, probably people setting you up to buy THEIR cpu's, VOUCHING on the yields and they're/their totally incorrect. You live and you learn.

I hthink I have the YouTube videos on this computer. If I find them I'll post them. I do know one of them for sure I commented on it and told him that his calculations were way off. I even told him that he needed to reimburse me some money for cpu's because of his "video".

If/when I do a video on anything, I'm going to be correct on my yields. PERIOD!

Kevin