Old but new I200 Pentiums?

Well this my first question and post.

Other than my intro.

I am trying to evaluate some "scrap" and decide wether it is better to try and sell these old CPUs or crush them up and give a go at refining.

Or rather should I save them until I am set up to do a bit of refining and as I have a quantity of definite scrap its not like I am compelled to start crushing.

Other than the fact that they were probably manufactured and shipped in 1996 they are literally brand new and still in the box they came in.

Or at least 42 are like brand new one I've sort of scratched a bit for tests.

They have gold pins on one side and a white metal plate on the other

The plate does not react to the platinum test acid when scratched (I've got one of those ebay kits of acid) its hard to make a streak with the white metal but best I can tell the streak also shows no reaction.

Nor do the various golds testing solutions produce any noticeable reaction.

Here is how they are marked:

Pentium I200
FV80502200
SY044/VSS

So is the top plate platinum or am I reading the acids wrong?

Should I crush or is there a market somewhere for these old but brand new chips?


Thanks for any advice

Sebastian,

You've got what are called Pentium 200 MMX cpus. Most guys just call them fiber cpus because they are made from layers of laminated fibers. The heat spreader (silver plate on top) is copper plated with nickel. It comes off easily with a heat gun to expose a square of gold plated copper foil. The pins are kovar plated with nickel and gold. The core can be punched out as seen in my MMX core removal video on my website. The wires connecting the core to the housing are hair fine gold.

Steve

Gold wires are actually embedded in the black plastic covering the chip opposite of the copper plate.

Astro,

Yes, the connecting wires were still gold in these cpus. I have just taken this closeup photo to prove it:



here's a second shot that is slightly out of focus, but the gold wires are clearly visible:



The early PII's were made using these wires as well. PIII's and beyond were bonded directly to the core without wire interconnects in a ball grid array arrangement.

Steve

I don't understand what you are trying to address. I only made a statement that the wires are inside the plastic which is shown in your photos. You will have to crush the plastic to recover the gold from the wires.

I don't understand what you are trying to address.

I don't think he's trying to address your statement. He is only trying to educate for the less informed students and those who will find this thread months from now. :wink:

Astro,

I was confirming your statement for the general public with some photos. The photos just make it easier to understand exactly where in the cpu you are refering to.

Steve

Thanks for all the help - I am still wondering if its better to sell the chips or refine.

Any idea what the recovery of PMs might be?

Too hot to do anything right now.

I haven't put these cpus to the yield tests yet. I have several pounds of them myself and may get around to processing them soon. I'll add the data to the yield installments I've been posting.

Steve

lazersteve said:

Sebastian,

You've got what are called Pentium 200 MMX cpus. Most guys just call them fiber cpus because they are made from layers of laminated fibers. The heat spreader (silver plate on top) is copper plated with nickel. It comes off easily with a heat gun to expose a square of gold plated copper foil. The pins are kovar plated with nickel and gold. The core can be punched out as seen in my MMX core removal video on my website. The wires connecting the core to the housing are hair fine gold.

Steve

Click to expand...

Steve...

Did you conclusively test the heat spreader to confirm it is nickel plated onto copper or is this a speculation?

thank you.

I read the information from the manufacturers datasheet on the CPU.

I have uploaded the section of the official Intel datasheet to the documents section of my website ( http://www.goldrecovery.us ) that displays the heat spreader contents. Check the 'Intel CPU Package Data' document, page 1 Table 13-1, 'Package Body Material' heat slug construction for the PPGA cpu package (Plastic Pin Grid Array = Intel MMX CPU Family).

Steve

Thank you very much.

Quite a detailed response.

As this chip did not respond to a magnet, and also indicated positive on the platinum acid test.

What might be a good quick field method for differing between nickel and platinum?

Clean the area. Drop of 50/50 nitric - let it work; several drops (2, or 3 - you must neutralize the acid) of ammonia; drop of dimethyl gloxime (1% in alcohol or dist. water - it dissolves very, very slowly in water). Nickel will turn bright red. I hope I remember that right - I didn't look it up. If the nitric goes slow (cold outside), first heat the area a little bit with a lighter. It the part is thin, heat from underneath. Or, put it in a warm place and let it warm up first.

Two drops of muriatic acid and 1 drop of hydrogen peroxide will settle it. Apply the reagents to a clean cpu top and wait 5 minutes. Slightly buff the surface with the cotton swab. Nickle will dissolve as a blue green liquid where Platinum won't.

Here's the results that I got this morning from this test:



Do you have documentation that the cpu lids (heat slugs/spreaders) are plated with platinum which contradicts Intel's data sheet ?

Steve

goldsilverpro said:

Clean the area. Drop of 50/50 nitric - let it work; several drops (2, or 3 - you must neutralize the acid) of ammonia; drop of dimethyl gloxime (1% in alcohol or dist. water - it dissolves very, very slowly in water). Nickel will turn bright red. I hope I remember that right - I didn't look it up. If the nitric goes slow (cold outside), first heat the area a little bit with a lighter. It the part is thin, heat from underneath. Or, put it in a warm place and let it warm up first.

Click to expand...

Again my thanks - do you have preferred supplier - I've done a bit of a google but most suppliers do not seem to cater to a someone without corporate credentials.

If I read the wiki article correctly dimethylgloxime is also used to ID palladium - the article did not say what to look for however.

lazersteve said:

Two drops of muriatic acid and 1 drop of hydrogen peroxide will settle it. Apply the reagents to a clean cpu top and wait 5 minutes. Slightly buff the surface with the cotton swab. Nickle will dissolve as a blue green liquid where Platinum won't.

....Imaged snipped.....

Do you have documentation that the cpu lids (heat slugs/spreaders) are plated with platinum which contradicts Intel's datasheet ?

Steve

Click to expand...

No I stated the test acid in my kit showed positive for platinum I did not mean t make a blanket statement that platinum was present.

- I am quite the neophyte but I do understand that one must understand the mechanisms of the test to understand the result and I don't

- everyone has been of immense help.

lazersteve said:

Two drops of muriatic acid and 1 drop of hydrogen peroxide will settle it. Apply the reagents to a clean cpu top and wait 5 minutes. Slightly buff the surface with the cotton swab. Nickle will dissolve as a blue green liquid where Platinum won't.

Click to expand...

Excellent sounds much easier to obtain HCL 30% and hydrogen peroxide - is that 3% Hydrogen peroxide?

thanks Steve

Sebastain said:

As this chip did not respond to a magnet, ...

Click to expand...

I observed that the CPU heat spreader DOES respond to a magnet as seen here:




Steve

lazersteve said:

Sebastain said:

As this chip did not respond to a magnet, ...

Click to expand...

I observed that the CPU heat spreader DOES respond to a magnet as seen here:

[img:632:537]http://www.goldrecovery.us/images/cpu_magnet.jpg[/img]


Steve

Click to expand...

Thanks for the detailed reply Steve.
I took 4 pieces of clean heat spreaders and dissolved them in in diluted AR. I observed that the shiny metal ( nickel ? ) disolved early and completely before any reaction on the copper which was pinkish in color.
The solution became black on contact with stannous chloride which led me to believe it was gold or palladium.
I checked for gold and none was present . I have not done any other conclusive tests.
In retrospect and in view of your info , could it be that the solder contains a PGM ( perhaps palladium ) ?.

Sebastian said:

Excellent sounds much easier to obtain HCL 30% and hydrogen peroxide - is that 3% Hydrogen peroxide?

Click to expand...

Yes, that is 3% household peroxide.

Steve