Gold content of Pentium Pros

[Alentia]

I confirm, last row is total output. 0.25 is output per lb.

Thank you for taking the time to format the data.

 

[Alentia]

More CPU processing data. Since it was a batch, the data seems vary from data posted before. But it should can be taken as per minimum values:

Sorry guys, I do not know how are formatting the data works

	Pounds	Total gram	Quantity	Piece gram	Output piece	Output lb	Total Output
							
Ppro-256K Cache	0.39	176	2	88	0.25	1.29	0.5
Mix broken	0.56	256	10	26	0.07	1.24	0.7
486DX	0.78	354	15	24	0.18	3.46	2.7
Cyrix 6x86	0.35	160	4	40	0.2	2.27	0.8
Motorola	0.05	24	1	24	0.07	1.32	0.07
Dec	0.14	62	1	62	0.28	2.05	0.28
AMD Athlon	0.43	194	11	18	0.05	1.29	0.55
Intel Pentium	1.15	520	17	31	0.05	0.74	0.85
AMD K6	0.29	132	6	22	0.07	1.44	0.42
Sun TurboXGX	0.07	32	1	32	0.3	4.25	0.3
386 + Mot	0.10	46	3	15	0.06	1.77	0.18
Varios yield 286-10 Cyrix cap	0.09	40	6	7	0.03	2.04	0.18
386 AMD solder	0.05	22	1	22		0.00	0
AMD K5 gold top	0.17	76	2	38	0.39	4.66	0.78
AMD K5 gold bottom	0.06	26	1	26	0.39	6.80	0.39
P90	0.10	46	1	46	0.38	3.75	0.38
Total Estimated	4.78	2166	82				9.08

 

[necromancer]

export file as text and open, then copy and paste into GRF post, notepad might give you trouble (microsoft)

edit: i don't understand the names + list names + final numbers
they make no sence to me unless i missed something ?
and if you use or download this data the least you can do is say thank you to "Alentia"

 

[necromancer]

There was lots of good stuff posted in the early days. Except for stuff like auctions, nothing on here is timely. I like seeing the old stuff dredged up. Shows people are searching.

no searching, i read every post as a new member ")

 

[AUH-R]

Hi Alentia,

Good work on the data, how can you be sure of each processor type if you processed them together? I look forward to comparing data later on in the year.

Best wishes,

 

[Alentia]

Some data is already known and some went through tweaks. You can use this data to determine minimum values. So if you buy the CPU, you do not overpay and if you get about the same yield you did it right.

Taking Cyrix/IBM 6x86 CPUs for example. Not sure how people claim them to be over 0.25. They consistently yield 0.2 per CPU on average in a batch of alone.
486s are consistent all the time with 0.18 yield per CPU.

 

[lazersteve]

I don't get anywhere near your results on the Cyrix/IBM. There are several varieties, the ones I run are the ones with the black bottom covers. They run 1.2 to 1.4 grams per pound average from mixed Cyrix, IBM, and VIA mixed gold tops.

Steve

 

[AUH-R]

Some data is already known and some went through tweaks. You can use this data to determine minimum values. So if you buy the CPU, you do not overpay and if you get about the same yield you did it right.

Taking Cyrix/IBM 6x86 CPUs for example. Not sure how people claim them to be over 0.25. They consistently yield 0.2 per CPU on average in a batch of alone.
486s are consistent all the time with 0.18 yield per CPU.

Thank you, I do find it helpful.

Best wishes,

 

[Alentia]

I have obtained P-Pro with 1Mb Cache (Fiber) for the experiment. Will post results once become available with some pictures when possible.

 

[Alentia]

1. I have removed the aluminum top from the CPU
2. I have taken out Cache/Microchip with aluminum backing and placed into HCl - aluminum backing dissolved within about 30 min
3. I have removed bottom layer of fiber to expose more PM. Pins are somewhat just a bit magnetic.
4. I have split CPU in to 2 halfs. One to be treated with Nitric and another in AP
5. Dropped Cache/Microchip into AP seprately

 

[Alentia]

The first interesting discovery Pentium Pro 1Mb Cache has no silver content or very insignificant.
It looks like the best way to process them is with AP.

On the other hand, following will help to calculate amount of Cu content in Pentium Pro 512K Cache

Heatspreader weights 27.5 gram
Naked CPU without Pins and Heatspreader - 44 gram
Top gold foil - 0.085 gram
Considering P-Pro has about 0.5 of silver
Cu, Sn + anything else = 88gr - (27.5 + 44 + 0.5 + 0.5+5(bottom ceramic covering)) = about 10.5 gram
That would require about 45ml of HNO3 per CPU to digest base materials in an open vessel system (beaker, bucket, etc where NOx is to escape freely)
I have used 750ml for 26 CPUs or about 30 ml per CPU using NOx condenser on exhaust and it was a bit excessive.

I have decided to deep the HNO3 treated half into concentrated H2SO4 to dissolve SnO2 before moving it to AR.

 

[cashtronx]

Hello all. Newbie here. Recently started recycling electronics for gold recovery. Great stuff here! For all PP, the amount of gold in them is based upon the size of it; 256KB, 512KB, 1MB.... There is little to none new info on the web about PP and its technology, only old stuff. I think they are still used in supercomputers and the government has bought it out. lol Just a theory!

 

[AUH-R]

Alentia, I think your contribution to this forum is excellent and I really enjoy reading your informative posts and seeing the pictures you provide.

Thank you & best wishes,
AuH-R

 

[Alentia]

I have a suspicion that gold from both halves got plated on what looks like some kind of four "most likely copper sticks". First AR treatment resulted in very green (blue+yellow) solution. Most likely copper from those sticks. What I saw in precipitation, I can only make visual estimate about 0.01 or 0.02 grams from each half and about 0.1 from cache. I will post actual data when gold is washed.
I have decided to place both halves in one beaker and treat with AR again. I will post results from second AR treatment.

 

[Alentia]

To all the disappointment fiber 1Mb Cache P-Pro yields only 0.1g of gold

0.08 for Cache/Microchip module
0.02 for pins and inside body plating

Yet to try Ceramic...

 

[Alentia]

Some more interesting data... Not all 512K created equal.

Processed 25 x 512K Cache Pentium Pro chips.

Output 10.5 gram

Meaning: Some chips produced 0.5 gram as expected and some only 0.25 gram. Now it is too late to analyze the difference.

following can apply: 17 x 0.5 + 8 x 0.25 = 10.5 gram

Some 8 512K Cache were odd one's.